Hana Dev en

PUBLIC
SAP HANA Appliance Software SPS 05

Document Version: 1.2 - 2013-06-19
SAP HANA Developer Guide
Table of Contents
1 Document History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2 Introduction to SAP HANA Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3 SAP HANA Architecture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.1 SAP HANA In-Memory Database. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.1.1 Columnar Data Storage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.1.2 Parallel Processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13
3.1.3 Simplifying Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.2 SAP HANA Database Architecture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.3 SAP HANA Extended Application Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.4 SAP HANA-Based Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4 SAP HANA Development Platform. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.1 Developer Scenarios. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.1.1 Scenario: Developing Native SAP HANA Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.1.2 Scenario: Using Database Client Interfaces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.2 Development Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .22
4.3 Repository. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.4 SAP HANA Studio. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.4.1 Modeler Perspective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.4.2 SAP HANA Development Perspective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.4.3 Debug Perspective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.4.4 Administration Console Perspective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.5 Getting Started. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.5.1 Adding a System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
5 Setting Up Your Application. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.1 Before you Start. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.2 Setting up Delivery Units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.2.1 Maintaining the Delivery-Unit Vendor ID. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.2.2 SAP HANA Delivery Unit Naming Conventions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.2.3 Creating a Delivery Unit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
5.3 Using SAP HANA Projects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5.3.1 SAP HANA Repository: Workspaces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.3.2 Creating a Repository Workspace. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.3.3 SAP HANA Studio Projects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
5.3.4 Creating a Project for SAP HANA XS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.3.5 Sharing a Project for SAP HANA XS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45
5.3.6 Importing a Project in SAP HANA XS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2
PUBLI C
2013 SAP AG or an SAP affiliate company. All rights reserved.

Table of Contents
5.4 Maintaining Repository Packages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
5.4.1 SAP HANA Repository Packages and Namespaces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48
5.4.2 Defining Repository Package Privileges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.4.3 Creating a Package. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
5.4.4 Defining a Package Hierarchy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5.5 Creating the Application Descriptors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.5.1 The SAP HANA XS Application Descriptor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.5.2 Create an Application Descriptor File. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55
5.5.3 The Application-Access File. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
5.5.4 Application-Access File Keyword Options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
5.5.5 Application-Access URL Rewrite Rules. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.5.6 Enabling Access to SAP HANA XS Application Packages. . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
5.5.7 The Application-Privileges File. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.5.8 Create an SAP HANA XS Application Privileges File. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.6 Tutorial: My First SAP HANA Application. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.6.1 Open the Development Perspective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72
5.6.2 Add a System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.6.3 Add a Workspace. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75
5.6.4 Add a Project. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77
5.6.5 Share Your Project. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78
5.6.6 Write Server-Side JavaScript. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.6.7 Retrieve Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
6 Setting Up the Persistence Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
6.1 Schema. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
6.2 Creating Schemas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.3 Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.4 Table Configuration Schema. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87
6.5 Creating Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
6.6 Sequences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
6.7 Creating Sequences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
6.8 SQL Views. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.9 Creating SQL Views. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
6.10 Data Provisioning Using Table Import. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
6.10.1 Table-Import Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
6.10.2 Table-Import Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
6.10.3 Table-Import Extensions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
6.10.4 Table-Import Configuration-File Syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
6.11 Importing Data Using Table Import. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
6.12 Using Imported Table Data in SAP HANA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
6.13 Using Imported Table Data in an SAP HANA XS Application. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.14 Extending a Table Import. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
Table of Contents
PUBLI C
2013 SAP AG or an SAP affiliate company. All rights reserved. 3

7 Setting Up the Analytic Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
7.1 Setting Up the Modeling Environment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
7.1.1 Setting Modeler Preferences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109
7.1.2 Configuring the Import Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
7.1.3 Importing Table Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
7.1.4 Loading Data into Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
7.1.5 Copying Content Delivered by SAP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
7.1.6 Mapping the Authoring Schema to the Physical Schema. . . . . . . . . . . . . . . . . . . . . . . . . . . 116
7.1.7 Generating Time Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
7.2 Creating Views. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
7.2.1 Attributes and Measures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
7.2.2 Attribute Views. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
7.2.3 Creating Attribute Views. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
7.2.4 Analytic Views. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
7.2.5 Creating Analytic Views. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
7.2.6 Calculation Views. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
7.2.7 Creating Calculation Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
7.2.8 Assigning Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
7.2.9 Creating Input Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
7.2.10 Creating Hierarchies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .148
7.2.11 Using Currency and Unit of Measure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
7.2.12 Activating Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
7.3 Creating Decision Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
7.3.1 Changing the Layout of a Decision Table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
7.3.2 Using Parameters in a Decision Table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
7.4 Managing Object Versions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
7.4.1 Switching Ownership of Inactive Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
7.4.2 Toggling Versions of Content Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
7.4.3 Viewing Version History of Content Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
7.5 Working with Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
7.5.1 Managing Layout. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
7.5.2 Filtering Packages and Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
7.5.3 Refactoring Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
7.5.4 Validating Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
7.5.5 Generating Object Documentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
7.5.6 Enabling Multilanguage Support for Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
7.5.7 Checking Model References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
7.5.8 Viewing the Job Log. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
7.5.9 Maintaining Search Attributes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
7.5.10 Previewing Data of Content Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
7.5.11 Functions used in Expressions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
4
PUBLI C

Table of Contents
7.5.12 Searching Tables, Models and Column Views. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
7.5.13 Setting Keyboard Shortcuts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
7.5.14 Copying an Object. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
7.6 Importing BW Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178
8 Developing Procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
8.1 Editing SQLScript. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .184
8.1.1 Defining Local Table Types in Procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186
8.2 Debugging SQLScript. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
8.3 Developing Procedures in the Modeler Editor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
9 Defining Web-based Data Access. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
9.1 Data Access with OData in SAP HANA XS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
9.1.1 OData in SAP HANA XS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
9.1.2 Defining the Data an OData Service Exposes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192
9.1.3 OData Service Definitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
9.1.4 Creating an OData Service Definition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
9.1.5 Tutorial: Using the SAP HANA OData Interface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
9.1.6 OData Service-Definition Examples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
9.1.7 OData Service Definition Language Syntax. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
9.1.8 OData Service Definition: SQL-EDM Type Mapping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
9.1.9 OData URI Parameters, Query Options, and Features. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
9.1.10 OData Security Considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
9.2 Data Access with XMLA in SAP HANA XS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .230
9.2.1 XML for Analysis (XMLA). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
9.2.2 XMLA Service Definition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
9.2.3 XMLA Security Considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
9.2.4 Multidimensional Expressions (MDX). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
9.2.5 MDX Functions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
9.2.6 MDX Extensions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
9.2.7 Defining the Data an XMLA Service Exposes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .239
9.2.8 Creating an XMLA Service Definition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
9.2.9 Tutorial: Using the SAP HANA XMLA Interface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
10 Writing Server-Side JavaScript Code. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
10.1 Data Access with JavaScript in SAP HANA XS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
10.2 Server-Side JavaScript in SAP HANA XS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
10.2.1 JavaScript Editor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
10.2.2 Server-Side JavaScript Security Considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
10.2.3 Writing Server-Side JavaScript Application Code. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
10.3 Server-Side JavaScript Libraries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
10.3.1 Writing Server-Side JavaScript Libraries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Table of Contents
PUBLI C

10.3.2 Importing Server-Side JavaScript Libraries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
10.4 Server-Side JavaScript APIs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
10.5 The SQL Connection Configuration File. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
10.6 Connection-language Settings in SAP HANA XS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
10.7 Server-Side JavaScript Tracing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
10.7.1 Tracing Server-Side JavaScript Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
10.7.2 Viewing Server-Side JavaScript Application Trace Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
10.8 Debugging Server-Side JavaScript. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
10.8.1 Opening a Port for Server-Side JavaScript Debugging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
10.8.2 Troubleshooting Server-Side JavaScript Debugging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
11 Building UIs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
11.1 Building UIs with SAPUI5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .271
11.1.1 Installing SAPUI5 Application Development Tools. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
11.1.2 Creating an SAPUI5 Application. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
11.1.3 Supporting Translation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .276
11.2 Using UI Integration Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
11.2.1 Creating an Application Site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
11.2.2 Designing an Application Site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
11.2.3 Creating a Widget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
11.2.4 Developing Widgets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
12 Enabling Search. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
12.1 Creating Full Text Indexes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
12.1.1 Full Text Index Types. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
12.1.2 Synchronization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
12.1.3 Text Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .298
12.1.4 Dropping Full Text Indexes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
12.1.5 Altering Full Text Index Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .303
12.1.6 Full Text Index Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
12.2 Building SQL Search Queries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
12.2.1 Search Queries with CONTAINS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
12.2.2 EXACT Search. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
12.2.3 LINGUISTIC Search. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
12.2.4 FUZZY Search. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
12.3 Building Search Apps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
12.3.1 Introduction to the UI Toolkit for Info Access. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
12.3.2 Installing the Service and the Toolkit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
12.3.3 Getting to Know the Demo App. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
12.3.4 Getting to Know the Demo HTML. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
12.3.5 Preparing Your Source Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
12.3.6 Connecting Your Source Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .402
6
PUBLI C

Table of Contents
12.3.7 Defining the Page Layout. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
12.3.8 Configuring the Widgets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .404
12.3.9 Defining the Layout of Result Lists and Details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
13 Setting Up Roles and Authorizations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
13.1 The Authorization Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409
13.2 Authentication Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .410
13.3 Roles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
13.3.1 Roles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
13.3.2 Roles as Repository Objects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
13.3.3 Creating Roles in the Repository. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
13.4 Privileges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
13.4.1 System Privileges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .418
13.4.2 Object Privileges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418
13.4.3 Package Privileges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
13.4.4 Analytic Privileges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422
13.4.5 Creating Analytic Privileges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
13.4.6 Granting Privileges to Users. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
13.5 Application Access. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
14 Implementing Lifecycle Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
14.1 SAP HANA Delivery Units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
14.2 The SAP HANA Delivery-Unit Lifecycle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
14.3 Exporting Delivery Units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
14.4 Importing Delivery Units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
14.5 Translating Delivery Units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .444
14.5.1 SAP HANA Delivery-Unit Translation Details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
14.5.2 Maintaining Translation Details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
15 Using Database Client Interfaces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
15.1 Connecting via ODBC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447
15.1.1 Using the User Store. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
15.1.2 Testing the ODBC Installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
15.2 Connecting via JDBC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
15.2.1 Tracing JDBC Connections. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
15.2.2 Valid Java-to-SQL Conversions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
15.3 Connecting via ODBO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .454
15.3.1 Connecting with Microsoft Excel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454
15.3.2 Multidimensional Expressions (MDX). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
15.3.3 MDX Functions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
15.3.4 MDX Extensions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462
Table of Contents
PUBLI C

16 SAP HANA Developer References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
8
PUBLI C

Table of Contents
1 Document History
The document history includes all versions of the document that have been published.
Version Date SAP HANA Revision Description
1.1 21 Dec 2012 47 References to the
following have been
added:
Sections 7.2.3,
Creating Attribute
Views and 7.2.5,
Creating Analytic
Views have been
updated for Label
Mapping
functionality.
Section 7.6, Importing
BW Objects has been
updated with the
support to import role
based authorizations
into the Modeler as
analytic privileges.
Section 8.1.1, Defining
Local Table Types in
Procedures.
Section 11.2, Using UI
Integration Services.
1.2 08 Feb 2013 50 References to the
following have been
added:
Section 7.1.1
Generating Time Data
has been revised.
Section 7.2.5 Creating
Analytic Views has
been revised with the
addition of
documentation
regarding Temporal
Join.
Section 7.5.13 Setting
Keyboard Shortcuts
has been updated
with the change in a
Document History
PUBLI C

Version Date SAP HANA Revision Description
keyboard shortcut for
Data Preview.
Section 7.6 Importing
BW Objects has been
updated.
Section 8.2
Debugging SQLScript
has been updated
with a prerequisite
describing how to
grant debugger
system privileges.
Section 11. 2 Using UI
Integration Services
has been revised and
updated with the
following changed
features:
The design panel
has a new look
and feel, and
additional
capabilities, such
as search for
widgets,
configuration of
site layout, and
improved edit
mode.
The GadgetPrefs
API enables
personalization of
widgets by
storing widget
preferences
separately for
each user.
Section 13.4.4
Analytic Privileges has
been revised.
10
PUBLI C

Document History
2 Introduction to SAP HANA Development
This guide presents a developers view of SAP HANA
. It explains how to use the SAP HANA development tools to

create comprehensive analytical models and to build applications with SAP HANA's programmatic interfaces and
integrated development environment, including to develop native code that runs inside SAP HANA.
The guide is organized as follows:
SAP HANA Architecture [page 12] describes the basic capabilities and architecture of SAP HANA.
Developer Scenarios [page 19] describes the main developer scenarios for which you can use SAP HANA to
develop applications.
#unique_5 shows you how to get started building a simple SAP HANA-based application, including how to use
the SAP HANA studio tools and work with the SAP HANA repository.
Most of the remaining chapters explain how to develop various SAP HANA development objects that you can
include in your SAP HANA application.
Note
For the latest information about the availability of features for SAP HANA Extended Application Services (SAP
HANA XS) and related development tools, please see SAP Note 1779803.
Audience
This guide is aimed at two developer roles:
Modeler: This person, often a business/data analyst or database expert, is concerned with the definition of
the model and schemas that will be used in SAP HANA, the specification and definition of tables, views,
primary keys, indexes, partitions and other aspects of the layout and inter-relationship of the data in SAP
HANA.
The data modeler is also concerned with designing and defining authorization and access control, through the
specification of privileges, roles and users.
The modeler generally uses the Administration Console and Modeler perspectives and tools of the SAP HANA
studio.
Application Programmer: The programmer is concerned with building SAP HANA applications, which could
take many forms but are designed based on the model-view-controller architecture. Programmers develop
the code for:
View, which could run inside a browser or on a mobile device
Controller, which typically runs in the context of an application server
Model, which interacts closely with the data model, performs efficient queries, and may be developed to
run within the SAP HANA data engine, using embedded procedures or libraries
The programmer generally uses the SAP HANA Development perspective and tools of the SAP HANA studio.
Introduction to SAP HANA Development
PUBLI C

3 SAP HANA Architecture
SAP HANA is an in-memory data platform that is deployable on premise or on demand. At its core, it is an
innovative in-memory relational database management system that makes full use of the capabilities of current
hardware to increase application performance, reduce cost of ownership, and enable new scenarios and
applications that were not possible before.
With SAP HANA, you can build applications that integrate the business control logic and the database layer with
unprecedented performance. As a developer, one of the key questions is how you can minimize data movements.
The more you can do directly on the data in memory next to the CPUs, the better the application will perform.
3.1 SAP HANA In-Memory Database
SAP HANA runs on multi-core CPUs with fast communication between processor cores, and containing terabytes
of main memory. With SAP HANA, all data is available in main memory, which avoids the performance penalty of
disk I/O. Either disk or solid-state drives are still required for permanent persistency in the event of a power failure
or some other catastrophe. This does not slow down performance, however, because the required backup
operations to disk can take place asynchronously as a background task.
3.1.1 Columnar Data Storage
A database table is conceptually a two-dimensional data structure organized in rows and columns. Computer
memory, in contrast, is organized as a linear structure. A table can be represented in row-order or column-order.
A row-oriented organization stores a table as a sequence of records. Conversely, in column storage the entries of
a column are stored in contiguous memory locations. SAP HANA supports both, but is particularly optimized for
column-order storage.
Columnar data storage allows highly efficient compression. If a column is sorted, often there are repeated
adjacent values. SAP HANA employs highly efficient compression methods, such as run-length encoding, cluster
coding and dictionary coding. With dictionary encoding, columns are stored as sequences of bit-coded integers.
That means that a check for equality can be executed on the integers; for example, during scans or join
operations. This is much faster than comparing, for example, string values.
12
PUBLI C

SAP HANA Architecture
Columnar storage, in many cases, eliminates the need for additional index structures. Storing data in columns is
functionally similar to having a built-in index for each column. The column scanning speed of the in-memory
column store and the compression mechanisms especially dictionary compression allow read operations with
very high performance. In many cases, it is not required to have additional indexes. Eliminating additional indexes
reduces complexity and eliminates the effort of defining and maintaining metadata.
3.1.2 Parallel Processing
SAP HANA was designed to perform its basic calculations, such as analytic joins, scans and aggregations in
parallel. Often it uses hundreds of cores at the same time, fully utilizing the available computing resources of
distributed systems.
With columnar data, operations on single columns, such as searching or aggregations, can be implemented as
loops over an array stored in contiguous memory locations. Such an operation has high spatial locality and can
efficiently be executed in the CPU cache. With row-oriented storage, the same operation would be much slower
because data of the same column is distributed across memory and the CPU is slowed down by cache misses.
Compressed data can be loaded into the CPU cache faster. This is because the limiting factor is the data transport
between memory and CPU cache, and so the performance gain exceeds the additional computing time needed for
decompression.
Column-based storage also allows execution of operations in parallel using multiple processor cores. In a column
store, data is already vertically partitioned. This means that operations on different columns can easily be
processed in parallel. If multiple columns need to be searched or aggregated, each of these operations can be
assigned to a different processor core. In addition, operations on one column can be parallelized by partitioning
the column into multiple sections that can be processed by different processor cores.
PUBLI C

3.1.3 Simplifying Applications
Traditional business applications often use materialized aggregates to increase performance. These aggregates
are computed and stored either after each write operation on the aggregated data, or at scheduled times. Read
operations read the materialized aggregates instead of computing them each time they are required.
With a scanning speed of several gigabytes per millisecond, SAP HANA makes it possible to calculate aggregates
on large amounts of data on-the-fly with high performance. This eliminates the need for materialized aggregates
in many cases, simplifying data models, and correspondingly the application logic. Furthermore, with on-the fly
aggregation, the aggregate values are always up-to-date unlike materialized aggregates that may be updated only
at scheduled times.
3.2 SAP HANA Database Architecture
A running SAP HANA system consists of multiple communicating processes (services). The following shows the
main SAP HANA database services in a classical application context.
Such traditional database applications use well-defined interfaces (for example, ODBC and JDBC) to
communicate with the database management system functioning as a data source, usually over a network
14
PUBLI C

connection. Often running in the context of an application server, these traditional applications use Structured
Query Language (SQL) to manage and query the data stored in the database.
The main SAP HANA database management component is known as the index server, which contains the actual
data stores and the engines for processing the data. The index server processes incoming SQL or MDX
statements in the context of authenticated sessions and transactions.
The SAP HANA database has its own scripting language named SQLScript. SQLScript embeds data-intensive
application logic into the database. Classical applications tend to offload only very limited functionality into the
database using SQL. This results in extensive copying of data from and to the database, and in programs that
slowly iterate over huge data loops and are hard to optimize and parallelize. SQLScript is based on side-effect free
functions that operate on tables using SQL queries for set processing, and is therefore parallelizable over multiple
processors.
In addition to SQLScript, SAP HANA supports a framework for the installation of specialized and optimized
functional libraries, which are tightly integrated with different data engines of the index server. Two of these
functional libraries are the SAP HANA Business Function Library (BFL) and the SAP HANA Predictive Analytics
Library (PAL). BFL and PAL functions can be called directly from within SQLScript.
SAP HANA also supports the development of programs written in the R language.
SQL and SQLScript are implemented using a common infrastructure of built-in data engine functions that have
access to various meta definitions, such as definitions of relational tables, columns, views, and indexes, and
definitions of SQLScript procedures. This metadata is stored in one common catalog.
The database persistence layer is responsible for durability and atomicity of transactions. It ensures that the
database can be restored to the most recent committed state after a restart and that transactions are either
completely executed or completely undone.
The index server uses the preprocessor server for analyzing text data and extracting the information on which the
text search capabilities are based. The name server owns the information about the topology of SAP HANA
system. In a distributed system, the name server knows where the components are running and which data is
located on which server. The statistics server collects information about status, performance and resource
consumption from the other servers in the system. Monitoring clients, such as the SAP HANA studio, access the
statistics server to get the status of various alert monitors. The statistics server also provides a history of
measurement data for further analysis.
Related Links
SAP HANA SQLScript Reference
SAP HANA Business Function Library (BFL) Reference
SAP HANA Predictive Analysis Library (PAL) Reference
SAP HANA R Integration Guide
3.3 SAP HANA Extended Application Services
Traditional database applications use interfaces such as ODBC and JDBC with SQL to manage and query their
data. The following illustrates such applications using the common Model-View-Controller (MVC) development
architecture.
PUBLI C

SAP HANA greatly extends the traditional database server role. SAP HANA functions as a comprehensive
platform for the development and execution of native data-intensive applications that run efficiently in SAP HANA,
taking advantage of its in-memory architecture and parallel execution capabilities.
By restructuring your application in this way, not only do you gain from the increased performance due to the
integration with the data source, you can effectively eliminate the overhead of the middle-tier between the user-
interface (the view) and the data-intensive control logic, as shown in the following figure.
In support of this data-integrated application paradigm, SAP HANA Extended Application Services provides a
comprehensive set of embedded services that provide end-to-end support for Web-based applications. This
includes a lightweight web server, configurable OData support, server-side JS execution and, of course, full
access to SQL and SQLScript.
These SAP HANA Extended Application Services are provided by the SAP HANA XS server, which provides
lightweight application services that are fully integrated into SAP HANA. It allows clients to access the SAP HANA
system via HTTP. Controller applications can run completely natively on SAP HANA, without the need for an
additional external application server.The following shows the SAP HANA XS server as part of the SAP HANA
system.
16
PUBLI C

The application services can be used to expose the database data model, with its tables, views and database
procedures, to clients. This can be done in a declarative way using OData services or by writing native application-
specific code that runs in the SAP HANA context . Also, you can use SAP HANA XS to build dynamic HTML5 UI
applications.
In addition to exposing the data model, SAP HANA XS also hosts system services that are part of the SAP HANA
system. The search service is an example of such a system application. No data is stored in the SAP HANA XS
server itself. To read tables or views, to modify data or to execute SQLScript database procedures and
calculations, it connects to the index server (or servers, in case of a distributed system).
Note
For the latest information about the availability of features for SAP HANA Extended Application Services (SAP
HANA XS) and related development tools, please see SAP Note 1779803.
Related Links
Building UIs with SAPUI5 [page 271]
This section provides introductory information about UI development toolkit for HTML5.
Enabling Search [page 291]
With a SAP HANA database, your users will want to search tables and views much like they would when searching
for information on the Internet. In SAP HANA, you can either directly query data using SQL queries or you can
build search apps using a UI toolkit.
PUBLI C

3.4 SAP HANA-Based Applications
The possibility to run application-specific code in SAP HANA raises the question: What kind of logic should run
where? Clearly, data-intensive and model-based calculations must be close to the data and, therefore, need to be
executed in the index server, for instance, using SQLScript or the code of the specialized functional libraries.
The presentation (view) logic runs on the client for example, as an HTML5 application in a Web browser or on a
mobile device.
Native application-specific code, supported by SAP HANA Extended Application Services, can be used to provide
a thin layer between the clients on one side, and the views, tables and procedures in the index server on the other
side. Typical applications contain, for example, control flow logic based on request parameters, invoke views and
stored procedures in the index server, and transform the results to the response format expected by the client.
The communication between the SAP HANA XS server and index server is optimized for high performance.
However, performance is not the only reason why the SAP HANA XS server was integrated into SAP HANA. It also
leads to simplified administration and a better development experience.
The SAP HANA XS server completes SAP HANA to make it a comprehensive development platform. With the SAP
HANA XS server, developers can write SAP HANA-based applications that cover all server-side aspects, such as
tables and database views, database procedures, server-side control logic, integration with external systems, and
provisioning of HTTP-based services. The integration of the SAP HANA XS server into the SAP HANA system also
helps to reduce cost of ownership, as all servers are installed, operated and updated as one system.
18
PUBLI C

4 SAP HANA Development Platform
To simplify and support the development of SAP HANA-based applications, SAP HANA serves as the integrated
development environment (IDE) for developing and delivering these SAP HANA applications.
Built around the Eclipse-based SAP HANA studio the SAP HANA IDE supports integrated and collaborative
development, debugging and deployment of applications that use native database procedures for data
processing, server-side JavaScript for control and an HTML5 SDK for the development of user interface
representation.
The design-time environment's central component is the SAP HANA repository, which stores and manages all
design-time objects. The following figure illustrates the principle of co-development with SAP HANA.
Using the SAP HANA studio on your workstation, you design and create development objects such as data models
or server-side code files, and then store, manage and share them with other developers, by interacting with the
SAP HANA repository. The repository enables teams to work together on a set of development objects, and
ultimately turn them into runtime objects which can be used by clients or other applications. The repository
provides the basis for concepts like namespaces (through packages), transport and delivery support.
4.1 Developer Scenarios
The possibility to run application specific code in SAP HANA creates several possibilities for developing SAP
HANA based applications, representing various integration scenarios, and corresponding development
processes.
Broadly, we distinguish SAP HANA based applications into two broad categories:
Web-based scenarios that take full advantage of the SAP HANA Extended Application Services. In these
scenarios, clients access SAP HANA data using standard OData or XMLA interfaces, or directly use a Web-
based GUI that was developed using the SAPUI5 toolkit, and that uses custom-developed server-side
JavaScript, as well as native SQLScript procedures.
Traditional client-based scenarios, where an external application accesses the SAP HANA data model (tables,
analytic views, etc.) via client interfaces such as ODBC, ODBO and JDBC, and only uses SQL and native
SQLScript procedures.
In either case, as a developer you need to understand the SAP HANA development environment, which enables
you to design and develop your data and analytical models, and your other HANA-based development objects in
the form of portable and deployable delivery units.
Common activities include:
SAP HANA Development Platform
PUBLI C

Table 1: Common Activities
Activity More Information
Setting up your development environment. Setting Up Your Application [page 37]
Setting up your schema and tables. Setting Up the Persistence Model [page 83]
Setting up views of your data. Setting Up the Analytic Model [page 109]
Developing procedures for data-intensive logic. Developing Procedures [page 183]
Setting Up Roles and Authorizations. The Authorization Model [page 409]
Managing the delivery of your application. SAP HANA Delivery Units [page 440]
For these activites, you will want to keep handy the following references:
SAP HANA SQL Reference, to help you writing SQL statements within procedures and from your server-side
JavaScript.
SAP HANA SQLScript Reference, to help you if you are writing procedures.
The next two sections describes the main scenarios and what activities you may need to perform for them.
4.1.1 Scenario: Developing Native SAP HANA Applications
Here, you want to create a Web-based scenario that takes full advantage of SAP HANA Extended Application
Services. In this scenario, clients access SAP HANA data using standard OData or XMLA interfaces, or directly use
a Web-based GUI that was developed using the SAPUI5 toolkit, and that uses custom-developed server-side
JavaScript, as well as native SQLScript procedures.
For this scenario, you may need to perform the following activities:
20
PUBLI C

Table 2: Developing Native SAP HANA Applications
Develop server-side JavaScript for control logic. Writing Server-Side JavaScript Code [page 243]
Define OData data interfaces. Data Access with OData in SAP HANA XS [page 191]
Define XMLA data interfaces. Data Access with XMLA in SAP HANA XS [page 230]
Build HTML pages with SAPUI5. Building UIs with SAPUI5 [page 271]
You will also want to keep handy the following references:
SAP HANA SQL Reference, to help you write SQL statements within procedures and from your server-side
JavaScript.
SAP HANA XS JavaScript Reference, to help you use the SAP HANA XS JavaScript API.
4.1.2 Scenario: Using Database Client Interfaces
Here, you want to build an application outside of SAP HANA, for example, within SAP NetWeaver, that accesses
the SAP HANA data model (for example, tables and analytic views) via client interfaces such as ODBC, ODBO and
JDBC, and only uses SQL and native SQLScript procedures.
For this scenario, you may need to perform the following activities:
Table 3: Using Database Client Interfaces
Install and use the SAP HANA client interfaces. Using Database Client Interfaces [page 447]
PUBLI C

Develop procedures for data-intensive logic. Developing Procedures [page 183]
You will also want to keep handy several references:
SAP HANA SQL Reference, to help you write SQL statements.
For information on MDX in SAP HANA, see MDX Functions [page 233].
SAP HANA SQLScript Reference, to help you if you are writing procedures.
4.2 Development Objects
The design-time building blocks of your SAP HANA applications are called development objects. Some of these
development objects, such as projects and packages, are concepts that help you structure your application.
Others, like schemas, table definitions, analytical and attribute views, help you organize the structure of your data.
Procedures and server-side JavaScript code are the backbone of the SAP HANA application. Other types of
development objects help you control the access to runtime objects.
Understanding the different development objects and their use is one of the main objectives of this guide.
The following are the building blocks for an SAP HANA application, showing the file extension and where to get
information for building it:
Table 4: Application Building Blocks
Object Description File Extension More Information
Structure
Project An Eclipse project for developing your
application or part of an application.
The .project file can be stored in the SAP
HANA repository.
.project SAP HANA Studio Projects
[page 44]
Package A container in the repository for
development objects.
Packages are
represented by
folders.
Maintaining Repository
Packages [page 47]
Modeling Data
Schema A database schema for organizing database
objects.
.hdbschema Schema [page 84]
Table A database table. .hdbtable Tables [page 86]
SQL View A virtual table based on a SQL query. .hdbview SQL Views [page 97]
Attribute,
Analytic and
Calculation
View
A view created with modeling tools and
designed to model a business use case.
Created with the
Navigator view.
Setting Up the Analytic
Model [page 109]
Decision Table A database table used to define business
rules, for example, for validating data.
Creating Decision Tables
[page 154]
22
PUBLI C

Analytic
Privilege
A set of rules that allows users to seeing a
subset of data in a table or view.
Creating Analytic Privileges
[page 433]
Sequence A set of unique numbers, for example, for
use as primary keys for a specific table.
.hdbsequence Sequences [page 94]
Procedure A database function for performing
complex and data-intensive business logic
that cannot be performed with standard
SQL.
.procedure Developing Procedures
[page 183]
Web Access
Application
Descriptor
A file in a repository package that defines a
root folder of a native SAP HANA
application. All files in that package are
available to be called via URL.
.xsapp Creating the Application
Descriptors [page 54]
Application
Access File
A file that defines permissions for a native
SAP HANA application, that is, permissions
for accessing and running objects in the
package.
.xsaccess Enabling Access to SAP
HANA XS Application
Packages [page 66]
Application
Privilege
A file that defines a privilege related to an
SAP HANA Extended Application Services
application, for example, the right to start or
administer the application.
.xsprivileges The Application-Privileges
File [page 68]
Server-Side
JavaScript
Code
JavaScript code that can run in SAP HANA
Extended Application Services and that can
be accessed via URL.
.xsjs Writing Server-Side
JavaScript Application
Code [page 255]
Server-Side
JavaScript
Library
JavaScript code that can run in SAP HANA
Extended Application Services but cannot
be accessed via URL. The code can be
imported into an .xsjs code file.
.xsjslib
OData
Descriptor
A file that defines an OData service that
exposes SAP HANA data.
.xsodata Data Access with OData in
SAP HANA XS [page 191]
XMLA
Descriptor
A file that defines an XMLA service that
exposes SAP HANA data.
.xsxmla Data Access with XMLA in
SAP HANA XS [page 230]
SQL Connection
Configuration
A file that enables execution of SQL
statements from inside server-side
JavaScript code with credentials that are
different than those of the requesting user.
.xssqlcc The SQL Connection
Configuration File [page
261]
Other
Role A file that defines an SAP HANA role. .hdbrole Creating Roles in the
Repository [page 414]
PUBLI C

Search Rule Set A file that defines a set of rules for use with
fuzzy searches. The rules help decide what
is a valid match in a search.
.searchruleset Search Rules [page 368]
Resource
Bundle
A file for defining translatable UI texts for an
application.
.hdbtextbundle Creating an SAPUI5
Application [page 272]
Application Site A file that defines an application site. .xsappsite Creating an Application Site
[page 277]
Widget A file that defines a standalone SAP HANA
application for the purpose of integration
into an application site. A file that defines a
standalone SAP HANA application for the
purpose of integration into an application
site.
.xswidget Creating a Widget [page
280]
4.3 Repository
The SAP HANA repository is the design-time storage system for development objects and is built into SAP HANA.
The repository is the source control for all your development work on SAP HANA. You can add objects to the
repository, update the objects, publish the objects, and compile these design-time objects into runtime objects.
The repository supports the following:
Version Control
Sharing of objects between multiple developers
Transport
The repository manages all development objects. You can browse the repository for a specific system with the
SAP HANA Repositories view.
Packages
Within the repository, development objects are managed as files within packages. Packages enable you to:
Create namespaces, so you can uniquely identify development objects.
Group objects that logically belong together.
Provide containers for your objects so you can easily insert relevant objects into delivery units and transport
them.
Assign permissions at the package level.
You might create a top-level package for your company, and then a subpackage for each project, and then
subpackages for parts of your project, for example, a subpackage for your HTML files, another for server-side
JavaScript files, and so forth.
24
PUBLI C

Workspaces
When working with development objects, you need to retrieve them from the repository, work on them on your
workstation, and then return them to the repository. To make this simpler, the SAP HANA studio enables you to
create a repository workspace, which establishes a link between two locations:
The repository of the SAP HANA system where you wish to maintain the development object files of your
projects (in the form of a package hierarchy).
A directory/folder on your workstation where you checkout these project files while you are working on them.
Working with Objects
When creating development objects, the following is the workflow:
1. Create a project.
2. Share the project to associate it with a workspace, and thus a specific SAP HANA system. Shared projects,
once they are activated, are available for import by other members of the application-development team.
3. Create a new object file within the project, and save it to your workstation. Depending on the extension of the
file you create, you may use a dedicated editor for that type of object, for example, a JavaScript editor
for .xsjs files.
4. Commit the file.
Committing the file saves it in the repository, but the object is inactive and cannot be run, and only you can
see it.
5. Activate the file.
Activating a file does the following:
Publishes the file so others can see it.
Adds the previously active version of the file to the history. The repository maintains a history of changes
to the file.
Validates the file and compiles it or exposes it as a runtime object.
The design-time object is now active.
Related Links
Setting Up Your Application [page 37]
In SAP HANA Extended Application Services (SAP HANA XS), the design-time artifacts that make up your
application are stored in the repository like files in a file system. You first choose a root folder for your application-
development activities, and within this folder you create additional subfolders to organize the applications and the
application content according to your own requirements.
PUBLI C

4.4 SAP HANA Studio
The SAP HANA studio is an Eclipse-based development and administration tool for working with SAP HANA,
including creating projects, creating development objects, and deploying them to SAP HANA. As a developer, you
may want to also perform some administrative tasks, such as configuring and monitoring the system.
There are several key Eclipse perspectives that you will use while developing:
Modeler: Used for creating various types of views and analytical privileges.
SAP HANA Development: Used for programming applications, that is, creating development objects that
access or update the data models, such as server-side JavaScript or HTML files.
Debug: Used to debug code, such as server-side JavaScript or SQLScript.
Administration: Used to monitor the system and change settings.
To open a perspective, go to Window Open Perspective , or select on the toolbar.
4.4.1 Modeler Perspective
The Modeler perspective is used to define your analytic model by creating various types of views.
26
PUBLI C

The perspective contains the following main areas:
Navigator view: A view of the database objects, that is, those objects you create from the Modeler perspective.
Quick Launch Area: A collection of shortcuts for performing the most common modeling tasks. If you close
the Quick Launch tab, you can reopen it by selecting Help Quick Launch .
4.4.1.1 Navigator View
The Navigator view displays the database objects in SAP HANA. You can browse both the activated (runtime)
objects and those objects you have created but not yet activated.
The Navigator view is divided into the following main sections:
Security
Contains the objects you use to set up system authorizations and privileges, for example: roles and users
defined for this system.
Catalog
Contains the database objects that have been activated. The objects are grouped into schemas, which is the
common way to organize activated database objects, for example: tables, views, column views, and
procedures.
Content
Contains design-time database objects, both those that have been activated and those not activated. This
includes all the artifacts collected in your application packages. If you want to see other development objects,
use the SAP HANA Repositories view.
PUBLI C

Related Links
SAP HANA Repositories View [page 29]
4.4.2 SAP HANA Development Perspective
The SAP HANA Development perspective is where you will do most of your programming work, creating projects,
associating them to SAP HANA systems, creating development objects, and deploying them.
28
PUBLI C

The perspective contains the following main areas:
Development Objects: Several views for browsing your development objects: the objects on your workstation,
and the objects in the repository of the SAP HANA system you are working with.
Editors: Specialized editors for working with different types of development objects.
4.4.2.1 SAP HANA Repositories View
The SAP HANA Repositories view enables you to browse the repository of a specific SAP HANA system, viewing
the package hierarchy and to download files to your workstation.
PUBLI C

The view essentially is a list of repository workspaces that you have created for developing on various systems.
Generally, you create a workspace, check out files from the repository, and then do most of your development
work in the Project Explorer.
If you are creating a project from scratch, you do not need anything from this view, except to see the hierarchy of
the repository. You create the project in the Project Explorer.
4.4.2.2 Project Explorer View
The Project Explorer is the standard Project Explorer view in Eclipse, which shows you the development files
located on your workstation.
Here you can create files, edit files, and deploy them to the repository.
4.4.2.3 Working with the Repository
To work with the repository, you need to either:
Share your project with the repository via a workspace. Sharing a project associates it with a SAP HANA
system, so that files within the project can be added to the repository on that system.
Check out an existing package in the repository from a workspace. This creates copies on your workstation of
the package, its subpackages and their objects.
Team Menu
You can interact with the repository by right-clicking on a file or project (in the Project Explorer view) or package
(in the SAP HANA Repositories view), and selecting an option from the Team menu.
30
PUBLI C

You can perform the following on a file, folder or project in the Project Explorer view, that is, development objects
that you are working on and located on your workstation:
Table 5: Repository Actions in Project Explorer (project, folder, or file)
Action Description
Commit
Saves the object to the repository.
The object (or changes to the object) are only visible to you. There is no versioning of
committed objects; the repository stores only the latest committed changes.
Activate Makes the development object active, and does the following:
Publishes the object so it is visible to others.
Compiles the object into runtime objects.
Deploys the runtime object.
Creates a new version of the design-time object.
PUBLI C

Action Description
Check Simulates activation of the object in order to check if the object is valid and can be
activated.
Checkout Retrieves the latest version of the object from the repository.
Revert Removes all your committed changes (inactive object) with the latest version of the object
that was activated.
Share/Unshare
(project only)
Associates a project with a specific SAP HANA system, so that files within the project can
be added and updated on the repository on that system. For more information, see Using
SAP HANA Projects [page 42].
The following options are available for files only.
Show Local History Shows a history of all versions of your object saved on your workstation. This is the Eclipse
local history feature.
Show History
Shows a history of all the versions of the object that were activated in the repository.
Currently, you can only view a list of versions.
You can perform the following on a package and file in the SAP HANA Repositories view:
Table 6: Repository Actions in SAP HANA Repositories (package or file)
Action Description
Checkout (package
only)
Retrieves the latest version of the objects from the repository in this package and its
subpackages.
Checkout and
Import Projects
(package only)
Retrieves the latest version of the objects from the repository in this package, and imports
a project into the SAP HANA studio. For more information, see Using SAP HANA Projects
[page 42].
Show History (file
only)
Shows a history of all the versions of the object that were activated in the repository.
32
PUBLI C

Status of Development Objects
Each object displayed in your project within the Project Explorer view is shown with an icon that indicates its
status.
Table 7: Status of Development Objects
Icon Description
The file is not committed. You made a change to the file and saved it to your workstation, but have
not yet committed the changes to the repository.
The file is committed.
The file is activated.
4.4.3 Debug Perspective
The Debug perspective in the SAP HANA studio is the standard Eclipse Debug perspective, enabling you to start
debug sessions, create breakpoints and watch variables.
In the SAP HANA studio, the debug tools have been enhanced to let you also debug SAP HANA code, both server-
side JavaScript (.xsjs files) and SQLScript code (.procedure files). The following shows the tools available in
the Debug perspective and how it is organized:
PUBLI C

Debug Session: The debug sessions that have been started
Watch: Breakpoints and variables to watch
Code: Code files that you have opened
Related Links
Debugging Server-Side JavaScript [page 266]
SAP HANA studio enables you to debug XS JavaScript files, including setting breakpoints and inspecting
variables.
Debugging SQLScript [page 187]
The SAP HANA SQLScript debugger allows you to debug and analyze procedures. In a debug session, your
procedures will be executed in a serial mode, not in parallel (not-optimized). This allows you to test the
correctness of the procedure logic and is not for evaluating the performance.
4.4.4 Administration Console Perspective
The Administration Console perspective enables you to configure the SAP HANA server so that you can do your
development work. For example, when debugging, the debug port must be opened and debugging enabled, which
is done by setting configuration parameters within the administration console.
The Administration Console perspective displays the Navigator view, for adding or selecting systems, and the
administration console, where you can monitor and configure the system. Configuration parameters can viewed
and changed from the Configuration tab.
Note
You may need additional permissions to work with the administration console.
34
PUBLI C

If the console is closed, you can open it by clicking the Administration icon in the Navigator view.
4.5 Getting Started
Before starting development work in the SAP HANA studio, make sure to have the following:
An installation of the SAP HANA studio on your workstation.
A live SAP HANA system to which to connect.
A user on the SAP HANA server that has at least the following roles or their equivalent:
PUBLI C

MODELING
CONTENT_ADMIN
4.5.1 Adding a System
To develop applications, you must first make a connection from your SAP HANA studio to an SAP HANA system.
1. In the Navigator view, right-click anywhere in the view and select Add System.
2. In the System window, enter the host name, instance number, and a description for the SAP HANA system
you want to add.
3. Select Next.
4. Enter a user name and password, and select Finish.
The Navigator view includes a new top-level node for the system. You can now create a repository workspace for
this system so you can start to develop objects to run in it.
Related Links
36
PUBLI C

5 Setting Up Your Application
Note
Application development with SAP HANA Extended Application Services (SAP HANA XS) is currently only
available as an SAP-led project solution, for pre-approved customers and partners. This applies to server-side
JavaScript programming, support for ODATA and XMLA, Web server features and the Web application
development environment. For more information, seeSAP Note 1779803.
As part of the application-development process, you typically need to perform the tasks described in the following
list. Each of the tasks in more detail is described in its own section:
1. Check application-development prerequisites
Before you start developing applications using the features and tools provided by the SAP HANA XS,
developers who want to build applications to run on SAP HANA XS need to be granted access to development
tools, SAP HANA systems, database accounts, and so on.
2. Set up delivery units.
To create and manage delivery units, you must set the identity of the vendor with whom the delivery units are
associated. To avoid conflicts with applications from SAP or other providers, we recommend that you name
the root application-development folder for your company using the DNS name of your company. For
example, you could use the name acme.com.hr.newHires for the root folder for a new application
managing new hires in a company called acme.
3. Set up SAP HANA projects.
In SAP HANA, projects enable you to group together all the artifacts you need for a specific part of the
application-development environment. To start the application-development work flow, you first create a
repository workspace in the SAP HANA Development perspective, which creates a directory structure to store
files on your PC; the workspace you create enables you to synchronize changes in local files with changes in
the repository. Then you can use the SAP HANA studio to create a project to manage the development
activities for the new application.
4. Maintain repository packages.
To perform the high-level tasks that typically occur during the process of maintaining repository packages,
you need to be familiar with the concepts of packages and package hierarchies, which you use to manage the
artifacts in your applications.
5. Maintain application descriptors.
The framework defined by the application descriptors includes the root point in the package hierarchy where
content is to be served to client requests. The framework also defines if the application is permitted to expose
data to client requests, what kind of access to the data is allowed, and what if any privileges are required to
perform actions on packages and package content.
5.1 Before you Start
To enable application-developers to start building native applications that take advantage of the SAP HANA
Extended Application Services (SAP HANA XS), the SAP HANA administrator must set up the application-
Setting Up Your Application
PUBLI C

development environment in such a way that developers have access to the tools and objects that they need to
perform the tasks required during the application-development process.
Before you start developing applications using the features and tools provided by the SAP HANA XS, bear in mind
the following prerequisites. Developers who want to build applications to run on SAP HANA XS need the following
tools, accounts, and privileges:
Note
The following tasks can only be performed by someone who has the required authorizations in SAP HANA, for
example, a SAP HANA administrator.
Access to a running SAP HANA development system (with SAP HANA XS)
A valid user account in the SAP HANA database on that system
Access to SAP HANA studio tools
Access to the SAP HANA client (which SAP HANA studio uses to connect to the repository)
Access to the SAP HANA repository
Access to selected run-time catalog objects
Note
To provide access to the repository for application developers, you can use a predefined role or create your
own custom role to which you assign the privileges that the application developers need to perform the
everyday tasks associated with the application-development process.
To provide access to the repository from the SAP HANA studio, the EXECUTE privilege is required for
SYS.REPOSITORY_REST, the database procedure through with the REST API is tunneled. To enable the activation
and data preview of information views, the technical user _SYS_REPO also requires SELECT privilege on all
schemas where source tables reside.
In SAP HANA, you can use roles to assign one or more privileges to a user according to the area in which the user
works; the role defines the privileges the user is granted. For example, a role enables you to assign SQL privileges,
analytic privileges, system privileges, package privileges, and so on. To create and maintain artifacts in the SAP
HANA repository, you can assign application-development users the following roles:
One of the following:
MODELING
The predefined MODELING role assigns wide-ranging SQL privileges, for example, on _SYS_BI and
_SYS_BIC. It also assigns the analytic privilege _SYS_BI_CP_ALL, and some system privileges. If these
permissions are more than your development team requires, you can create your own role with a set of
privileges designed to meet the needs of the application-development team.
Custom DEVELOPMENT role
A user with the appropriate authorization can create a custom DEVELOPMENT role specially for
application developers. The new role would specify only those privileges an application-developer needs
to perform the everyday tasks associated with application development, for example: maintaining
packages in the repository, executing SQL statements, displaying data previews for views, and so on.
PUBLIC
This is a role that is assigned to all users by default.
Related Links
Creating Roles in the Repository [page 414]
You model roles in the SAP HANA repository in a domain-specific language (DSL).
38
PUBLI C

Defining Repository Package Privileges [page 51]
In the SAP HANA repository, you can set package authorizations for a specific user or for a role. Authorizations
that are assigned to a repository package are implicitly assigned to all sub-packages, too. You can also specify if
the assigned user authorizations can be passed on to other users.
5.2 Setting up Delivery Units
A delivery unit is a collection of packages that are to be transported together. You assign all the packages
belonging to your application to the same delivery unit to ensure that they are transported consistently together
within your system landscape. Each delivery unit has a unique identity.
The identity of a delivery unit consists of two parts: a vendor name and a delivery-unit name. The combined ID
ensures that delivery units from different vendors are easy to distinguish and follows a pattern that SAP uses for
all kinds of software components.
To create and manage delivery units you first need to maintain the identity of the vendor, with whom the delivery
units are associated, and in whose namespace the packages that make up the delivery unit are stored. As part of
the vendor ID maintenance process, you must perform the following tasks:
1. Understand delivery units
You must be familiar with the conventions that exist for delivery-unit names and understand the phases of the
delivery-unit lifecycle.
2. Maintain details of the vendor ID associated with a delivery unit.
Delivery units are located in the namespace associated with the vendor who creates them and who manages
the delivery-unit's lifecycle.
3. Create a delivery unit.
Related Links
Maintaining the Delivery-Unit Vendor ID [page 39]
In SAP HANA, the vendor ID is used primarily to define the identity of the company developing a software
component that they plan to ship for use with SAP HANA. If you want to create a delivery unit, it is a prerequisite
to maintain a vendor ID in your system.
Creating a Delivery Unit [page 40]
A delivery unit is a group of transportable objects used for content delivery. You can use a delivery unit to
transport the design-time objects that are stored in the SAP HANA repository between two systems, for example,
from a development system to a consolidation system.
5.2.1 Maintaining the Delivery-Unit Vendor ID
Before creating your first own delivery unit you must set the identity of the vendor in the development system's
configuration. To maintain details of the delivery-unit vendor ID, perform the following steps:
PUBLI C

1. Start the SAP HANA studio.
2. Switch to the Administration Console perspective.
In the SAP HANA studio's Navigator view, select the desired system, and choose Administration from the
context-sensitive menu. Alternatively, use the menu path: Window > Open Perspective > Administration
Console .
3. Display configuration details for the SAP HANA instance; choose the Configuration tab page.
4. Maintain details of the vendor ID.
In the Configuration tab page, perform the following steps:
a) Locate indexserver.ini in the list of configuration files displayed in the Name column.
b) Expand the indexserver.ini entry.
c) Expand the repository entry.
d) Edit the content_vendor parameter.
e) Double-click content_vendor and enter the name of your vendor. Note that guidelines and conventions
exist for vendor names.
Note
We recommend that you use your DNS name to set the vendor ID, for example, acme.com.
f) Save your changes.
5.2.2 SAP HANA Delivery Unit Naming Conventions
In SAP HANA, conventions and guidelines exist for the naming of delivery units (DU). The delivery unit is the
vehicle that lifecycle management (LCM) uses to ship one or more software components from SAP (or a partner)
to a customer. The DU is also the container you use to transport application content in your system landscape.
If you are creating a delivery unit, you must adhere to the following naming conventions
The name of a delivery unit must contain only capital letters (A-Z), digits (0-9), and underscores (_)
You cannot use an underscore (_) as the first character of a delivery-unit name.
Note
The naming conventions for packages in a delivery unit differ from the naming conventions that apply to the
delivery unit itself. For example, the maximum length of a package name is not restricted to 30 characters; it
must be less than 190 characters (including the namespace hierarchy).
5.2.3 Creating a Delivery Unit
A delivery unit is a group of transportable objects used for content delivery. You can use a delivery unit to
transport the design-time objects that are stored in the SAP HANA repository between two systems, for example,
from a development system to a consolidation system.
In the SAP HANA studio, you can create a delivery unit in the Quick Launch view of the Modeler perspective :
40
PUBLI C

Note
You cannot create a delivery unit unless you have already defined the delivery unit's vendor ID; the vendor ID
defines the namespace in which the new delivery unit resides.
To create a new delivery unit, perform the following steps:
1. In the SAP HANA studio, start the Modeler perspective.
2. In the Setup screen area of the Quick Launch tab, choose Delivery Units....
3. Create a new delivery unit.
In the Delivery Units dialog, choose Create...
4. Maintain delivery unit details.
a) Enter a name for the new delivery unit.
The delivery unit Name is mandatory.
b) Fill in the other information as required:
Note the following points when entering information:
The name of the Vendor is set to the vendor ID that is specified in the system configuration, for
example, acme.com.
The Responsible text box enables you to specify the name of the person responsible for managing the
delivery unit.
In the text boxes Version, Support Package Version, and Patch Version enter integer values only; the
combined values define the version of the delivery unit that is currently being developed. For
example, enter Version = 1, Support Package Version = 3, and Patch Version = 17 to specify that the
current version of your delivery unit is 1.3.17. The version number is transported to other systems
with every DU transport.
Note
The numbers you enter here refer to the application component that you are developing; the
numbers do not refer to the patch or service-pack level deployed on the SAP HANA server.
The PPMS ID is the product ID used by the SAP Product and Production Management System
(PPMS).
Note
Customers and partners should leave the PPMS ID text box empty.
Related Links
Maintaining the Delivery-Unit Vendor ID [page 39]
SAP HANA Delivery Unit Naming Conventions [page 40]
PUBLI C

5.3 Using SAP HANA Projects
Projects group together all the artifacts you need for a specific part of the application-development environment.
Before you can start the application-development workflow, you must create a project, which you use to group
together all your application-related artifacts. However, a project requires a repository workspace, which enables
you to synchronize changes in local files with changes in the repository. You can create the workspace before or
during the project-creation step. As part of the project-creation process, you perform the following tasks:
1. Create a development workspace.
The workspace is the link between the SAP HANA repository and your local filesystem, where you work on
project-related objects.
2. Create a project.
Create a new project for a particular application or package; you can use the project to collect in a convenient
place all your application-related artifacts.
3. Share a project.
Sharing a project enables you to ensure that changes you make to project-related files are visible to other
team members and applications. Shared projects are available for import by other members of the
application-development team.
Note
Files checked out of the repository are not locked; conflicts resulting from concurrent changes to the same
file must be resolved manually, using the Merge tools provided in the context-sensitive Team menu.
4. Import a project.
Import a project (and its associated artifacts) that has been shared by another member of the application-
development team.
Related Links
Creating a Repository Workspace [page 43]
A workspace is a local directory that you map to all (or part) of a package hierarchy in the SAP HANA repository.
When you check out a package from the repository, SAP HANA copies the contents of the package hierarchy to
your workspace, where you can work on the files.
Creating a Project for SAP HANA XS [page 45]
Before you can start the application-development workflow, you must create a project, which you use to group all
your application-related artifacts.
Sharing a Project for SAP HANA XS [page 45]
Before you can start working on files associated with a new project, you must share the project; sharing a project
enables you to track and synchronize local changes with the repository.
Importing a Project in SAP HANA XS [page 47]
Before you can start the application-development workflow, you must either create a new project and share it
(with the repository), or import a shared project from the repository into your workspace. Importing a project
enables you to track and synchronize local changes with the colleagues working on the objects in the imported
project.
42
PUBLI C

5.3.1 SAP HANA Repository: Workspaces
The place where you work on project-related objects is called a repository workspace. A workspace is an
environment that maps a local directory to all (or part) of a package hierarchy in the SAP HANA repository.
In SAP HANA studio, the repository tools enable you to view and browse the entire hierarchy of design-time
objects stored in the repository. However, when you checkout a package from the repository, SAP HANA copies
the contents of the package hierarchy to your workspace, where you can work on the files in your local file system.
Note
Before you can create a workspace you must maintain connection information in the SAP HANA database user
store.
To start development work with SAP HANA studio, for example, to checkout the contents of a package, you must
create a repository workspace. The workspace contains a system folder with metadata and package folders for
the repository content. The file-system folders and their subfolders reflect the package hierarchy in the
repository; the repository client ensures that changes are synchronized.
5.3.2 Creating a Repository Workspace
A workspace is a local directory that you map to all (or part) of a package hierarchy in the SAP HANA repository.
When you check out a package from the repository, SAP HANA copies the contents of the package hierarchy to
your workspace, where you can work on the files.
Before you can start work on the development of the application, you need to set up a workspace, where you store
checked-out copies of your applications source-code files.
To create a new workspace in the SAP HANA studio, perform the following steps:
1. Open the SAP HANA studio.
2. Open the SAP HANA Development perspective.
3. Choose the SAP HANA Repositories view.
4. Choose Create Workspace
The Create Workspace button is located in the top right-hand corner of the SAP HANA Repositories view.
5. Specify the workspace details. In the Create New Repository Workspace dialog, enter the following
information and choose Finish:
a) Specify the SAP HANA system, for which you want to create a new workspace.
b) Enter a workspace name, for example the name of the SAP HANA system where the repository is located.
To avoid the potential for confusion, it is recommended to associate one workspace with one repository.
c) Specify where the workspace root directory should be located on your local file system, for example: C:
\users\username\workspaces
The new workspace is displayed in the SAP HANA Repositories view.
Note
PUBLI C

Although the packages and objects in the chosen repository are visible in the SAP HANA Repositories view,
you cannot open or work on the objects here. To work on objects, you must create a project and use the
Project Explorer view.
5.3.3 SAP HANA Studio Projects
A basic project contains folders and files. More advanced projects are used for builds, version management,
sharing, and the organization and maintenance of resources.
Projects enable multiple people to work on the same files at the same time. You can use SAP HANA studio to
perform the following project-related actions in the repository:
Checkout folders and files from the repository
Commit changes to the repository
Activate the committed changes
Revert inactive changes to the previously saved version
By committing project-related files to the repository and activating them, you enable team members to see the
latest changes. The commit operation detects all changes in packages that you configure SAP HANA studio tool
to track and writes the detected changes back to the repository. The repository client tools also support
synchronization with changes on the server, including conflict detection and merging of change. All workspace-
related repository actions are available as context-sensitive menu options in SAP HANA studio. For example, if
you right click a repository object at the top of the package hierarchy in the Project Explorer in SAP HANA studio,
you can commit and activate all changed objects within the selected hierarchy.
Note
If you create a new project using SAP HANA studio, you can assign the new project to an existing workspace.
You can share and unshare projects. Sharing a project associates it with a particular package in the repository
linked to a particular workspace. The act of sharing the project sets up a link between the workspace and the
repository and enables you to track and synchronize local changes with the versions of the objects stored in the
repository. When a project is shared, it becomes available to other people with authorization to access to the
repository, for example, colleagues in an application-development team. Team members can import a shared
project and see and work on the same files as the creator of the project.
Note
Always unshare a project before deleting it.
In the SAP HANA studio you can create a project at any package level, which enables a fine level of control of the
artifacts that may (or may not) be exposed by sharing the project.
44
PUBLI C

5.3.4 Creating a Project for SAP HANA XS
Projects group together all the artifacts you need for a specific part of your application-development environment.
A basic project contains folders and files. More advanced projects are used for builds, version management,
sharing, and the organization and maintenance of resources.
To create a new project in the SAP HANA studio, perform the following steps:
3. Choose the Project Explorer view.
4. Right-click the white space in the Project Explorer view and choose New > Project in the popup menu.
The type of project you create determines the details you have to provide in the New Project dialog that
appears.
a) Enter a project name that describes what the project is about, for example: XS_JavaScript, XS_OData
or XS_SAPUI5.
b) Click Finish to create the new project.
The new project is displayed in the Project Explorer view.
Note
The contents of the project depend on the type of project you create. For example, a general project is
empty immediately after creation; a JavaScript project contains all the resource files associated with a
JavaScript project, such as libraries and build-environment artifacts.
5.3.5 Sharing a Project for SAP HANA XS
Before you can start working on files associated with a new project, you must share the project; sharing a project
enables you to track and synchronize local changes with the repository.
When you share a project, you set up a connection to the SAP HANA repository associated with a particular SAP
HANA instance. Sharing the project enables you to ensure that changes you make to project-related files are
visible to other team members and applications. Other developers can import a shared project and work on the
same files.
Note
Use the Project Explorer view in the SAP HANA studio to check if a project is shared. In addition to the project
name, a shared project displays the SAP HANA system ID of the repository where the shared artifacts are
located, a SAP HANA user name, and the path to the repository package to which the shared project is
assigned, for example. "XSJS_myproject [SID (dbusername, 'sap.hana.xs.app1')].
To share a project in the SAP HANA studio, perform the following steps:
1. Open the SAP HANA studio
PUBLI C

3. Open the Project Explorer view.
4. Share the project
Right-click the project you want to share and choose Team Share Project in the pop-up menu.
5. Select the repository type.
The Share Project dialog displays a list of all available repository types; choose SAP HANA Repository and
choose Next.
6. Select the repository workspace where the project should be located.
7. Specify the package that you want to associate the shared project with.
The Share Project dialog displays the suggested location for the shared project in the New Project location
screen area. The default location is the name of the workspace with the name of the project you want to
share. Choose Browse... to locate the package you want to associate the shared project with. The selected
package is displayed in the Path to package text box.
Note
The Keep project folder option appends the name of the project you are sharing to the name of the
workspace in which you are sharing the project and creates a new package with the name of the shared
project under the workspace location displayed. Use this option only if you want to create multiple projects
for a selected package, for example, if you are creating a root project in your root application package.
8. Click Finish to complete the project-sharing procedure.
9. Add new files as required
At this point you can start adding project-specific files to the shared project. These artifacts can then be
committed to the repository, where they reside as inactive objects until they are activated, for example, using
the Team Activate option in the context-sensitive menus available in the Project Explorer view.
Note
The Project Explorer view decorates the file icons to indicate the current state of the repository files, for
example: local (not yet committed), committed (inactive), and active (available for use by others).
10. Make the project available for import, for example, so that others can join it and make changes to project
content.
The project-sharing procedure creates some artifacts (for example, the .project file) that must be
committed to the repository and activated so that other team members can import the project more easily
into their workspace. The .project file is used in several dialogs to populate the list of available projects.
Note
Use the SAP HANA Repositories view to import projects (and checkout project content).
Related Links
Importing a Project in SAP HANA XS [page 47]
project.
46
PUBLI C

5.3.6 Importing a Project in SAP HANA XS
project.
To import an existing project from the repository into your workspace, perform the following steps.
1. Open the SAP HANA studio
3. Choose the HANA Repositories view.
4. Right-click the package where the project you want to import is located and choose Checkout and Import
Projects... in the popup menu.
Projects can be assigned to a package at any level of the package hierarchy. If you know where the project is
located, browse to the package first before choosing the Checkout and Import Projects... option. This reduces
the amount of files to checkout and download to your local file system.
Note
The existence of a .project file in a package identifies the package as being associated with a project.
The SAP HANA studio checks out the content of the selected package and displays any projects it finds in the
Projects screen area.
5. Select the projects to import.
If multiple projects are available for import, select the projects you want to import.
6. Choose Finish to import the selected projects.
You can add the imported project to your Working Sets.
Note
A working set is a concept similar to favorites in a Web browser, which contain the objects you work on
most frequently.
5.4 Maintaining Repository Packages
All content delivered as part of the application you develop for SAP HANA is stored in packages in the SAP HANA
repository. The packages are arranged in a hierarchy that you define to help make the process of maintaining the
packages transparent and logical.
To perform the high-level tasks that typically occur during the process of maintaining repository packages, you
need to be familiar with the concepts of packages and package hierarchy that create and maintain the artifacts for
your applications. You also need to know about the privileges the application developers will need to have access
to (and perform operations on) the packages.
Note
PUBLI C

You can also create and delete packages in the Project Explorer, for example, by creating or deleting folders in
shared projects and committing and activating these changes. However, to maintain advanced package
properties (for example, privileges, component, the package maintainer, and so on) you must use the Modeling
perspective in the SAP HANA studio..
As part of the process of maintaining your application packages, you typically perform the following tasks:
1. Create a package
Packages are necessary to group logically distinct artifacts together in one object location that is easy to
transport.
2. Define the package hierarchy
The package hierarchy is essential for ease of maintenance as well as the configuration of access to packages
and the privileges that are required to perform actions on the packages.
3. Define package privileges
You can set package authorizations for a specific user or for a role. Authorizations that are assigned to a
repository package are implicitly assigned to all sub-packages, too.
Related Links
Creating a package [page 51]
In SAP HANA, a package contains a selection of repository objects. You assemble a collection of packages into a
delivery unit, which you can use to transport the repository objects between SAP HANA systems.
Defining the package hierarchy [page 52]
Packages belonging to an application-development delivery unit (DU) should be organized in a clear hierarchical
structure under a single root package representing the vendor, for example, acme.com.
Defining package privileges [page 51]
5.4.1 SAP HANA Repository Packages and Namespaces
In SAP HANA, a package typically consists of a collection of repository objects, which can be transported between
systems. Multiple packages can be combined in a delivery unit (DU).
A SAP HANA package specifies a namespace in which the repository objects exist. Every repository object is
assigned to a package, and each package must be assigned to a specific delivery unit. In the repository, each
object is uniquely identified by a combination of the following information:
Package name
Object name
Object type
Note
Multiple objects of the same type can have the same object name if they belong to different packages.
Before you start the package development process, consider the following important points:
48
PUBLI C

Package hierarchy
Each vendor uses a dedicated namespace, and the package hierarchy you create enables you to store the
various elements of an application in a logical order that is easy to navigate.
Package type
Packages can be structural or non-structural; some packages contain content; other packages contain only
other (sub)packages.
Package naming conventions
There are recommendations and restrictions regarding package names, for example, the name's maximum
length and what if any characters must not be used.
Package Hierarchy
You can create a package hierarchy, for example, by establishing a parent-child type relationship between
packages. The assignment of packages to delivery units is independent of the package hierarchy; packages in a
parent-child relationship can belong to different delivery units. SAP recommends that you assign to one specific
delivery unit all packages that are part of a particular project or project area.
The package hierarchy for a new project typically includes sub-packages, for example, to isolate the data model
from the business logic. Although there are no package interfaces to enforce visibility of objects across packages,
this separation of logical layers of development is still a recommended best practice.
Note
You can only assign one project per package; this is important to remember if you have a mixture of design-
time objects that need to be used in multiple projects, for example: server-side JavaScript (XSJS), SAPUI5, and
a general project (for procedures).
The following simple example shows a package structure containing tutorials for the use of a new application:
sap
\
hana
\
app1
\
code
demos
docs
\
tutorials
manuals
help
All content delivered by SAP should be in a sub-package of "sap". Partners and customers should choose their
own root package to reflect their own name (for example, the domain name associated with the company) and
must not create packages or objects under the "sap" root structural package. This rules ensures that customer or
partner created content will not be overwritten by an SAP update or patch.
Note
SAP reserves the right to deliver without notification changes in packages and models below the "sap" root
structural package.
PUBLI C

There are no system mechanisms for enforcing the package hierarchy. The "sap" root structural package is not
automatically protected. However, by default you cannot change the content of packages that did not originate in
the system. In addition, an authorization concept exists, which enables you to control who can change what inside
packages.
Package Types
SAP HANA Application Services provide or allow the following package types:
Structural
Package only contains sub-packages; it cannot contain repository objects.
Non-Structural
Package contains both repository objects and subpackages.
The following packages are delivered by default with the repository:
sap
Transportable package reserved for content delivered by SAP. Partners and customers must not use the sap
package; they must create and use their own root package to avoid conflicts with software delivered by SAP,
for example when SAP updates or overwrites the sap package structure during an update or patch process.
system-local
Non-transportable, structural packages (and subpackages). Content in this package (and any subpackages)
is considered system local and cannot be transported. This is similar to the concept of the $tmp development
class in SAP NetWeaver ABAP.
system-local.generated
Non-transportable, structural packages for generated content, that is; content not created by manual user
interaction
system-local.private
Non-transportable, structural package reserved for objects that belong to individual users, for example,
system-local.private.<user_name> . To avoid compatibility issues with future functionality, do not use
the system-local.private package or any of its sub-packages.
Package Naming Conventions
The following rules apply to package names:
Permitted characters
Lower/upper case letters (aA-zZ), digits (0-9), hyphens (-), and dots (.) are permitted in package names.
Dots in a package name define a logical hierarchy. For example, "a.b.c" specifies a package "a" that contains
sub-package "b", which in turn contains sub-package "c".
Forbidden characters
A package name must not start with either a dot (.) or a hyphen (-) and cannot contain two or more
consecutive dots (..).
Package name length
The maximum permitted length of a package name is 190 characters. To avoid problems, we recommend you
restrict the length of package names to well under the 190-character limit.
50
PUBLI C

Package namespace length
The name of the complete package namespace hierarchy (for example, "aa.bb.cc.zz" including dots) must not
be more than 190 characters long. To avoid problems, we recommend you restrict the length of the package
namespace to well under the 190-character limit.
5.4.2 Defining Repository Package Privileges
To set user (or role) authorizations for repository packages, perform the following steps:
1. Open the Navigator view in the SAP HANA studio's Modeler perspective.
2. In the Navigator view, expand the Security Roles/Users node for the system hosting the repository that
contains the packages you want to grant access to.
You can also define roles via source files; roles defined in this way can be assigned to a delivery unit and
transported to other systems.
3. Double click the user (or role) to whom you want to assign authorizations.
4. Open the Package Privileges tab page.
5. Choose [+] to add one or more packages. Press and hold the Ctrl key to select multiple packages.
6. In the Select Repository Package dialog, use all or part of the package name to locate the repository package
that you want to authorize access to.
7. Select one or more repository packages that you want to authorize access to; the selected packages appear
in the Package Privileges tab page.
8. Select the packages to which you want authorize access and, in the Privileges for screen page, check the
required privileges, for example:
a) REPO.READ
Read access to the selected package and design-time objects (both native and imported)
b) REPO.EDIT_NATIVE_OBJECTS
Authorization to modify design-time objects in packages originating in the system the user is working in
c) REPO.ACTIVATE_NATIVE_OBJECTS
Authorization to activate/reactivate design-time objects in packages originating in the system the user is
working in
d) REPO.MAINTAIN_NATIVE_PACKAGES
Authorization to update or delete native packages, or create sub-packages of packages originating in the
system in which the user is working
5.4.3 Creating a Package
In SAP HANA, a package contains a selection of repository objects. You assemble a collection of packages into a
delivery unit, which you can use to transport the repository objects between SAP HANA systems.
PUBLI C

You can use packages to manage the various elements of your application development project. To create a
project, perform the following steps:
1. In the SAP HANA studio, start the SAP HANA Development perspective.
2. In the Navigator view, select the SAP HANA system where you want to create a new package and expand the
Content node to display the namespace hierarchy for package content.
3. Right-click the package where you want to add a new package and choose New Package... in the
context-sensitive popup menu.
SAP HANA studio displays the New Package dialog.
4. Maintain the package details.
In the New Package dialog, enter information in the following fields:
a) Enter a name for the new package.
The package Name is mandatory. Add the new name to the end of the full package path, for example,
acme.com.package1.
b) Fill in the other optional information as required:
Use the Delivery Unit drop-down list to assign the new package to a delivery unit.
Choose Translation if you intend to have the package content localized. You must maintain the translation
details.
5. Create the new package.
In the New Package dialog, click OK to create a new package in the specified location.
6. Activate the new package.
In the Navigator view, right-click the new package and choose Activate from the context-sensitive popup
menu.
Related Links
SAP HANA Delivery-Unit Translation Details [page 445]
The SAP HANA repository includes features for translating package-related metadata texts. If you plan to
translate the contents of the new delivery unit you create, you must maintain translation details.
5.4.4 Defining a Package Hierarchy
Packages belonging to an application-development delivery unit (DU) should be organized in a clear hierarchical
structure under a single root package representing the vendor, for example, acme.com.
The package hierarchy for a new project might include sub-packages, for example, to isolate the data model from
the business logic. Although there are no package interfaces to enforce visibility of objects across packages, this
separation of logical layers of development is still a recommended best practice.
Note
You can only assign one project per package; this is important to remember if you have a mixture of design-
time objects that need to be used in multiple projects, for example: server-side JavaScript (XSJS), SAPUI5, and
a general project (for procedures).
The following simple example shows a package structure containing tutorials for the use of a new application:
acme
\
52
PUBLI C

hana
\
app1
\
docs
\
tutorials
Package hierarchy
Each vendor uses a dedicated namespace, for example, acme.
Note
Do not use the namespace sap to build your application hierarchy. The namespace sap is reserved for use
by SAP; packages created in the sap namespace are overwritten by system updates.
Package type
Some packages contain content; other packages contain only other (sub)packages. Packages can also
contain both objects and (sub)packages.
Package naming conventions
There are recommendations and restrictions regarding package names.
To set up a package hierarchy in the SAP HANA repository, perform the following steps:
1. Create a new root package.
Open the Modeler perspective and perform the following steps:
a) Choose New > Package .
b) Choose Create...
2. Maintain the package details.
In the Create Package dialog, perform the following steps:
a) Enter the name of the package (mandatory).
Guidelines and conventions apply to package names.
b) Enter a package description (optional).
c) Specify the delivery unit that the package is assigned to.
You can add additional packages to a delivery unit at a later point in time, too.
d) Specify a language for the package content.
e) Assign responsibility of the package to a specific user (optional).
By default, the responsible user for a new package is the database user connected to the SAP HANA
repository in the current SAP HANA studio session.
f) Maintain translation details.
If you plan to have the content translated, you need to maintain the translation details; this is covered in
another topic.
3. Create a new subpackage.
In the Navigator view of the Modeler perspective, perform the following steps:
a) Right-click the package to which you want to add a new subpackage.
b) In the pop-up menu, choose New > Package...
4. Maintain the subpackage details.
In the Create Package dialog, perform the following steps:
a) Enter the name of the subpackage (mandatory).
PUBLI C

Guidelines and conventions apply to package names.
b) Enter a description for the new subpackage (optional).
c) Specify the delivery unit that the subpackage is assigned to.
You can add additional packages to a delivery unit at a later point in time, too.
d) Specify a language for the subpackage content.
e) Assign responsibility of the subpackage to a specific user (optional).
By default, the responsible user for a new package is the database user connected to the SAP HANA
repository in the current SAP HANA studio session.
f) Maintain translation details.
If you plan to have the content translated, you need to maintain the translation details; this is covered in
another topic.
Related Links
SAP HANA Delivery Unit Naming Conventions [page 40]
Maintaining Translation Details [page 445]
Translation details provide a brief overview of the type and current status of the text objects to be translated, for
example, by specifying the technical area of the new delivery unit and indicating the current status of translation
texts.
5.5 Creating the Application Descriptors
When you develop and deploy applications in the context of SAP HANA Extended Application Services (SAP
HANA XS), you must define the application descriptors, which describe the framework in which the application
runs. The framework defined by the application descriptors includes the root point in the package hierarchy where
content is to be served to client requests, whether the application is permitted to expose data to client requests,
what kind of access to the data is allowed, and what if any privileges are required to perform actions on packages
and package content.
To perform the high-level tasks that make up the process of defining the application descriptors, you must be
familiar with the concept of the application descriptor, the application-access file, and if required, the application-
privileges file. Maintaining the application descriptors involves the following tasks:
1. Creating the application descriptor file.
The package that contains the application descriptor file becomes the root path of the resources exposed to
client requests by the application you develop.
2. Creating the application-access file.
The application-access file enables you to specify who or what is authorized to access the content exposed by
the application package and what content they are allowed to see.
3. Creating the application-privileges file. (Optional)
The application-privileges file can be used to define the authorization privileges required for access to an
application, for example, to start the application or to perform administrative actions on an application.
Related Links
54
PUBLI C

Create an application descriptor [page 55]
Each application that you want to develop and deploy on SAP HANA Extended Application Services (SAP HANA
XS) must have an application-descriptor file. The application descriptor is the core file that you use to describe an
application's framework within SAP HANA XS.
Create an application-access file [page 66]
The application-access file enables you to specify who or what is authorized to access the content exposed by the
application package and what content they are allowed to see.
Create an application-privileges file [page 69]
The application-privileges (.xssprivileges) file can be used to define the authorization levels required for
access to an application, for example, to start the application or perform administrative actions on an application.
You can then assign the application privileges to the individual users who require them.
5.5.1 The SAP HANA XS Application Descriptor
XS) must have an application descriptor file. The application descriptor is the core file that you use to describe an
The package that contains the application descriptor file becomes the root path of the resources exposed to client
requests by the application you develop.
Note
The application-descriptor file has no contents and no name; it only has the file extension xsapp, for
example, .xsapp. For backward compatibility, content is allowed in the .xsapp file but ignored.
The application descriptor file performs the following operations:
Determines the called application
The application root is determined by the package containing the .xsapp file. If the package sap.test
contains the file .xsapp, the application will be available under the URL http://<host>:<port>/
sap.test/.
5.5.2 Create an Application Descriptor File
The package that contains the application-descriptor file becomes the root path of the resources exposed by the
application you develop.
1. Create a root package for your application, for example, MyPackage.
Note
The namespace sap is restricted. Place the new package in your own namespace, which you can create
alongside the sap namespace.
PUBLI C

a) Start the SAP HANA studio and open the SAP HANA Development perspective.
b) In the Project Explorer view, right-click the folder where you want to create the new (MyPackage)
package.
c) In the context-sensitive popup menu, choose New Folder .
d) Enter the name MyPackage and choose Finish.
2. Activate the new package in the repository.
a) In the SAP HANA Development perspective, open the Project Explorer view and right-click the new
(MyPackage) package.
b) In the context-sensitive popup menu, choose Team Activate .
3. Create an application descriptor for your new application and place it in the package (MyPackage) you
created in the previous step.
The application descriptor is the core file that you use to indicate an application's availability within SAP
HANA XS. The application descriptor marks the point in the package hierarchy at which an application's
content is available to clients. The application-descriptor file has no contents and no name; it only has the file
extension .xsapp.
Note
For backward compatibility, content is allowed in the.xsapp file but ignored.
a) In the SAP HANA studio and open the SAP HANA Development perspective.
b) In the Project Explorer view, right-click the folder where you want to create the new (.xsapp) file.
c) In the context-sensitive popup menu, choose New File .
d) Enter the name .xsapp and choose Finish.
Files with names that begin with the period (.), for example, .xsapp, are sometimes not visible in the
Project Explorer. To enable the display of all files in the Project Explorer view, use the Customize View
Available Customization option and clear all check boxes.
4. Save and activate your changes and additions.
(.xsapp) package.
5.5.3 The Application-Access File
SAP HANA XS enables you to define access to each individual application package that you want to develop and
deploy.
The application-access file enables you to specify who or what is authorized to access the content exposed by a
SAP HANA XS application package and what content they are allowed to see. For example, you use the
application-access file to specify if authentication is to be used to check access to package content and if rewrite
rules are in place that hide or expose target and source URLs.
The application-access file does not have a name; it only has the file extension .xsaccess. The content of
the .xsaccess file is formatted according to JSON rules and is associated with the package it belongs to as well
as any subpackages lower in the package hierarchy. Multiple .xsaccess files are allowed, but only at different
56
PUBLI C

levels in the package hierarchy. This enables you to specify different application-access rules for individual
packages and subpackages in the package hierarchy.
Note
You cannot place two .xsaccess files in the same package. Furthermore, the rules specified in a .xsaccess
file that is associated with a subpackage take precedence over any rules specified in a .xsaccess file
associated with any parent package higher up the package hierarchy.
The application-access file performs the following operations:
Data exposure
Use the exposed keyword to specify if package content is to be exposed to client requests via HTTP.
Note
Exposed package content can include design-time objects, for example, tables and views.
Authentication rules
Use the authentication keyword to enable authorization for the requests in URLs either at the application level
or for single packages in an application. The following authentication methods are supported:
SAP logon tickets
SAP logon tickets enable single sign-on across SAP boundaries, for example, to set up single sign-on
(SSO) between SAP NetWeaver and SAP HANA XS. SAP logon tickets can also be used as authentication
for access to third-party applications.
To configure the trust relationship between the issuer of the SAP logon ticket and SAP HANA, you must
specify the path to the SAP logon ticket trust store, which contains the trust chain for the ticket issuer.
You can use the SapLogonTicketTrustStore keyword in the xsengine.ini file. Default values are:
$SECUDIR/saplogon.pse or $HOME/.ssl/saplogon.pem.
Note
SAP HANA XS does not issue SAP logon tickets; it only accepts them. Since the tickets usually reside
in a cookie, the issuer and SAP HANA XS need to be in the same domain to make sure that your
browser sends the SAP logon ticket cookie with each call to SAP HANA XS.
Form-based authentication
Redirect the logon request to a form to fill in, for example, on a Web page.
Note
If you need to troubleshoot problems when developing a form-based logon solution for your
application, you can configure the generation of useful trace information in the XSENGINE section of
the database trace component using the following entry: xsa:sap.hana.xs.formlogon.
Basic (user name and password)
Log on with a recognized user name and password
The authentication methods can also be written as an array, for example, to allow applications to support
multiple authentication methods. The order of the authentication methods in the array should proceed from
strongest to weakest, for example:
1. SAP logon ticket
PUBLI C

2. Form-based
3. Basic authentication
Application authorization
Use the authorization keyword in the .xsaccess file to specify which authorization level is required by a user
for access to a particular application package. The authorization levels you can choose from are defined in
the .xsprivileges file, for example, "execute" for basic privileges, or "admin" for administrative privileges
on the specified package.
URL rewrite rules
Use the rewrite_rules keyword in the .xsaccess file to hide internal URL path details from external users,
clients, and search engines. It is not possible to define global rewrite rules; the rules you define apply to the
specified local application only. Rules are specified as a source-target pair where the source is written in the
JavaScript regex syntax, and the target is a simple string where references to the groups found can be
inserted using $groupnumber.
Connection security
Use the force_ssl keyword in the .xsaccess file to enforce the use of secure HTTP (SSL/HTTPS) for client
connections. Browser requests that do not use SSL/HTTPS are refused and the 403 Forbidden page is
displayed. Note that if you set the force_ssl option, then you must ensure that the SAP Web Dispatcher is
configured to accept and manage HTTPS requests. For more information about configuring the SAP Web
Dispatcher to use HTTPS/SSL see the SAP HANA Security Guide.
Note
If the SAP Webdispatcher sends the header x-sap-webdisp-ap with the HTTPS port, the request is
redirected to a HTTPS page, for example: http://..:80/test -> https://..:433/test. Otherwise, a
403 error is displayed.
Entity Tags
You can allow or prevent the generation of entity tags (etags) for static Web content using the enable_etags
keyword in the .xsaccess file. Etags are used to improve caching performance, for example, so that the
same data is not resent if no change has occurred since the last request.
MIME Mapping
MIME means Multipurpose Internet Mail Extensions. You can use the mime_mapping keyword in
the .xsaccess file to define the way in which to map certain file suffixes to required MIME types:
"mime_mapping": [ {"extension":"jpg", "mimetype":"image/jpeg"} ]
Cross-Site Request Forgery (XSRF)
You can use the prevent_xsrf keyword in the .xsaccess file to guard against cross-site request-forgery
attacks. XSRF attacks attempt to trick a user into clicking a specific hyperlink, which shows a (usually well-
known) Web site and perform some actions on the users behalf, for example, in a hidden iframe. The
prevent_xsrf keyword checks the validity of a session-specific security token to ensure that it matches the
token that SAP HANA XS generates in the backend for the corresponding session.
Example
The Application-Access (.xsaccess) File
The following example, shows the composition and structure of the SAP HANA XS application access
(.xsaccess) file. In this file, data is available to client requests. The authentication methods specified are SAP
logon ticket and then, as a fall-back option if the logon with the SAP logon ticket fails, a logon with a user name
58
PUBLI C

and password. Allowing a fall-back log-on mechanism is useful if the requesting client has problems handling
the SAP logon ticket mechanism.
{
"exposed" : true, // Expose data via http
"authentication" : // Authentication method
[
{
"method": "LogonTicket",

},
{
"method" : "Basic"

},
],
"authorization": // Grant package privileges
[
"sap.xse.test::Execute",
"sap.xse.test::Admin"
]
"rewrite_rules" : // URL rewriting rules
[
{
"source": "/entries/(\\d+)/(\\d+)/(\\d+)/",
"target": "/logic/entries.xsjs?year=$1&month=$2&day=$3"
}
],
"mime_mapping" : // Map file-suffix to MIME type
[
{
"extension":"jpg", "mimetype":"image/jpeg"
}
],
"force_ssl" : true, // Refuse request if not SSL
"enable_etags" : false // Prevent etag generation
"prevent_xsrf" : true // Prevent cross-site request forgery
}

Related Links
http://help.sap.com/hana/hana_sec_en.pdf
Application-Access File Keyword Options [page 59]
The application-access (.xsaccess) file enables you to specify whether or not to expose package content, which
authentication method is used to grant access, and what content is visible.
5.5.4 Application-Access File Keyword Options
Example
The Application Access (.xsaccess) File
The following example shows all possible keyword combinations in the SAP HANA XS application-access
(.xsaccess) file.
PUBLI C

Note
In the form shown below, the .xsaccess file is not a working model; it is used to illustrate all possible
options.
{
"exposed" : false,
"authentication" : // null = no auth required,
[
{
},
{
"method": "Form",
},
{
"method" : "Basic"
}
],
"authorization":
[
]
"rewrite_rules" :
[{
"source" : "...",
"target" : "..."
}],
"mime_mapping" :
[
{
}
],
"force_ssl" : true,
"enable_etags" : false
"prevent_xsrf" : false
}

exposed
{
"exposed" : false,
}

The exposed keyword enables you define if content in a package (and its subpackages) is to be made available by
HTTP to client requests. Values are Boolean true or false. If no value is set for exposed, the default setting (false)
applies.
60
PUBLI C

authentication
{
"authentication" :
[
{
},
{
"method": "Form",
},
{
"method" : "Basic",
}
],
}

The authentication keyword enables you to define the authentication method to use for Browser requests either
at the application level or for single packages in an application. SAP HANA Extended Application Services support
the following logon authentication mechanisms:
Note
The default setting is basic. You can specify multiple authentication methods in the application-access file.
Basic authentication
Logon with a recognized database user name and password.
SAP logon ticket
SAP logon tickets enable single sign-on across SAP boundaries, for example, to set up single sign-on (SSO)
between SAP NetWeaver and SAP HANA XS. SAP logon tickets can be also be used as authentication for
access to third-party applications.
To configure the trust relationship between the issuer of the SAP logon ticket and SAP HANA, you must
specify the path to the SAP logon ticket trust store, which contains the trust chain for the ticket issuer. You
can use the SapLogonTicketTrustStore keyword in the xsengine.ini file. Default values are: $SECUDIR/
saplogon.pse or $HOME/.ssl/saplogon.pem.
Note
SAP HANA XS does not issue SAP logon tickets; it only accepts them. Since the tickets usually reside in a
cookie, the issuer and SAP HANA XS need to be in the same domain to make sure that your browser sends
the SAP logon ticket cookie with each call to SAP HANA XS.
Redirect the logon request to a form to fill in, for example, a Web page.
Form-based authentication requires the libxsauthenticator library, which must not only be available but
also be specified in the list of trusted applications in the xsengine application container. The application list is
displayed in the SAP HANA studio's Administration Console perspective in the following location:
Administration Configuration tab xsengine.ini application_container application_list . If it is not
displayed, ask the SAP HANA administrator to add it.
Note
PUBLI C

If you need to troubleshoot problems with form-based logon, you can configure the generation of useful
trace information in the XSENGINE section of the database trace component using the following entry:
xsa:sap.hana.xs.formlogon.
If you use the authentication in the .xsaccess file, you must specify the authentication method to apply, for
example, basic or SAP logon ticket. If you use the authentication keyword in the application-access file, but do not
set an authentication method, your .xsaccess file is a valid JSON file but is not a semantically valid .xsaccess
file and will return a parsing validation error . If you do not set the authentication keyword, your application applies
the default basic authentication method (user name and password) to enable access to data.
Note
You can disable authentication checks with the null options, as follows: {"authentication" : null}.
authorization
{
"authorization":
[
]
}
The authorization keyword in the .xsaccess file enables you to specify which authorization level is required for
access to a particular application package, for example, execute or admin on the package sap.xse.text.
Note
The authorization levels you can choose from are defined in the .xsprivileges file for the package, for
example, "execute" for basic privileges, or "admin" for administrative privileges on the specified package. If you
use the authorization keyword in the .xsaccess file, for example, to require execute privileges for a specific
application package, you must create a .xsprivileges file for the same application package (or a parent
package higher up the hierarchy, in which you define the execute privilege level declared in the .xsaccess
file.
rewrite_rules
{
"rewrite_rules" :
[{
"source" : "...",
"target" : "..."
}]
}

62
PUBLI C

The rewrite_rules keyword enables you hide the details of internal URL paths from external users, clients, and
search engines. Any rules specified affect the local application where the .xsaccess file resides (and any
subpackage, assuming the subpackages do not have their own .xsaccess files); it is not possible to define global
rewrite rules. URL rewrite rules are specified as a source-target pair where the source is written in the JavaScript
regex syntax and the target is a simple string where references to found groups can be inserted using
$groupnumber.
mime_mapping
{
"mime_mapping" :
[
{
}
]
}
The mime_mapping keyword enables you to define how to map certain file suffixes to required MIME types. For
example, you can map files with the .jpg file extension to the MIME type image/jpeg.
force_ssl
{
"force_ssl" : false,
}
The force_ssl keyword enables you to refuse Browser requests that do not use secure HTTP (SSL/HTTPS) for
client connections. If no value is set for exposed, the default setting (false) applies and non-secured connections
(HTTP) are allowed.
enable_etags
{
"enable_etags" : true,
}
You can allow or prevent the generation of entity tags (etags) for static Web content using the enable_etags
keyword. If no value is set, the default setting (true) applies, in which case etags are generated. Etags are used to
improve caching performance, for example, so that the same data is not resent from the server if no change has
occurred since the last time a request for the same data was made.
PUBLI C

prevent_xsrf
{
"prevent_xsrf" : true
}
You can use the prevent_xsrf keyword in the .xsaccess file to protect applications from cross-site request-
forgery (XSRF) attacks. XSRF attacks attempt to trick a user into clicking a specific hyperlink, which shows a
(usually well-known) Web site and performs some actions on the users behalf, for example, in a hidden iframe. If
the targeted end user is logged in and browsing using an administrator account, the XSRF attack can compromise
the entire Web application.
The prevent_xsrf keyword prevents the XSRF attacks by ensuring that checks are performed to establish that a
valid security token is available for a given Browser session. The existence of a valid security token determines if
an application responds to the client's request to display content; if no valid security token is available, a 403
Forbidden message is displayed. A security token is considered to be valid if it matches the token that SAP
HANA XS generates in the back end for the corresponding session.
Note
The default setting is false, which means there is no automatic prevention of XSRF attacks. If no value is
assigned to the prevent_xsrf keyword, the default setting (false) applies.
To include the XSRF token in the HTTP headers, you must first fetch the token as part of a GET request, as
illustrated in the following example:
xmlHttp.setRequestHeader("X-CSRF-Token", "Fetch");
You can use the fetched XSRF token in subsequent POST requests, as illustrated in the following code example:
xmlHttp.setRequestHeader("X-CSRF-Token", xsrf_token);
Related Links
Server-Side JavaScript Security Considerations [page 245]
If you choose to use server-side JavaScript to write your application code, you need to bear in mind the potential
for (and risk of) external attacks such as cross-site scripting and forgery, and insufficient authentication.
5.5.5 Application-Access URL Rewrite Rules
Rewriting URLs enables you to hide internal URL path details from external users, clients, and search engines. You
define URL rewrite rules in the application-access file (.xsaccess) for each application or for an application
hierarchy (an application package and its subpackages).
The rewrite rules you define in the .xsaccess file apply only to the local application to which the .xsaccess file
belongs; it is not possible to define global rules to rewrite URLs. Rules are specified as a source-target pair where
the source is written in the JavaScript regex syntax, and the target is a simple string where references to found
groups can be inserted using $groupnumber.
The following examples show how to use a simple set of rewrite rules to hide internal URLs from requesting clients
and users.
64
PUBLI C

The first example illustrates the package structure that exists in the repository for a given application; the
structure includes the base package apptest, the subpackages subpackage1 and subpackage2, and several
other subpackages:
sap---apptest
|---logic
| |---users.xsjs
| |---posts.xsjs
|---posts
| |---2011...
|---subpackage1
| |---image.jpg
|---subpackage2
| |---subsubpackage
| | |---secret.txt
| |---script.xsjs
|---subpackage3
| |---internal.file
|---users
| |---123...
|---.xsapp
|---.xsaccess
|---index.html
The application-access file for the package apptest (and its subpackages) includes the following rules for
rewriting URLs used in client requests:
{
"rewrite_rules": [
{
"source": "/users/(\\d+)/",
"target": "/logic/users.xsjs?id=$1"
},
{
"source": "/posts/(\\d+)/(\\d+)/(\\d+)/",
"target": "/logic/posts.xsjs?year=$1&month=$2&day=$3"
}
]
}
Assuming we have the package structure and URL rewrite rules illustrated in the previous examples, the following
valid URLs would be exposed; bold URLs require authentication:
/sap/apptest/
/sap/apptest/index.html
/sap/apptest/logic/users.xsjs
/sap/apptest/logic/posts.xsjs
The rewriting of the following URLs would be allowed:
/sap/apptest/users/123/ ==> /sap/appTest/logic/users.xsjs?id=123
/sap/apptest/posts/2011/10/12/ ==> /sap/appTest/logic/posts.xsjs?
year=2011&month=10&day=12
PUBLI C

5.5.6 Enabling Access to SAP HANA XS Application
Packages
You can use a set of keywords in the application-access file .xsaccess to specify if authentication is required to
enable access to package content, which data is exposed, and if rewrite rules are in place to hide target and
source URLs, for example, from users and search engines. You can also specify what, if any, level of authorization
is required for the package and whether SSL is mandatory for client connections.
1. If it does not already exist, create a root package for the application you want to enable access to, for
example, MyPackage.
Note
a) Start the SAP HANA studio and open the SAP HANA Development perspective.
package.
2. Activate the new package in the repository.
(MyPackage) package.
3. If it does not already exist, create an application descriptor for the application and place it in the package
(MyPackage) you created in the previous step.
content is available to clients.
Note
The application-descriptor file has no contents and no name; it only has the file extension .xsapp.
b) In the Project Explorer view, right-click the folder where you want to create the new (.xsapp) file.
d) Enter the name .xsapp and choose Finish.
Files with names that begin with the period (.), for example, .xsapp, are sometimes not visible in the
Project Explorer. To enable the display of all files in the Project Explorer view, use the Customize View
Available Customization option and clear all check boxes.
e) Activate the new .xsapp file in the repository.
66
PUBLI C

4. Create the application access file.
The application-access file is a JSON-compliant file with the file suffix .xsaccess. Note that the application-
access file does not have a name before the dot (.); it only has the suffix .xsaccess.
Create a file called .xsaccess and place it in the root package of the application to which you want to enable
access. A basic .xsaccess file must, at the very least, contain a set of curly brackets, for example, {}. Note
that the .xsaccess file uses keyword-value pairs to set access rules; if a mandatory keyword-value pair is not
set, then the default value is assumed.
b) In the Project Explorer view, right-click the folder where you want to create the new (.xsaccess) file.
d) Enter the name .xsaccess and choose Finish.
5. Enable application access to data.
You use the expose keyword to enable or disable access to content at a package or subpackage level.
{
"exposed" : true
}
6. Define the application authentication method.
You use the authentication keyword to define how to manage the authentication process for requests to
access package content, for example, SAP logon ticket, form-based logon, or a basic user name and
password .
{
"authentication" :
[
{ "method" : "Basic" }
]
}
7. Specify the application privileges if required. (Optional)
Use the authorization keyword in the .xsaccess file to specify which authorization level is required by a user
for access to a particular application package. The authorization keyword requires a corresponding entry in
the .xsprivileges file, for example, execute for basic privileges or admin for administrative privileges on
the specified package.
{
"authorization":
["sap.xse.test::Execute",
]
}
8. Save the .xsaccess file in the package with which you want to associate the rules you have defined.
9. Commit the .xsaccess file to the repository and activate it.
In the Project Explorer view, right click the object you want to activate and choose Team > Activate in the
popup menu.
Related Links
Application-Access File Keyword Options [page 59]
PUBLI C

The Application-Privileges File [page 68]
In SAP HANA Extended Application Services (SAP HANA XS), the application-privileges (.xsprivileges) file
can be used to create or define the authorization privileges required for access to an SAP HANA XS application,
for example, to start the application or to perform administrative actions on an application. These privileges can
be checked by an application at runtime.
5.5.7 The Application-Privileges File
The application-privileges file has only the file extension .xsprivileges; it does not have a name and is
formatted according to JSON rules. Multiple .xsprivileges files are allowed, but only at different levels in the
package hierarchy; you cannot place two .xsprivileges files in the same application package. The package
privileges defined in a .xsprivileges file are bound to the package to which the .xsprivileges file belongs
and can only be used in this package and its subpackages.
Inside the .xsprivileges file, a privilege is defined by specifying an entry name with an optional description.
This entry name is then automatically prefixed with the package name to form the unique privilege name, for
example, sap.hana::Execute.
As an application privilege is created during activation of an .xsprivileges file, the only user who has the
privilege by default is the _SYS_REPO user. To grant or revoke the privilege to (or from) other users you must use
the GRANT_APPLICATION_PRIVILEGE or REVOKE_APPLICATION_PRIVILEGE procedure in the _SYS_REPO
schema.
Note
The .xsprivileges file lists the authorization levels that are available for access to an application package;
the .xsaccess file defines which authorization level is assigned to which application package.
In the following above, if the application-privileges file is located in the application package sap.hana.xse, then
the following privileges are created:
sap.hana.xse::Execute
sap.hana.xse::Admin
The privileges defined apply to the package where the .xsprivileges file is located as well as any packages
further down the package hierarchy unless an additional .xsprivileges file is present, for example, in a
subpackage. The privileges do not apply to packages that are not in the specified package path, for example,
sap.hana.app1.
Example
The SAP HANA XS Application-Privileges File
The following example shows the composition and structure of a basic SAP HANA XS application-privileges file.
{
"privileges" :
68
PUBLI C

[
{ "name" : "Execute", "description" : "Basic execution
privilege" },
{ "name" : "Admin", "description" : "Administration privilege" }
]
}
If the .xsprivileges file shown in the example above is located in the package sap.hana.xse, you can assign
the Execute privilege for the package to a particular user by calling the GRANT_APPLICATION_PRIVILEGE
procedure, as illustrated in the following code:
call "_SYS_REPO"."GRANT_APPLICATION_PRIVILEGE"('"sap.hana.xse::Execute"', '<user>')
5.5.8 Create an SAP HANA XS Application Privileges File
The application-privileges (.xssprivileges) file can be used to define the authorization levels required for
access to an application, for example, to start the application or perform administrative actions on an application.
You can then assign the application privileges to the individual users who require them.
The .xssprivileges file must reside in the same application package that you want to define the access
privileges for.
Note
If you use the .xsprivileges file to define application-specific privileges, you must also add a corresponding
entry to the .xsaccess file, for example, using the authorization keyword.
1. If you have not already done so, create a root package for your new application, for example, MyPackage.
a) In the SAP HANA studio, open the SAP HANA Development perspective.
package.
2. If you have not already done so, create an application descriptor for your new application and place it in the
root package (MyPackage) you created in the previous step.
content is available to clients.
Note
3. If you have not already done so, create an application-access file for your new application and place it in the
package to which you want to grant access.
The application-access file does not have a name; it only has the file extension .xsaccess.
a) In the SAP HANA studio, open the SAP HANA Development perspective.
PUBLI C

b) In the Project Explorer view, right-click the folder where you want to create the new (.xsaccess) file.
d) Enter the name .xsaccess and choose Finish.
e) Specify the privileges required for access to the application or application package.
Use the authorization keyword in the .xsaccess file to specify which authorization level is required by a
user for access to a particular application package.
Note
If you enable the authorization keyword in the .xsaccess file, you must add a corresponding entry to
the .xsprivileges file, too.
{
"exposed" : true
"authentication" :
[
]
"authorization":
[
"com.acme.myApp::Execute",
"com.acme.myApp::Admin"
]
}
4. Create the application-privileges (.xsprivileges) file and place it in the application package whose access
privileges you want to define.
The application-privileges file does not have a name; it only has the file extension .xsprivileges. The
contents of the .xsprivileges file must be formatted according to JavaScript Object Notation (JSON)
rules.
Note
Multiple .xsprivileges files are allowed, but only at different levels in the package hierarchy; you cannot
place two .privileges files in the same application package. The privileges defined in a .xsprivileges
file are bound to the package to which the file belongs and can only be applied to this package and its
subpackages.
b) In the Project Explorer view, right-click the folder where you want to create the new (.xsprivileges)
file.
d) Enter the name .xsprivileges and choose Finish.
e) Activate the new (.xsprivileges) file
5. Define the required application privileges.
In the .xsprivileges file, you define a privilege for an application package by specifying an entry name with
an optional description. This entry name is then automatically prefixed with the package name in which
the .xsprivileges file is located to form a unique privilege name. For example,
com.acme.myapp::Execute would enable execute privileges on the package com.acme.myapp. The
privilege name is unique to the package to which it belongs and, as a result, can be used in
multiple .xsprivileges files in different packages.
70
PUBLI C

Note
The .xsprivileges file lists the authorization levels defined for an application package. A corresponding
entry is required in the same application's access file .xsaccess file to define which authorization level is
assigned to which application package.
{
"privileges" :
[
{ "name" : "Execute", "description" : "Basic execution
privilege" },
{ "name" : "Admin", "description" : "Administration privilege" }
]
}
The activation of the application privileges creates the corresponding objects, which you can use to assign the
specified privileges to an author.
7. Assign the application privilege to the users who require it.
After activation of the .xsprivileges object, the only user who by default has the application privileges
specified in the .xsprivileges file is the _SYS_REPO user. To grant the specified privilege to (or revoke
them from) other users, use the GRANT_APPLICATION_PRIVILEGE or REVOKE_APPLICATION_PRIVILEGE
procedure in the _SYS_REPO schema.
To grant the execute application privilege to a user, run the following command in the SAP HANA studio's
SQL Console:
call
"_SYS_REPO"."GRANT_APPLICATION_PRIVILEGE"('"com.acme.myApp::Execute"','<UserName>
')
To revoke the execute application privilege to a user, run the following command in the SAP HANA studio's
SQL Console:
call
"_SYS_REPO"."REVOKE_APPLICATION_PRIVILEGE"('"com.acme.myApp::Execute"','<UserName
>')
Related Links
Create an Application Descriptor File [page 55]
Enabling Access to SAP HANA XS Application Packages [page 66]
PUBLI C

5.6 Tutorial: My First SAP HANA Application
This tutorial shows you how SAP HANA development is done in the SAP HANA studio, including setting up a
project, and developing a simple JavaScript file to extract data from the database.
Though extremely simple, the tutorial shows the development process for programming all types of applications.
The project creates server-side JavaScript code that retrieves data by executing SQL in the database, and then
returns the data to the client, which in this case is a browser. The project can be viewed as adhering to the model-
view-controller architecture:
Model: You have (dummy) data in the database, which we can extract via SQL.
Controller: You have JavaScript code that controls the extraction of the data and delivers it to the view.
View: You have a simple browser that calls the JavaScript code and simply displays the data.
5.6.1 Open the Development Perspective
Before you do anything, you have to start the SAP HANA studio and open the SAP HANA Development
perspective.
1. Open the SAP HANA studio. The first screen you see is the Welcome screen, with quick links to the main SAP
HANA perspectives and to the documentation.
2. Select Open Development.
72
PUBLI C

The default SAP HANA Development perspective contains three views on the left for navigating SAP HANA
and the repository, which holds the objects that you will create. In addition, the editor area is on the right,
where you will build your objects.
Related Links
SAP HANA Development Perspective [page 28]
5.6.2 Add a System
You need to add a connection to the development SAP HANA system you will be working with.
1. In the Navigator view, right-click anywhere in the view and select Add System.
PUBLI C

2. Enter the following fields for the SAP HANA system:
Server name
Instance number on that server
A display name for this system. When you start working with a lot of systems, you will want to have a way
to label the systems in the SAP HANA studio. Enter Development System.
3. Select Next.
4. Enter a user name and password for the connection, and select Finish.
74
PUBLI C

After adding the system, you will see the system in the Navigator view.
5.6.3 Add a Workspace
After you add a system, you need to specify where on your workstation you will be saving development files while
working with this system.
1. In the SAP HANA Repositories view, click in the upper-right of the view.
2. Provide the following:
SAP HANA system, which is the same system you just created.
Workspace name, which can be anything you like. For this tutorial, enter DevWS.
A folder with this name is created below the workspace root.
Workspace root, which can be anywhere on your workstation. For this tutorial, create a folder at C:
\SAPHANAworkspaces and make this the root.
PUBLI C

3. Click Finish.
In the SAP HANA Repositories view, you will see your workspace, which enables you to browse the repository of
the system tied to this workspace. You will see the repository's packages, displayed as folders.
At the same time, a folder will be added to your file system to hold all your development files.
76
PUBLI C

5.6.4 Add a Project
Now that you've set up your development environment for your SAP HANA system, you can add an Eclipse
project to contain all the development objects you want to create.
There are a variety of project types for different types of development objects. Generally, these projects types
import necessary libraries for working with specific types of development objects. Here, you will create an XS
Project.
1. From the menu, select File New Project .
2. Under SAP HANA Development, select XS Project, and select Next.
3. Enter the following for the project:
Name: Enter mycompany.myorg.testing.
Since Eclipse project names must be unique within the same Eclipse workspace, a good convention is to
use the fully qualified package name as the project name.
Project Location: You can keep this as the default Eclipse workspace.
4. Select Finish.
Now, in the Project Explorer view, you have a project that is ready to be shared, that is, associated with your
workspace.
PUBLI C

5.6.5 Share Your Project
After creating a project, you must associate it with a workspace, which enables the project files to be saved to the
repository of your development SAP HANA system.
1. In the Project Explorer view, right-click on the project, and select Team Share Project .
Since you only have one workspace, the wizard selects it for you. If you had several workspaces, you would
choose with which one you wanted to share your project.
The dialog also shows the Current project location (the current location of your project, in the Eclipse
workspace), and the New project location (where your project will be copied so it can be associated with the
repository workspace).
Also, since Add project folder as subpackage is checked, subpackages will be created based on the name of
your project.
78
PUBLI C

2. Click Finish.
Now your project appears in the Project Explorer view associated with your workspace.
The .project file is shown with an asterisk , which indicates that the file has changed but has yet to be
committed, or saved, to the repository.
3. Right-click on the project, and select Team Commit . This adds your project and its files to the
repository, though only you can see them. The .project file is now displayed with a diamond, , indicating
that the latest version on your workstation has been committed.
In addition, the SAP HANA Repositories view shows that a new hierarchy of packages has been created based
on the name of your project, mycompany.myorg.testing.
5.6.6 Write Server-Side JavaScript
Now let's write some code. As a first step, to make sure everything is working OK, you will simply output Hello,
World!.
You will have to create three files:
.xsjs: This contains your code.
PUBLI C

.xsapp: This indicates that everything in your package can be exposed via HTTP. You still need to explicitly
expose the content and assign access controls.
.xsaccess: Exposes your content, meaning it can be accessed via HTTP, and assigned access controls.
1. Right-click your project, and select New Other .
2. Select SAP HANA Development XS JavaScript Source File .
3. In File name, call your JavaScript file MyFirstSourceFile.xsjs, and select Finish.
The file, which is blank, opens in the JavaScript editor.
4. In the MyFirstSourceFile.xsjs file, enter the following code:
$.response.contentType = "text/html";
$.response.setBody( "Hello, World !");
This uses the SAP HANA XS JavaScript API's response object to write out HTML. By typing $. you have
access to the API's objects.
5. Add a blank file called .xsapp (no name, just a file extension) by right-clicking to the root of your project.
More on this later.
To add a file, right-click the project and select New File , enter a file name, and select Finish.
6. Add a file called .xsaccess (no name, just a file extension), and copy the following code:
{
"exposed" : true,
"authentication" :
[
]
}
This code exposes the contents via HTTP, and requires you to log in with your SAP HANA credentials to
access the file.
7. Right-click on the project, and select Team Commit . This adds your four new files
(.xsjs, .xsapp, .xsaccess and .project) to the repository, though only you can see the files and no one
can run them.
8. Right-click on the project, and select Team Activate . This publishes your work and you can now test it.
To access your JavaScript code, open a browser and enter the URL:
http://myServer:8000/mycompany/myorg/testing/MyFirstSourceFile.xsjs
Note
Change the server name to your server.
The port is 80 plus two digits for your instance. If your instance is 00, then the port is 8000.
You should get the following:
80
PUBLI C

After logging in with your SAP HANA user name and password, you should get the following:
5.6.7 Retrieve Data
To make things more interesting, let's make a call to the database via the JavaScript and then display the results.
You will be using the SQL statement:
select * from DUMMY
This is test SQL to check connectivity, and returns one row with one field called DUMMY, whose value is X.
1. In MyFirstSourceFile.xsjs, delete or comment out all your existing code.
2. Add the following code:
$.response.contentType = "text/html";
var output = "Hello, World !<br><br>";
var conn = $.db.getConnection();
var pstmt = conn.prepareStatement( "select * from DUMMY" );
var rs = pstmt.executeQuery();
PUBLI C

if (!rs.next()) {
$.response.setBody( "Failed to retrieve data" );
$.response.status = $.net.http.INTERNAL_SERVER_ERROR;
} else {
output = output + "This is the response from my SQL: " + rs.getString(1);
}
rs.close();
pstmt.close();
conn.close();
$.response.setBody(output);
3. Save the file.
4. Commit the file by right-clicking the file and selecting Team Commit .
5. Activate the file by right-clicking the file and selecting Team Activate .
In your browser, refresh the page. You should get the following:
82
PUBLI C

6 Setting Up the Persistence Model
In SAP HANA Extended Application Services (SAP HANA XS), the persistence model defines the schema, tables,
and views that specify what data to make accessible and how. The persistence model is mapped to the
consumption model that is exposed to client applications and users, so that data can be analyzed and displayed.
SAP HANA XS enables you to create database schema, tables, views, and sequences as design-time files in the
repository. Repository files can be read by applications that you develop.
Note
All repository files including your view definition can be transported (along with tables, schema, and
sequences) to other SAP HANA systems, for example, in a delivery unit. A delivery unit is the medium SAP
HANA provides to enable you to assemble all your application-related repository artifacts together into an
archive that can be easily exported to other systems.
You can also set up data-provisioning rules and save them as design-time objects so that they can be included in
the delivery unit that you transport between systems.
As part of the process of setting up the basic persistence model for SAP HANA XS, you perform the following
tasks:
Create a schema.
Define a design-time schema and maintain the schema definition in the repository. The transportable schema
has the file extension .hdbschema, for example, MYSCHEMA.hdbschema
Create a table.
Define a design-time table and maintain the table definition in the repository. The transportable table has the
file extension .hdbtable, for example, MYTABLE.hdbtable
Create a view.
Define a design-time view and maintain the view definition in the repository. The transportable view has the
file extension .hdbview, for example, MYVIEW.hdbview
Create a sequence.
Define a design-time sequence and maintain the sequence definition in the repository. The transportable
sequence has the file extension .hdbsequence, for example, MYSEQUENCE.hdbsequence
Import table content
Define data-provisioning rules that enable you to import data from comma-separated values (CSV) files into
SAP HANA tables using the SAP HANA XS table-import feature. The complete configuration can be included
in a delivery unit and transported between SAP HANA systems.
Note
On activation of a repository file, the file suffix, for example, .hdbview, .hdbschema, or .hdbtable, is used to
determine which runtime plug-in to call during the activation process. The plug-in reads the repository file
selected for activation, for example, a table, sees the object descriptions in the file, and creates the appropriate
runtime object.
Setting Up the Persistence Model
PUBLI C

6.1 Schema
Relational databases contain a catalog that describes the various elements in the system. The catalog divides the
database into sub-databases known as schema. A database schema enables you to logically group together
objects such as tables, views, and stored procedures. Without a defined schema, you cannot write to the catalog.
SAP HANA Extended Application Services (SAP HANA XS) enables you to create a database schema as a
transportable design-time file in the repository. Repository files can be read by applications that you develop.
If your application refers to the repository (design-time) version of a schema rather than the runtime version in
the catalog, for example, by using the explicit path to the repository file (with suffix), any changes to the
repository version of the file are visible as soon as they are committed to the repository. There is no need to wait
for the repository to activate a runtime version of the schema.
If you want to define a transportable schema using the design-time hdbschema specifications, use the
configuration schema illustrated in the following example:
string schema_name
The following example shows the contents of a valid transportable schema-definition file for a schema called
MYSCHEMA:
schema_name=MYSCHEMA;
The schema is stored in the repository with the schema name MYSCHEMA as the file name and the
suffix .hdbschema, for example, MYSCHEMA.hdbschema.
Schema Activation
If you want to create a schema definition as a design-time object, you must create the schema as a flat file. You
save the file containing the schema definition with the suffix .hdbschema in the appropriate package for your
application in the SAP HANA repository. You can activate the design-time objects at any point in time.
Note
On activation of a repository file, the file suffix, for example, .hdbschema, is used to determine which runtime
plugin to call during the activation process. The plug-in reads the repository file selected for activation, parses
the object descriptions in the file, and creates the appropriate runtime objects.
If you activate a schema-definition object in SAP HANA, the activation process checks if a schema with the same
name already exists in the SAP HANA repository. If a schema with the specified name does not exist, the
repository creates a schema with the specified name and makes _SYS_REPO the owner of the new schema.
Note
The schema cannot be dropped even if the deletion of a schema object is activated.
If you define a schema in SAP HANA XS, note the following important points regarding the schema name:
84
PUBLI C

Name mapping
The schema name must be identical to the name of the corresponding repository object.
Naming conventions
The schema name must adhere to the SAP HANA rules for database identifiers.
Name usage
The Data Definition Language (DDL) rendered by the repository contains the schema name as a delimited
identifier.
6.2 Creating Schemas
A schema defines the container that holds database objects such as tables, views, and stored procedures.
This task describes how to create a file containing a schema definition. Schema definition files are stored in the
SAP HANA repository. To complete this task successfully, note the following prerequisites:
You must have access to a SAP HANA system.
You must have already created a development workspace and a project.
You must have shared the project so that the newly created files can be committed to (and synchronized
with) the repository.
SAP HANA Extended Application Services (SAP HANA XS) enables you to create a database schema as a design-
time file in the repository.
To create a schema definition file in the repository, perform the following steps:
4. Create the schema definition file.
Browse to the folder in your project workspace where you want to create the new schema-definition file and
perform the following tasks:
a) Right-click the folder where you want to save the schema-definition file and choose New in the context-
sensitive popup menu.
b) Enter the name of the schema in the File Name box and add the file suffix .hdbschema, for example,
MYSCHEMA.hdbschema.
c) Choose Finish to save the new schema in the repository.
5. Define the schema name.
To edit the schema file, in the Project Explorer view double-click the schema file you created in the previous
step, for example, MYSCHEMA.hdbschema, and add the schema-definition code to the file:
Note
The following code example is provided for illustration purposes only.
schema_name=MYSCHEMA;
6. Save the schema file.
PUBLI C

7. Commit the schema file to the repository.
a) Locate and right-click the new schema file in the Project Explorer view.
b) In the context-sensitive pop-up menu, choose Team Commit .
8. Activate the schema.
a) Locate and right-click the new schema file in the Project Explorer view.
b) In the context-sensitive pop-up menu, choose Team Activate .
9. Grant SELECT privileges to the owner of the new schema.
After activation in the repository, the schema object is only visible in the catalog to the _SYS_REPO user. To
enable other users, for example the schema owner, to view the newly created schema in the SAP HANA
studio's Modeler perspective, you must grant the user the required SELECT privilege.
a) In the SAP HANA studio Navigator view, right-click the SAP HANA system hosting the repository where
the schema was activated and choose SQL Console in the context-sensitive popup menu.
b) In the SQL console, execute the statement illustrated in the following example, where <SCHEMANAME> is
the name of the newly activated schema, and <username> is the database user ID of the schema owner:
call
_SYS_REPO.GRANT_SCHEMA_PRIVILEGE_ON_ACTIVATED_CONTENT('select','<SCHEMANAME>',
'<username>');
Related Links
Using SAP HANA Projects [page 42]
6.3 Tables
In the SAP HANA database, as in other relational databases, a table is a set of data elements that are organized
using columns and rows. A database table has a specified number of columns, defined at the time of table
creation, but can have any number of rows. Database tables also typically have meta-data associated with them;
the meta-data might include constraints on the table or on the values within particular columns.
SAP HANA Extended Application Services (SAP HANA XS) enables you to create a database table as a design-
time file in the repository. All repository files including your table definition can be transported to other SAP HANA
systems, for example, in a delivery unit.
Note
A delivery unit is the medium SAP HANA provides to enable you to assemble all your application-related
repository artifacts together into an archive that can be easily exported to other systems.
If your application is configured to use the design-time version of a database table in the repository rather than
the runtime version in the catalog, any changes to the repository version of the table are visible as soon as they
are committed to the repository. There is no need to wait for the repository to activate a runtime version of the
table.
86
PUBLI C

If you want to define a transportable table using the design-time .hdbtable specifications, use the configuration
schema illustrated in the following example:
struct TableDefinition {
string SchemaName;
optional bool temporary;
optional TableType tableType;
optional bool public;
optional TableLoggingType loggingType;
list<ColumnDefinition> columns;
optional list<IndexDefinition> indexes;
optional PrimaryKeyDefinition primaryKey;
optional string description
};
The following code illustrates a simple example of a design-time table definition:
table.schemaName = "MYSCHEMA";
table.tableType = COLUMNSTORE;
table.columns = [
{name = "Col1"; sqlType = VARCHAR; nullable = false; length = 20; comment =
"dummy comment";},
{name = "Col2"; sqlType = INTEGER; nullable = false;},
{name = "Col3"; sqlType = NVARCHAR; nullable = true; length = 20; defaultValue
= "Defaultvalue";},
{name = "Col4"; sqlType = DECIMAL; nullable = false; precision = 2; scale =
3;}];
table.indexes = [
{name = "MYINDEX1"; unique = true; indexColumns = ["Col2"];},
{name = "MYINDEX2"; unique = true; indexColumns = ["Col1", "Col4"];}];
table.primaryKey.pkcolumns = ["Col1", "Col2"];
If you want to create a database table as a repository file, you must create the table as a flat file and save the file
containing the table dimensions with the suffix .hdbtable, for example, MYTABLE.hdbtable. The new file is
located in the package hierarchy you establish in the SAP HANA repository. You can activate the repository files at
any point in time.
Note
On activation of a repository file, the file suffix, for example, .hdbtable, is used to determine which runtime
plug-in to call during the activation process. The plug-in reads the repository file selected for activation, in this
case a table, parses the object descriptions in the file, and creates the appropriate runtime objects.
Related Links
Table Configuration Schema [page 87]
time file in the repository. The design-time artifact that contains the table definition must adhere to
the .hdbtable syntax specified below.
6.4 Table Configuration Schema
time file in the repository. The design-time artifact that contains the table definition must adhere to
the .hdbtable syntax specified below.
PUBLI C

Table Definition
The following code illustrates a simple example of a design-time table definition using the .hdbtable syntax.
Note
Keywords are case-sensitive, for example, tableType and loggingType, and the schema referenced in the table
definition, for example, MYSCHEMA, must already exist.
table.loggingType = NOLOGGING;
table.columns = [
"dummy comment";},
= "Defaultvalue";},
3;}];
table.indexes = [
Table-Definition Configuration Schema
The following example shows the configuration schema for tables defined using the .hdbtable syntax. Each of
the entries in the table-definition configuration schema is explained in more detail in a dedicated section below:
string SchemaName;
optional string description
};
Schema Name
To use the .hdbtable syntax to specify the name of the schema that contains the table you are defining, use the
schemaName keyword. In the table definition, the schemaName keyword must adhere to the syntax shown in the
following example.
88
PUBLI C

Temporary
To use the .hdbtable syntax to specify that the table you define is temporary, use the boolean temporary
keyword. Since data in a temporary table is session-specific, only the owner session of the temporary table is
allowed to INSERT/READ/TRUNCATE the data. Temporary tables exist for the duration of the session, and data
from the local temporary table is automatically dropped when the session is terminated. In the table definition, the
temporary keyword must adhere to the syntax shown in the following example.
table.temporary = true;
Table Type
To specify the table type using the .hdbtable syntax, use the tableType keyword. In the table definition, the
TableType keyword must adhere to the syntax shown in the following example.
The following configuration schema illustrates the parameters you can specify with the tableType keyword:
enum TableType {
COLUMNSTORE;
ROWSTORE;
};
Table Logging Type
To enable logging in a table definition using the .hdbtable syntax, use the tableLoggingType keyword. In the
table definition, the tableLoggingType keyword must adhere to the syntax shown in the following example.
table.tableLoggingType = LOGGING;
The following configuration schema illustrates the parameters you can specify with the tableLoggingType
keyword:
enum TableLoggingType {
LOGGING;
NOLOGGING;
};
Table Column Definition
To define the column structure and type in a table definition using the .hdbtable syntax, use the columns
keyword. In the table definition, the columns keyword must adhere to the syntax shown in the following example.
table.columns = [
PUBLI C

"dummy comment";},
= "Defaultvalue";},
3;}];
The following configuration schema illustrates the parameters you can specify with the columns keyword:
struct ColumnDefinition {
string name;
SqlDataType sqlType;
optional bool nullable;
optional bool unique;
optional int32 length;
optional int32 scale;
optional int32 precision;
optional string defaultValue;
optional string comment;
};
SQL Data Type
To define the SQL data type for a column in a table using the .hdbtable syntax, use the sqlType keyword. In the
table definition, the sqlType keyword must adhere to the syntax shown in the following example.
table.columns = [
"dummy comment";},
...
];
The following configuration schema illustrates the data types you can specify with the sqlType keyword:
enum SqlDataType {
DATE; TIME; TIMESTAMP; SECONDDATE; INTEGER; TINYINT;
SMALLINT; BIGINT; REAL; DOUBLE; FLOAT; SMALLDECIMAL;
DECIMAL; VARCHAR; NVARCHAR; CHAR; NCHAR;CLOB; NCLOB;
ALPHANUM; TEXT; SHORTTEXT; BLOB; VARBINARY;
};
Table Order
To define the table order type using the .hdbtable syntax, use the order keyword. In the table definition, the
order keyword must adhere to the syntax shown in the following example.
table.order = ASC;
The following configuration schema illustrates the parameters you can specify with the order keyword:
enum Order {
ASC;
DSC;
};
90
PUBLI C

You can choose to filter the table contents either by ascending (ASC) or descending (DSC) order.
Primary Key Definition
To define the primary key for the specified table using the .hdbtable syntax, use the primaryKey and pkcolumns
keywords. In the table definition, the primaryKey and pkcolumns keywords must adhere to the syntax shown in
the following example.
The following configuration schema illustrates the parameters you can specify with the primaryKey keyword:
struct PrimaryKeyDefinition {
list<string> pkcolumns;
optional IndexType indexType;
};
Table Index Definition
To define the index type for the specified table using the .hdbtable syntax, use the indexes keyword. In the table
definition, the indexes keyword must adhere to the syntax shown in the following example.
table.indexes = [
The following configuration schema illustrates the parameters you can specify with the indexes keyword:
struct IndexDefinition {
string name;
bool unique;
optional Order order;
list<string> indexColumns;
};
Table Index Type
To define the index type for the specified table using the .hdbtable syntax, use the indexType keyword. In the
table definition, the indexType keyword must adhere to the syntax shown in the following example.
table.indexType = B_TREE;
The following configuration schema illustrates the parameters you can specify with the indexType keyword:
enum IndexType {
B_TREE;
PUBLI C

CPB_TREE;
};
B_TREE specifies an index tree of type B+, which maintains sorted data that performs the insertion, deletion, and
search of records. CPB_TREE stands for Compressed Prefix B_TREE and specifies an index tree of type CPB+,
which is based on pkB-tree. CPB_TREE is a very small index that uses a partial key, that is; a key that is only part
of a full key in index nodes.
Note
If neither the B_TREE nor the CPB_TREE is specified in the table-definition file, SAP HANA chooses the
appropriate index type based on the column data type, as follows:
CPB_TREE
Character string types, binary string types, decimal types, when the constraint is a composite key or a non-
unique constraint
B_TREE
All column data types other than those specified for CPB_TREE
Complete Table-Definition Configuration Schema
The following example shows the complete configuration schema for tables defined using the .hdbtable syntax.
enum TableType {
COLUMNSTORE; ROWSTORE;
};
enum TableLoggingType {
LOGGING; NOLOGGING;
};
enum IndexType {
B_TREE; CPB_TREE;
};
enum Order {
ASC; DSC;
};
enum SqlDataType {
DATE; TIME; TIMESTAMP; SECONDDATE;
INTEGER; TINYINT; SMALLINT; BIGINT;
REAL; DOUBLE; FLOAT; SMALLDECIMAL; DECIMAL;
VARCHAR; NVARCHAR; CHAR; NCHAR; CLOB; NCLOB;
ALPHANUM; TEXT; SHORTTEXT; BLOB; VARBINARY;
};
struct PrimaryKeyDefinition {
list<string> pkcolumns;
};
struct IndexDefinition {
string name;
bool unique;
optional Order order;
list<string> indexColumns;
};
struct ColumnDefinition {
string name;
SqlDataType sqlType;
optional bool nullable;
optional bool unique;
92
PUBLI C

optional int32 length;
optional int32 scale;
optional int32 precision;
optional string defaultValue;
optional string comment;
};
string schemaName;
optional string description;
};
TableDefinition table;
Related Links
Tables [page 86]
In the SAP HANA database, as in other relational databases, a table is a set of data elements that are organized
using columns and rows. A database table has a specified number of columns, defined at the time of table
creation, but can have any number of rows. Database tables also typically have meta-data associated with them;
the meta-data might include constraints on the table or on the values within particular columns.
Creating Tables [page 93]
6.5 Creating Tables
This task describes how to create a file containing a table definition. Table definition files are stored in the SAP
HANA repository. To complete this task successfully, note the following prerequisites:
You must have created a schema definition MYSCHEMA.hdbschema
To create a table file in the repository, perform the following steps:
4. Create the table definition file.
Browse to the folder in your project workspace where you want to create the new table file and perform the
following tasks:
PUBLI C

a) Right-click the folder where you want to save the table file and choose New in the context-sensitive popup
menu.
b) Enter the name of the table in the File Name box and add the file suffix .hdbtable, for example,
MYTABLE.hdbtable.
c) Choose Finish to save the new table-definition file.
5. Define the table.
To edit the table file, in the Project Explorer view double-click the table file you created in the previous step, for
example, MYTABLE.hdbtable, and add the table-definition code to the file:
Note
table.columns = [
"dummy comment";},
{name = "Col3"; sqlType = NVARCHAR; nullable = true; length = 20;
defaultValue = "Defaultvalue";},
3;}];
table.indexes = [
6. Save the table file.
7. Commit the changes to the repository.
a) Locate and right-click the new table file in the Project Explorer view.
Related Links
Creating Schemas [page 85]
A schema defines the container that holds database objects such as tables, views, and stored procedures.
6.6 Sequences
A sequence is a database object that generates an automatically incremented list of numeric values according to
the rules defined in the sequence specification. The sequence of numeric values is generated in an ascending or
descending order at a defined increment interval, and the numbers generated by a sequence can be used by
applications, for example, to identify the rows and columns of a table.
Sequences are not associated with tables; they are used by applications, which can use CURRVAL in a SQL
statement to get the current value generated by a sequence and NEXTVAL to generate the next value in the
defined sequence. Sequences provide an easy way to generate the unique values that applications use, for
example, to identify a table row or a field. In the sequence specification, you can set options that control the start
94
PUBLI C

and end point of the sequence, the size of the increment size, or the minimum and maximum allowed value. You
can also specify if the sequence should recycle when it reaches the maximum value specified. The relationship
between sequences and tables is controlled by the application. Applications can reference a sequence object and
coordinate the values across multiple rows and tables.
SAP HANA Extended Application Services (SAP HANA XS) enables you to create a database sequence as a
transportable design-time file in the repository. Repository files can be read by applications that you develop.
You can use database sequences to perform the following operations:
Generate unique, primary key values, for example, to identify the rows and columns of a table
Coordinate keys across multiple rows or tables
If you want to define a transportable sequence using the design-time sequence specifications, use the
configuration schema illustrated in the following example.
string schema;
int32 increment_by(default=1);
int32 start_with(default=-1);
optional int32 maxvalue;
bool nomaxvalue(default=false);
optional int32 minvalue;
bool nominvalue(default=false);
optional bool cycles;
optional string reset_by;
bool public(default=false);
optional string depends_on_table;
optional string depends_on_view;
The following example shows the contents of a valid sequence-definition file for a sequence called MYSEQUENCE.
Note that, in this example, no increment value is defined, so the default value of 1 (ascend by 1) is assumed. To set
a descending sequence of 1, set the increment_by value to -1.
schema= "TEST_DUMMY";
start_with= 10;
maxvalue= 30;
nomaxvalue=false;
minvalue= 1;
nominvalue=true;
cycles= false;
reset_by= "SELECT \"Col2\" FROM \"TEST_DUMMY\".\"com.sap.test.tables::MY_TABLE\"
WHERE \"Col2\"='12'";
depends_on_table= "com.sap.test.tables::MY_TABLE";
The sequence definition is stored in the repository with the suffix hdbsequence, for example,
MYSEQUENCE.hdbsequence.
If you activate a sequence-definition object in SAP HANA XS, the activation process checks if a sequence with the
same name already exists in the SAP HANA repository. If a sequence with the specified name does not exist, the
repository creates a sequence with the specified name and makes _SYS_REPO the owner of the new sequence.
Note
The sequence cannot be dropped even if the deletion of a sequence object is activated.
PUBLI C

6.7 Creating Sequences
A database sequence generates a serial list of unique numbers that you can use while transforming and moving
data to between systems.
This task describes how to create a file containing a sequence definition. Sequence-definition files are stored in
the SAP HANA repository. To complete this task successfully, note the following prerequisites:
SAP HANA Extended Application Services (SAP HANA XS) enables you to create a database sequence as a
design-time file in the repository.
To create a sequence definition file in the repository, perform the following steps:
4. Create the sequence-definition file.
Browse to the folder in your project workspace where you want to create the new sequence-definition file and
a) Right-click the folder where you want to save the sequence-definition file and choose New in the context-
b) Enter the name of the sequence in the File Name box and add the file suffix .hdbsequence, for example,
MYSEQUENCE.hdbsequence.
c) Choose Finish to save the new sequence in the repository.
5. Define the sequence properties.
To edit the sequence file, in the Project Explorer view double-click the sequence file you created in the
previous step, for example, MYSEQUENCE.hdbsequence, and add the sequence code to the file:
schema= "TEST_DUMMY";
start_with= 10;
maxvalue= 30;
nomaxvalue=false;
minvalue= 1;
nominvalue=true;
cycles= false;
reset_by= "SELECT \"Col2\" FROM \"TEST_DUMMY\".\"com.sap.test.tables::MY_TABLE\"
WHERE \"Col2\"='12'";
depends_on_table= "com.sap.test.tables::MY_TABLE";
6. Save the sequence file.
a) Locate and right-click the new sequence file in the Project Explorer view.
Related Links
96
PUBLI C

6.8 SQL Views
In SQL, a view is a virtual table based on the dynamic results returned in response to an SQL statement. Every
time a user queries an SQL view, the database uses the view's SQL statement to recreate the data specified in the
SQL view. The data displayed in an SQL view can be extracted from one or more database tables.
An SQL view contains rows and columns, just like a real database table; the fields in an SQL view are fields from
one or more real tables in the database. You can add SQL functions, for example, WHERE or JOIN statements, to a
view and present the resulting data as if it were coming from one, single table.
SAP HANA Extended Application Services (SAP HANA XS) enables you to create a database view as a design-time
file in the repository. Repository files can be read by applications that you develop. In addition, all repository files
including your view definition can be transported to other SAP HANA systems, for example, in a delivery unit.
If your application refers to the design-time version of a view from the repository rather than the runtime version
in the catalog, for example, by using the explicit path to the repository file (with suffix), any changes to the
repository version of the file are visible as soon as they are committed to the repository. There is no need to wait
for the repository to activate a runtime version of the view.
To define a transportable view using the design-time view specifications, use the configuration schema illustrated
in the following example:
string schema;
string query;
bool public(default=true);
optional list<string> depends_on_table;
optional list<string> depends_on_view;
The following example shows the contents of a valid transportable view-definition file for a view called MYVIEW:
schema="TEST_DUMMY";
query="SELECT * FROM \"TEST_DUMMY\".\"acme.com.test.tables::02_HDB_DEPARTMENT_VIEW
\"";
depends_on_view=["acme.com.test.tables::02_HDB_DEPARTMENT_VIEW"];
If you want to create a view definition as a design-time object, you must create the view as a flat file and save the
file containing the view definition with the suffix .hdbview, for example, MYVIEW.hdbview in the appropriate
package in the package hierarchy established for your application in the SAP HANA repository. You can activate
the design-time object at any point in time.
Note
On activation of a repository file, the file suffix (for example, .hdbview) is used to determine which runtime
plugin to call during the activation process. The plug-in reads the repository file selected for activation, parses
the object descriptions in the file, and creates the appropriate runtime objects.
Column Names in a View
If you want to assign names to the columns in a view, use the SQL query in the .hdbview file. In this example of
design-time view definition, the following names are specified for columns defined in the view:
PUBLI C

idea_id
identity_id
role_id
schema = "MYSCHEMA";
query = "select role_join.idea_id as idea_id, ident.member_id as identity_id,
role_join.role_id as role_id
from \"acme.com.odin.db.iam::t_identity_group_member_transitive\" as
ident
inner join \"acme.com.odin.db.idea::t_idea_identity_role\" as role_join
on role_join.identity_id = ident.group_id union distinct
select idea_id, identity_id, role_id
from \"acme.com.odin.db.idea::t_idea_identity_role\"
with read only";
6.9 Creating SQL Views
A view is a virtual table based on the dynamic results returned in response to an SQL statement. SAP HANA
Extended Application Services (SAP HANA XS) enables you to create a database view as a design-time file in the
repository.
This task describes how to create a file containing an SQL view definition. SQL view definition files are stored in
the SAP HANA repository. To complete this task successfully, note the following prerequisites:
To create a view-definition file in the repository, perform the following steps:
4. Create the view-definition file.
Browse to the folder in your project workspace where you want to create the new view-definition file and
a) Right-click the folder where you want to save the view-definition file and choose New in the context-
b) Enter the name of the view-definition file in the File Name box and add the file suffix .hdbview, for
example, MYVIEW.hdbview.
c) Choose Finish to save the new view-definition file in the repository.
5. Define the view.
To edit the view-definition file, in the Project Explorer view double-click the view-definition file you created in
the previous step, for example, MYVIEW.hdbview, and add the view-definition code to the file:
Note
98
PUBLI C

schema="TEST_DUMMY";
query="SELECT * FROM \"TEST_DUMMY\".\"com.sap.test.tables::02_HDB_DEPARTMENT_VIEW
\"";
depends_on_view=["com.sap.test.tables::02_HDB_DEPARTMENT_VIEW"];
6. Save the view-definition file.
a) Locate and right-click the new view-definition file in the Project Explorer view.
Related Links
6.10 Data Provisioning Using Table Import
You can import data from comma-separated values (CSV) into the SAP HANA tables using the SAP HANA
Extended Application Services (SAP HANA XS) table-import feature.
In SAP HANA XS, you create a table-import scenario by setting up an import-model file, a import-data file, and one
or more comma-separated value (CSV) files containing the content you want to import into the specified SAP
HANA table. The import-model file links the import operation to one or more target tables. The table definition (for
example, in the form of a .hdbtable file) can either be created separately or be part of the table-import scenario
itself.
To use the SAP HANA XS table-import feature to import data into an SAP HANA table, you need to understand the
following table-import concepts:
Table-Import Model
You define the table-import model in a configuration file that specifies the data fields to import and import
target tables of each data field.
Table-Import Data
You define the table-import data in a configuration file that specifies how to link the CSV data files to the
respective data field and, in turn, the target tables via the table-import model.
Table-Import Extension
You define the table-import extension in a configuration file that modifies the relationship to an existing table
import data file.
CSV Data File Constraints
The following constraints apply to the table-import feature in SAP HANA XS:
The number of table columns must match the number of CSV columns.
There must not be any incompatibilities between the data types of the table columns and the data types of the
CSV columns.
PUBLI C

Overlapping data in data files is not supported.
The target table of the import must not be modified (or appended to) outside of the data-import operation. If
the table is used for storage of application data, this data may be lost during any operation to re-import or
update the data.
6.10.1 Table-Import Model
The table-import model is a configuration file that you use to specify the parameters associated with the list of
comma-separated-value (CSV) files that you want to import. The table-import model also specifies, for each
parameter, the target table into which the data in the linked-data files is imported.
You can declare a parameter as optional. If a parameter is optional, it is not necessary to assign a value to this
parameter in the implementing table import data object. To declare a variable as optional in the table-import
model, add the keyword optional before the variable, as illustrated in the example below.
optional listCsvFile csvList;
If you want to add comments to your table-import model, use two back slashes (//) as illustrated in the example
below.
listCsvFile fileList; //this is the csv file list
Example
Example Table-Import Model File
listCsvFile csvFiles;
import csvFiles "TISCHEMA" "TiPackage::TiTable";
6.10.2 Table-Import Data
The table-import data configuration file enables you to specify which table-import model is to be implemented, for
example, TiModel.hdbtim and how to handle the data to be imported from the comma-separated list to the
target database table. The file-list parameters created in the table import-model file are assigned to lists of data-
file links, for example, comma-separated values (.csv) files. In the example illustrated below, the .csv is called
TiCsv.csv.
If you want to add comments to your table-import data-definition file, use two back slashes (//) as illustrated in
the following example:
implements TiPackage:TiModel.hdbtim; //This is a the import model
For each CSV file it is possible to define the following options in the data-definition (.hdbtid) file:
Final
Prevent a variable from being modified, for example, in a table-import extension file.
100
PUBLI C

Example
Example Table-Import Data File
implements TiPackage:TiModel.hdbtim; //This is a comment
csvFiles = ["TiPackage:TiCsv.csv"];
final
Use the final keyword with the csvFiles or csvFilesExt to you can prevent modifications of values assigned to file
lists, for example, in a table-import extension. To declare an option as final in the table-import data file, add the
keyword final before the option, as illustrated in the example below.
final csvFiles = ["acme.com.mypackage:data.csv"];
6.10.3 Table-Import Extensions
The table-import feature enables you to modify the initial settings for variables defined in table-import
configuration files by specifying new settings for defined variables in separate extension files.
If both a table-import model and a table-import data file exist, then you can modify initial variable assignments
made in one table-import data file in a subsequent table-import data files (so called extensions) - as long as the
variables are not declared explicitly as final in a previous file. The modification takes place only on the data values
of variables, for example, in a table-import-extension data file, where values can be augmented or be replaced.
The result is a chain of table-import data artifacts. The end result of a successful table import operation is based
on the resulting definition of the file-list parameter as a result of all data files in the chain.
Note
You can only create a linear chain of extensions. It is not permitted to create two table import data files
modifying the same base table import data files.
The following example illustrates a simple table-import model named timodel.hdbtim in the package
TiPackage.
csvFileList csvList;
import csvList "TISCHEMA" "TiPackage::TiTable";
The following example illustrates the contents of a simple table-import data file named tidata.hdbtid, which
implements the table-import model defined in timodel.hdbtim illustrated above.
implements TiPackage:timodel.hdbtim;
csvList = ["TiPackage:mydata.csv"];
PUBLI C

The following example illustrates how to extend the definition in the table-import data file tidata.hdbtid
illustrated above.
modifies sap.myPackage:tidata.hdbtid;
csvList = csvList + ["TiPackage:myotherdata.csv"];
After activation of the table-import files, the data from both CSV files (TiPackage:mydata.csv, and
TiPackage:myotherdata.csv) is imported into the target table.
6.10.4 Table-Import Configuration-File Syntax
When you define the details of the table-import operation, you use the following file-list types to enable the
processing of certain data formats:
csvFileList
Used to specify list of comma-separated values (CSV) file links. The linked CSV files are imported into the
target table specified in the table import model. The import process parses and validates the CSV file format.
Example
Table Import Model Example
The following example shows how to use the option csvFileList in the table-import model definition:
Example
Table Import Data Example
The following example shows how to use the option csvFileList in the table-import data definition:
implements TiPackage:TiModel.hdbtim;
6.11 Importing Data Using Table Import
The table-import function imports data from comma-separated values (CSV) files into SAP HANA tables.
Before you start this task, make sure that the following prerequisites are met:
An SAP HANA database instance is available.
The SAP HANA database client is installed and configured.
You have a database user account set up with the roles containing sufficient privileges to perform actions in
the repository, for example, add packages, add objects, and so on.
102
PUBLI C

The SAP HANA studio is installed and connected to the SAP HANA repository.
You have a development environment including a repository workspace, a package structure for your
application, and a shared project to enable you to synchronize changes to the project files in the local file
system with the repository.
Note
The names used in the following task are for illustration purposes only; replace the names of schema, tables,
files, and so on shown in the following examples with your own names.
1. Create a root package for your table import application.
In SAP HANA studio, open the SAP HANA Development perspective and perform the following steps:
a) In in the package hierarchy displayed in the Navigator view, right-click the package where you want to
create the new package for your table-import configuration and choose New > Package... .
b) Enter a name for your package, for example TiPackage.
Note
Naming conventions exist for package names, for example, a package name must not start with either
a dot (.) or a hyphen (-) and cannot contain two or more consecutive dots (..). In addition, the name
must not exceed 190 characters.
a) Choose OK to create the new package.
2. Create a set of table-import files.
The following files are required for a table import scenario.
The table-import data file, for example, TiData.hdbtid
Links the CSV data files to the data field and thus the target tables via the model
A table-import model, for example, TiModel.hdbtim
Defines the data fields and import target tables of each data field
A table import package, for example, TiPackage
Contains all the artifacts you need to complete the table-import operation
A CSV file, for example, TiCsv.csv
Contains the data to be imported into the SAP HANA table
A table definition, for example, TiTable.hdbtable
Defines the structure of the target import table
The schema definition, for example, TISCHEMA.hdbschema
Specifies the name of the schema in which the target import table is created
An application-logic, for example, TiConsumerApp.xsjs
Once all files are created, you can import data from a source file, such as a CSV file, into the desired table
during a commit.
3. Using any code editor, open the schema definition (TISCHEMA.schema) file that you just created and enter
the name of the schema you want to apply.
schemaname=TISCHEMA;
PUBLI C

4. Open the table definition of the target import table (TiTable.hdbtable) file that you just created and enter
the following lines of text.
table.schemaName = "TISCHEMA";
table.columns =
[
{name = "Surname"; sqlType = VARCHAR; nullable = false; length = 40;
comment = "Person's surname";},
{name = "Forename"; sqlType = VARCHAR; nullable = false; length = 40;
comment = "Person's forename";}
...
];
table.primaryKey.pkcolumns = ["Surname"];
5. Open the CSV file, for example, TiCsv.csv in a text editor and enter some values, for example, the following
lines.
Meyer,Hugo
Schmitt,Michael
Foo,Bar
6. Open the table-import model (TiModel.hdbtim) file that you just created and enter the following lines of
text.
7. Open the data definition file (TiData.hdbtid) file that you just created in the same package and enter the
following lines of text.
implements TiPackage:TiModel.hdbtim;
8. Deploy the table import.
a) Select the package that you created in the first step.
b) Click the alternate mouse button and choose Commit.
c) Click the alternate mouse button and choose Activate.
You have imported data into the SAP HANA table using the table import feature.
6.12 Using Imported Table Data in SAP HANA
Before you start, ensure that you have deployed a table import.
Note
If you activate the files and then log on to the system using the SAP HANA studio, you do not have sufficient
privileges to execute a SELECT statement on the table that was created. To be able to execute the select
104
PUBLI C

statement on the table that was created, execute the following statement using the SQL Console in the SAP
HANA studio:
call
_SYS_REPO.GRANT_PRIVILEGE_ON_ACTIVATED_CONTENT('select','"<Path.to.><TiPackageName
>::<TiTableName>"','<youruser>');
1. In the SAP HANA studio, write a Select statement to view the imported data.
For example:
SELECT * FROM "<Path.to.>TiPackage::TiTable";
2. Choose Execute.
You can view the contents of the comma-separated values (CSV) file in the SAP HANA table.
6.13 Using Imported Table Data in an SAP HANA XS
Application
Before you start, ensure that you have deployed a table import.
Note
If you activate the files and then log on to the system using the SAP HANA studio, you do not have sufficient
privileges to execute a SELECT statement on the table that was created. You must execute the following
statement using the SQL Console in the SAP HANA studio:
call
_SYS_REPO.GRANT_PRIVILEGE_ON_ACTIVATED_CONTENT('select','"<Path.to.>TiPackage::TiT
able"','<youruser>');
The example in the note above is for illustration purposes only. In the example shown, TiPackage is the path to
the table-import package in the repository, where package names are separated by a dot (.); TiTable is the name
of the table that is the target for the data-import operation; and <youruser> is the name of a valid SAP HANA
database user.
To use imported table in an SAP HANA XS application, perform the following steps:
1. Create an application-descriptor file and place it in your table-import package, for example, TiPackage.
2. Create an application-access file and place it in the package to which you want to grant access, for example, in
same TiPackage package as the application descriptor file you created in the previous step.
The application-access file does not have a name; it only has the file extension .xsaccess.
3. In the application-access (.xsaccess) file, enter the following lines of text:
{
"auth_required" : true
}
PUBLI C

4. Create an application-logic file using server-side JavaScript, for example, TiConsumerApp.xsjs and place it
in your table-import package, for example TiPackage.
The application-logic file is the SAP HANA XS JavaScript application you use to access the data imported into
the table.
5. Enter the following lines of text into your application-logic file, for example, TiConsumerApp.xsjs.
Note
The following example is for illustration purposes only. In the example shown, <Path.to>.TiPackage is
the absolute path to the table-import package in the repository and TiTable is the name of the table that
is the target for the data-import operation.
$.response.contentType = "text/plain";
var pstatement1 = conn.prepareStatement( "SELECT * FROM
\"<Path.to.>TiPackage::TiTable\" WHERE \"Surname\"='Meyer'");
var result = pstatement1.executeQuery();
if (result)
{
if (result.next())
{
$.response.setBody('The sql statement execution was successful.');
}
else
{
$.response.setBody("No data found");
}
$.response.status = $.net.http.OK;
}
else
{
}
result.close();
pstatement1.close();
conn.close();
6. Select the package that you created in the first step.
7. Click the alternate mouse button and choose Commit.
8. Click the alternate mouse button and choose Activate.
9. Access the SAP HANA XS application with the following URL http://<dbhost>:
80<DB_Instance_Number>/<package_name>/<application_logic_file_name>.xsjs.
You receive the message, The SQL statement execution was sucessful.
6.14 Extending a Table Import
Extend an existing table import file by creating a new table-import configuration file that you use to specify how to
modify the relation to an existing table-import file.
Before you start, ensure that you have deployed a table import; the following files are required for a table import
extension scenario.
106
PUBLI C

The table-import data file, for example, TiData.hdbtid
Links the new comma-separated-values CSV data files to the data fields and, in turn, the target tables via the
model
The table-import extension file, for example, TiDataExtend.hdbtid
Extends the table-import data file TiData.hdbtid
A table-import model, for example, TiModel.hdbtim
Defines the data fields and import target tables of each data field
A table import package, for example, TiPackage
Contains all the artifacts you need to complete the table-import operation
A CSV file, for example, TiCsv.csv
Contains the data to be imported into the SAP HANA table
A second CSV file, for example, TiCsvExtend.csv
Contains the new (extended) data to be imported into the table
A table definition, for example, TiTable.hdbtable
Defines the structure of the target import table
The schema definition, for example, TISCHEMA.schema
Specifies the name of the schema in which the target import table is created
An application-logic, for example, TiConsumerApp.xsjs
1. Open the repository package that contains the table-import data file (for example, TiData.hdbtid) that you
want to extend.
2. Create a CSV file and give the new CSV file a name, for example, TiCsvExtend.csv.
This is the CSV file that contains the new, extended data to be imported into the table. The extended data
either changes or adds to the date specified in the original CSV file TiCsv.csv.
3. Open the new CSV file TiCsvExtend.csv in any code editor and enter the new, extended values you want to
import separated by a comma.
For example, enter the name NewFoo,Bar.
NewFoo,Bar
Note
The value NewFoo in theTiCsvExtend.csv modifies the original Foo value specified in the first CSV file
TiCsv.csv.
4. Create a table-import data-extension file (for example, TiDataExtend.hdbtid) and place the new data-
extension file in the same package (for example, TiPackage) as the other files.
5. Enter the following lines into the data-extension file TiDataExtend.hdbtid.
Note
The following example is for illustration purposes only. In the example shown, <Path.to>.TiPackage is
the absolute path to the table-import package in the repository and TiData.hdbtid is the name of the
data-extension file for the data-import operation.
modifies <Path.to.>TiPackage:TiData.hdbtid;
csvFiles = csvFiles + ["<Path.to.>TiPackage:TiCsvExtend.csv"];
PUBLI C

6. In the SAP HANA XS application <application_logic_file_name>.xsjs, modify the script to query the
new value(s) specified in the table-import extension configuration.
In this example, the table-import extension adds the name NewFoo to the data to be imported to SAP HANA;
you now want to query the value NewFoo.
Note
The following example is for illustration purposes only. In the example shown, <Path.to.>TiPackage is
the absolute path to the table-import package in the repository and TiTable is the name of the table that
is the target for the data-import operation.
var pstatement1 = conn.prepareStatement( "SELECT * FROM
\"<Path.to.>TiPackage::TiTable\" WHERE \"Surname\"='NewFoo");
var result = pstatement1.executeQuery();
if (result)
{
if (result.next())
{
$.response.setBody('The sql statement execution was successful.');
}
else
{
$.response.setBody("No data found");
}
}
else
{
}
result.close();
pstatement1.close();
conn.close();
7. Access the SAP HANA XS application with the following URL http://<dbhost>:
80<DB_Instance_Number>/<package_name>/<application_logic_file_name>.xsjs.
You receive the message, The SQL statement execution was successful.
108
PUBLI C

7 Setting Up the Analytic Model
Modeling refers to an activity of refining or slicing data in database tables by creating views to depict a business
scenario. The views can be used for reporting and decision-making.
The modeling process involves the simulation of entities, such as CUSTOMER, PRODUCT, and SALES, and
relationships between them. These related entities can be used in analytics applications such as SAP
BusinessObjects Explorer and Microsoft Office. In SAP HANA, these views are known as information views.
Information views use various combinations of content data (that is, non-metadata) to model a business use case.
Content data can be classified as follows:
Attribute: Descriptive data, such as customer ID, city, and country.
Measure: Quantifiable data, such as revenue, quantity sold and counters.
You can model entities in SAP HANA using the Modeler perspective, which includes graphical data modeling tools
that allow you to create and edit data models (content models) and stored procedures. With these tools, you can
also create analytic privileges that govern the access to the models, and decision tables to model related business
rules in a tabular format for decision automation.
You can create the following types of information views:
Attribute Views
Analytic Views
Calculation Views
Related Links
SAP HANA Studio [page 26]
The SAP HANA studio is an Eclipse-based development and administration tool for working with SAP HANA,
including creating projects, creating development objects, and deploying them to SAP HANA. As a developer, you
may want to also perform some administrative tasks, such as configuring and monitoring the system.
7.1 Setting Up the Modeling Environment
7.1.1 Setting Modeler Preferences
You can set preferences for your Eclipse workspace to set up the default settings that the system uses whenever
you log on.
1. Choose Window Preferences Modeler
Note
You can also set the preferences choosing Quick Launch Manage Preferences
2. Identify the required preference and perform the corresponding substeps from the table below:
Setting Up the Analytic Model
PUBLI C

Requirement Preference Substeps
To specify the structure of content
packages in the Navigator view
Content Presentation Hierarchical - to view the package
structure in a hierarchical manner
such that, the child folder is inside the
parent folder. Flat - to view all the
packages at the same level for
example, sap, sap.ecc, sap.ecc.ui.
Show Object Type Folders - to group
together similar objects in a package
such as attribute views in the Attribute
View package.
To set the preferences for loading data
using flat file
Data From Local File
1. Browse the location to save error
log files for data load using flat
files.
2. Enter the batch size for loading
data. For example, if you specify
2000 and a file has records of
10000 rows the data load will
happen in 5 batches.
3. Enter a decision maker count that
will be used to propose data types
based on the file. For example,
enter 200 if you want the
proposal to be made based on the
200 rows of file data.
To set the default value for the client
that will be used while previewing
model data
Default Model Parameters Select the client from the Default
Client drop-down list.
To enforce various rules on objects
Note
Enforcing validation rules with
severity Error are mandatory.
Validation Rules Select the required rules to be applied
while performing object validation.
To determine the numbers of rows to
be displayed in a page
Data Preview Select the maximum rows for data
preview as required.
To specify a location for job log files Logs
1. Expand the Logs node.
2. Select Job Log.
3. Browse the location where you
want to save the job log files.
To enable logging the repository calls
and specify location for repository log
files
Logs
1. Expand the Logs node.
2. Select Job Log.
3. Select true from the drop-down
list.
4. Browse the location where you
want to save the repository log
files.
To enable search for the attributes
used in the views
Search Options Select Enable Search Attributes. .
110
PUBLI C

Requirement Preference Substeps
To allow lower case alphabets for
attribute view, analytic view,
calculation view, procedure and
analytic privilege names
Case Restriction Deselect the Model name in upper
case checkbox.
Note
After changing the preferences make sure you choose Apply and OK.
7.1.2 Configuring the Import Server
You need to establish a connection with the server to load data from external sources to SAP HANA .
You have the details of the BusinessObjects Data Services repository and ODBC drivers.
1. In the Quick Launch tab page, choose Configure Import Server.
2. Enter the IP address of the server from which you want to import data.
3. Enter the repository name.
4. Enter the ODBC data source, and choose OK.
Once the connection is established you can import the tables definition and then load the data into table
definitions.
7.1.3 Importing Table Definitions
One of the ways to create tables to load source system data into SAP HANA is to import the table definitions from
the source system.
You have configured the SAP HANA modeler for importing metadata using the Data Services infrastructure.
Use this procedure to import table definitions (metadata) from a source system to load source system data into
SAP HANA. You can point to these table definitions for creating various content models such as attribute,
analytic, and calculation views.
Based on your requirements, use one of the following approaches:
Mass Import: To import all table definitions from a source system. For example, you can use this approach if
this is a first import from the given source system.
Selective Import: To import only selected table definitions from a source system. For example, you can use
this approach if there are only few table definitions added or modified in the source system after your last
import.
1. If you want to import all table definitions from a source system, do the following:
a) In the File menu, choose Import.
b) Expand the SAP HANA Content node.
c) Choose Mass Import of Metadata, and choose Next.
PUBLI C

d) Select the target system where you want to import all the table definitions, and choose Next.
e) In the Connection Details dialog, enter the operating system user name and password of the target
system.
f) Select the required source system, and choose Finish.
Note
If the required system is not available from the dropdown list, you need to contact your administrator.
2. If you only want to import selective table definitions from a source system, do the following:
a) In the File menu, choose Import.
b) Expand the SAP HANA Content node.
c) Choose Selective Import of Metadata, and choose Next.
d) Select the target system where you want to import the table definitions, and choose Next.
e) Select the required source system.
Note
If the required system is not available from the dropdown list, you need to add the new source system
using Manage Connections.
f) In the Type of Objects to Import field, select the required type, and choose Next.
g) Add the required objects (tables or extractors) that you want to import.
Note
If you want to add dependent tables of a selected table, select the required table in the Target panel,
and choose Add Dependent Tables in the context menu.
h) Select the schema into which you want to import the metadata.
i) If you selected object type as extractor, select the package into which you want to place the
corresponding objects.
j) Choose Next, then review and confirm the import by choosing Finish.
If the source object is a table or non V-type extractor, the system creates physical tables and stores them in
the selected schema. However, if the source object is a V-Type extractor, the system creates content models
and stores these models in the selected package, and the underlying physical tables in the schema.
7.1.3.1 Configuring Data Services for Metadata Import
Use this procedure to enable the SAP HANA modeler to import table definitions from the source system using the
Data Services infrastructure.
1. Set the Passphrase
a) Log on to the Central Management Console of SAP BusinessObjects Enterprise (BOE).
b) Choose Manage Applications Data Services Application Settings .
c) In the Encryption Passphrase field, enter the passphrase that you have been using for the SAP HANA
studio, and choose Save.
d) Restart the TOMCAT and BOE services.
112
PUBLI C

2. Disable Session Security
a) Log on to the Data Services Management Console.
b) Choose Administrator.
c) In the Navigator view panel, choose Web Services.
d) Choose the Web Services Configuration tab page.
e) Select the Import_Repo_Object checkbox to save the connection details.
f) Select Disable Session Security from the dropdown menu, and choose Apply.
3. Creating a Data Source
a) Go to Start Control Panel Administrative Tools .
b) Choose Data Sources (ODBC).
c) Choose the System DSN tab page, and choose Add.
d) Select HDBODBC from the driver list, and choose Finish.
e) Enter a name and description for the data source.
f) Enter server details.
Note
The format in which you need to enter details is <<host>:3<instance number>15>. For example,
abc2012.wdf.sap.corp:30115.
g) Enter the required database details in the format, <<SID><instance number>>. For example, M4701.
h) Choose Connect.
7.1.3.2 Managing Source System Connections
Use this procedure to add or manage a source system connection that is required to import table definitions.
7.1.4 Loading Data into Tables
If you are using the Load Controller or Sybase Replication Server infrastructure, make sure that you have
imported all table definitions into the SAP HANA database. For more information, see Importing Table
Definitions.
If you are using the SLT component, the source system(s), target system and the target schema, are
configured by the administrator during the installation.
Use this procedure to load data into your table definitions. Depending on your requirements, you can perform the
following:
Initial Load - to load all data from a source SAP ERP system into the SAP HANA database by using Load
Controller or SAP Landscape Transformation (SLT). This is mostly applicable when you are loading data from
the source for the first time.
Data Replication - to keep the data of selected tables in the SAP HANA database up-to date with the source
system tables by using SyBase Replication Server or SAP Landscape Transformation (SLT).
PUBLI C

1. In the Quick Launch tab page, choose Data Provisioning.
2. If you are using SLT-based replication, choose Source.
3. Choose Load (for initial load) or Replicate (for data replication) as appropriate.
4. Select the required tables to load or replicate data in any of the following ways:
Search for the required tables.
1. Select the table from the list, and choose Add.
2. If you want to save the selected list of tables locally for future reference, select the Export selected
tables checkbox, and specify the target location.
Load the list of tables from a local file as follows:
1. Choose Load from file.
2. Select the file that contains the required list of tables.
Note
The supported file type is .csv.
5. If you are using the load controller infrastructure, choose Next and enter the operating system user name and
password.
6. Choose Finish.
7.1.4.1 Suspending and Resuming Data Load
When loading data into tables using SLT- based replication, you can choose to stop data replication temporarily
for a selected list of tables, and later resume data load for these.
1. In the Quick Launch tab page, choose Data Provisioning.
2. Select the source system for which you want to suspend or resume data load.
3. Choose Suspend or Resume .
4. Select the tables, and choose Add.
5. Choose Finish.
7.1.4.2 Uploading Data from Flat Files
You can upload data from flat files in a client file system to the SAP HANA database.
If the table schema corresponding to the file to be uploaded already exists in the SAP HANA database, the
new data records are appended to the existing table.
If the required table for loading the data does not exist in the SAP HANA database, create a table structure
based on the flat file.
The application suggests the column names and data types for the new tables, and allows you to edit them. There
is a 1:1 mapping between the file and table columns in the new table. The application does not allow you to
overwrite any columns or change the data type of existing data. The supported file types are: .csv, .xls, and .xlsx.
114
PUBLI C

1. In the File menu, choose Import.
2. In the Select an import source section, expand the SAP HANA content node.
3. Select Data from Local File, and choose Next.
4. In the Target System section, select the target system to which you want to import the data using the flat file,
and choose Next.
5. In the Define Import Properties page, browse and select the file containing the data you want to load.
a) If you have selected a CSV file, select a delimiter.
Note
A delimiter is used to determine columns and pick the correct data from them. In a csv file, the
accepted delimiters are ',', ';' and ':'.
b) If you have selected an .xls or .xlsx file, select a worksheet.
6. Select the New option if you want to load the data into a new table.
a) ChooseNext.
b) On the Manage Table Definition and Data Mapping screen, map the source and target columns.
Note
Only 1:1 column mapping is supported. You can also edit the table definition by changing the data
types, renaming columns, adding or deleting the columns, and so on.
You can choose to map the source and target columns using the Auto Map option. If you choose
the one to one option, then the first column from the source is mapped to the first column of the
target. If you choose the Map by name option, the source and target columns with the same name
are mapped.
7. Select the Existing option if you want to append the data to an existing table.
a) Choose Next.
b) On the Manage Table Definition and Data Mapping screen, map the source and target columns.
8. Perform the following steps if you want to provide a constant value for a column at the target:
a) Right-click the column. From the context menu, choose Make As Constant.
b) In the Constant dialog box, enter a value, and choose OK.
Note
You can set a column to constant if it is not mapped to a source column.
9. Enter a value in the Default Value column to provide a default value for a column at the target. ChooseFinish.
7.1.5 Copying Content Delivered by SAP
You can copy the standard content shipped by SAP or an SAP partner to your local package for your modeling and
reporting use cases; for example, from sap.ecc.fin to customer.ecc.fin.
You have the following privileges:
REPO.READ for the source package.
PUBLI C

REPO.MAINTAIN_NATIVE_PACKAGES and REPO.EDIT_NATIVE_OBJECTS for the target root package.
Tip
We recommend you copy the content shipped by SAP or an SAP partner to your local package to avoid
overwriting your changes during the subsequent import.
Restriction
For script-based calculation views and procedures, even if you copy the dependent objects, you need to change
the script manually and adjust the dependent object references.
1. In the Quick Launch tab page, choose Mass Copy.
2. To create a mapping between the source package and the target package, choose Add.
3. Select a source package and a target package, and choose Next.
Tip
If you want to create more package mapping, select the source and target packages as required. To copy
the objects, you need to map the source root packages to the target root packages.
4. Select the required objects, and choose Add.
Caution
If you copy an object, but not its dependent objects, the copied object in the target package will have
references to the dependent objects in the source package.
5. Choose Next to view the summary and choose Finish to confirm content copy.
Tip
You can deselect an object to avoid copying it to the target package.
You need to activate the copied objects in the target package to consume them for reporting purposes.
7.1.6 Mapping the Authoring Schema to the Physical Schema
Schema mapping is done when the physical schema in the target system is not the same as the physical schema
in the source system, usually in a transport scenario. The schema mapping is maintained via a mapping table.
Youcan map the authoring schemas to the physical database schemas in the target system to access and deploy
transported objects.
A physical schema is a schema in which the tables are available. It may differ in the source and target systems.
An authoring schema (logical schema) is the database schema used in the design time views. For example, this is
the database schema that is used in the content development system.
Content object definitions are stored in the repository and contain references to the physical database schemas.
When you copy the content objects to a different system, for example, from an SAP system to a customer system
116
PUBLI C

or between customer systems, the object definition still refers to the physical database schemas at the source. To
resolve this, you use schema mapping.
Remember
Schema mapping only applies to references from repository objects to catalog objects. It is not intended to be
used for repository to repository references.
Restriction
You need to map the references of the script-based calculation views and procedures manually, that is, by
changing the script if the tables are qualified with the schema. If the tables are not qualified, the default schema
of the view is used, and the schema mapping is also applied to the default schema.
You can map several authoring schemas to the same physical schema. For example, content objects delivered by
SAP refer to different authoring schemas, whereas in the customer system, all these authoring schemas are
mapped to a single physical schema where the tables are replicated.
Remember
The mapping between authoring and physical schemas is stored in the configuration table
_SYS_BI.M_SCHEMA_MAPPING
Note
If a system does not have schema mapping, the authoring schema is filled 1:1 from the physical schema;
otherwise, the default schema cannot be changed.
1. In the Quick Launch tab page, choose Schema Mapping.
2. Choose Add.
3. Enter the authoring schema and physical schema that need to be mapped.Choose OK.
Example
PUBLI C

Consider a source system, SAP, with an information object, AV1. This refers to the table MARA in the SAP_ERP
physical database schema. There is a target system, Customer, with the physical database schema EMEA.
After you import content, the object AV1 cannot be activated in the Customer system, because it still refers to
the SAP_ERP schema.
AV1 in the Customer system currently refers to SAP_ERP. To be able to activate the object, you need to modify
the mapping of AV1 in the Customer system.
To do this, a mapping is created in the target system Customer between the authoring and physical schema as
follows:
Authoring Schema Physical Schema
SAP_ERP EMEA
7.1.7 Generating Time Data
For modeling a business scenario that requires time dimension, you can populate time data in the default time-
related tables present in _SYS_BI schema, for example while creating time attribute views. You can choose to
generate the time data for a given time span based on your requirements such as, calendar type and granularity.
If you model a time attribute view without generating time data andyou use the data preview, then an empty view
is shown. The time range for which you can generate the time data for the selected granularity is mentioned in the
table below:
Granularity Range
Seconds <= 5 years
Minutes <= 15 years
Hour <= 30 years
Day <= 50 years
Week <= 50 years
Month <= 50 years
Year <= 50 years
Note
If you populate the data before modeling time attribute views, you can use the Value Help from the respective
time tables.
1. In the Quick Launch tab page, choose Generate Time Data.
2. If your financial year is same as he calendar year,for example, January to December, chooseGregorianin the
Calendar Type dropdown .
a) Enter the period for which you want to generate time data.
b) Select the required granularity
Note
118
PUBLI C

If the selected granularity is Week or Month, specify the First day of the week.
c) Choose Generate.
The system populates the generated time data in M_TIME_DIMENSION_YEAR,
M_TIME_DIMENSION_MONTH, M_TIME_DIMENSION_WEEK, M_TIME_DIMENSION tables in _SYS_BI
schema.
3. If your financial year is not the same asthecalendar year, for example, March to April, select Fiscalin the
Calendar Type dropdown .
a) Enter the period for which you want to generate the time data.
b) Select the Variant Schema where the tables that contain variant data are maintained.
Note
Tables T009 and T009B contain variant information.
c) Select the required variant that specifies the number of periods, along with their start and end dates,
according to your use case.
d) Choose Generate.
The system populates the generated time data in the M_FISCAL_CALENDAR table, which is located in the
_SYS_BI schema.
7.2 Creating Views
7.2.1 Attributes and Measures
Attributes and measures form the content data that is used for modeling. While attributes represent the
descriptive data like city and country, measures represent quantifiable data like revenue and quantity sold.
Attributes
Attributes are the individual non-measurable analytical elements.
Simple Attributes
Simple attributes are individual non-measurable analytical elements that are derived from the data
foundation.
For example, PRODUCT_ID and PRODUCT_NAME are attributes of a PRODUCT subject area.
Calculated Attributes
Calculated attributes are derived from one or more existing attributes or constants.
For example, deriving the full name of a customer (first and last name), assigning a constant value to an
attribute that can be used for arithmetic calculations.
Local Attributes
Local attributes used in an analytic view allow you to customize the behavior of an attribute for only that view.
PUBLI C

For example, if an analytic view or a calculation view include an attribute view, it inherits the behavior of the
attributes from the attribute view. If you set the parameter once, it is replicated in all views that are
consuming it.
Note
If you create an analytic view for one specific use case in which you want a particular attribute to behave
differently than it does in the attribute view to which it belongs, you can define it as a local attribute.
Measures
Measures are measurable analytical elements. Measures are derived from the analytic and calculation views.
Simple Measures
A simple measure is a measurable analytical element that is derived from the data foundation.
For example, PROFIT.
Calculated Measure
Calculated measures are defined based on a combination of data from OLAP cubes, arithmetic operators,
constants, and functions.
For example, calculated measures can be used to calculate the total sales of a product across five regions, or
to assign a constant value to a measure for a calculation.
Restricted Measure
Restricted measures are used to filter the value based on the user-defined rules for the attribute values.
Counters
Counters add a new measure to the calculation view definition to count the recurrence of an attribute. For
example, to count how many times product appears.
Note
You can choose to hide the attributes and measures that are not required for client consumption. For example,
for a complex calculation that is derived from a series of computations, you can hide the levels of computations
that are not required for reporting purposes.
7.2.2 Attribute Views
Attribute views are used to model an entity based on the relationships between attribute data contained in
multiple source tables.
For example, customer ID is the attribute data that describes measures (that is, who purchased a product).
However, customer ID has much more depth to it when joined with other attribute data that further describes the
customer (customer address, customer relationship, customer status, customer hierarchy, and so on).
You create an attribute view to locate the attribute data and to define the relationships between the various tables
to model how customer attribute data, for example, will be used to address business needs.
You can model the following elements within an attribute view:
120
PUBLI C

Columns
Calculated Columns
Note
In the Semantics node, you can classify the columns as attributes and build calculated columns of attribute
type.
Hierarchies
Note
For more information about the attributes and hierarchies mentioned above, see sections Attributes and
Measures, and Hierarchies.
You can choose to further fine-tune the behavior of the attributes of an attribute view by setting the properties as
follows:
Filters to restrict values that are selected when using the attribute view.
Attributes can be defined as Hidden so that they can be used in processes but are not visible to end users.
Attributes can be defined as key attributes and used when joining multiple tables.
The Drill Down Enabled property can be used to indicate if an attribute is available for further drill down when
consumed.
Attribute views can later be joined to tables that contain measures within the definition of an analytic view or a
calculation view to create virtual star schema on the SAP HANA data.
7.2.3 Creating Attribute Views
Attribute views are used to define joins between tables, and to select a subset (or all) of the table's columns and
rows. The rows selected can also be restricted by filters. One application of attribute views is to join multiple
tables together when using star schemas, to create a single dimension table. The resultant dimension attribute
view can then be joined to a fact table via an analytic view to provide meaning to its data. In this use case, the
attribute view adds more columns and also hierarchies as further analysis criteria to the analytic view. In the star
schema of the analytic view, the attribute view is shown as a single dimension table (although it might join multiple
tables), that can be joined to a fact table. For example, attribute views can be used to join employees to
organizational units which could then be joined to a sales transaction via an analytic view
You have imported T009 and T009B tables for creating attribute view of type Time.
Use this procedure to create a view that is used to model descriptive attribute data (that does not contain
measures) using attributes.
Tip
You need this view for creating a multidimensional view.
1. Set Parameter
1. In the Modeler perspective, expand the Content node of the required system.
PUBLI C

2. Expand the package to which you want to save your information object.
3. In the context menu of Attribute Views node, choose New .
4. Enter a name and description for the view.
5. To create data foundation for the view, perform substeps of the required scenario given in the table
below:
Scenario Substeps
Create a view with table
attributes.
In the Sub Type dropdown list, choose Standard.
Create a view with time
characteristics.
1. In the Sub Type dropdown list, choose Time.
2. Select the required calendar type as follows:
a. If the calendar type is Fiscal, select a variant schema, and a fiscal
variant.
b. If the calendar type is Gregorian, select the granularity for the
data.
3. To use the system-generated time attribute view, select Auto Create.
Note
The system creates a time attribute view based on the default time
tables, and defines the appropriate columns/attributes based on
the granularity. It also creates the required filters.
Note
The tables used for time attribute creation with calendar type Gregorian
are, M_TIME_DIMENSION, M_TIME_DIMENSION_ YEAR,
M_TIME_DIMENSION_ MONTH, M_TIME_DIMENSION_WEEK and for
calendar type Fiscal is M_FISCAL_CALENDAR. If you want to do a data
preview for the created attribute view, you need to generate time data
into the mentioned tables from the Quick Launch tab page.
Copy a view from an
existing view in this case,
you can modify the copied
view.
1. Choose Copy From.
2. Select the required attribute view.
Derive a view from an
existing view in this case,
you cannot modify the
derived view that acts as a
reference to the base
attribute view.
1. In the Sub Type dropdown, choose Derived.
2. Select the required attribute view.
6. Choose Finish.
The attribute view editor opens. The Scenario panel of the editor consist of two nodes - Data Foundation and
Semantics. The Data Foundation node represents the tables used for defining the output structure of the view.
122
PUBLI C

The Semantics node represents the output structure of the view, that is, the dimension. In the Details panel
you define the relationship between data sources and output elements.
2. Define Output Structure
a) Add the tables that you want to use in any of the following ways:
Drag the required tables present in the Catalog to the Data Foundation node.
Select the Data Foundation node in the Scenario panel, and in the context menu of the Details panel,
choose Add.
Note
You can choose to add the same table again in Data Foundation using table aliases in the editor. For
example, in cases where you want to have different cardinalities from the same table.
Restriction
It is not allowed to add column views to the Data Foundation.
b) If you want to query data from more than one table, in the Details panel context menu, choose Create
Join, and enter the required details.
Note
After creating the join, you can edit its properties like join type, cardinality, etc in the Properties
view. You can choose to create Text Join between table fields in order to get language-specific
data. For example, consider that you have a product table that contains product IDs and no
description about products, and you have a text table for products that has a language-specific
description for each product. You can create a text join between the two tables to get language-
specific details. In a text join, the right table should be the text table and it is mandatory to specify
the Language Column.
You can set the cardinality of the join as required. By default the cardinality of the join is "empty".
The empty cardinality can be specified in cases when you are not sure about the right cardinality. If
you specify the empty cardinatily, system during the join evaluation phase diagnose the best
suited cardinality and execute the join. Wrong selection of cardinality can lead to erroneous data
and out of memory issues.
c) Add the table columns to the output structure that is, the Semantics node that you want to use to define
attribute data. You can define the attribute data by doing one of the following:
Select the toggle button on the left of the table field.
Right-click the table field, and choose Add to Output.
d) If you want to specify a filter condition based on which system must display data for a table field in the
output do the following:
1. Right-click the table field, and choose Apply Filter.
2. Select the required operator, and enter filter values.
All the table fields that you have added to the output are automatically mapped as attributes.
3. Define Key Attributes
a) Select the Semantics node.
b) In the Attributes tab page of the Column panel, select the required attribute and select the Type as Key
Attribute.
PUBLI C

Remember
If there is more than one key attribute, all key attributes of the attribute view must point to the same
table in the data foundation. The central table of the attribute view is the one to which all the key
attributes point.
Note
In case of auto-generated time attribute views, the attributes and key attributes are automatically
assigned.
4. Optional Step: Create Calculated Columns
a) In the Output of Data Foundation panel, right-click Calculated Columns.
b) In the context menu, choose New.
c) Enter a name and description (label) for the calculated column.
d) Select a data type for the calculated column.
e) Enter length and scale for the calculated coulmn.
f) In the Expression Editor enter the expression. For example, you can write a formula such as,
if("PRODUCT" = 'ABC', "DISCOUNT" * 0.10, "DISCOUNT"). This means if attribute PRODUCT equals the
string ABC then DISCOUNT equals to DISCOUNT multiplied by 0.10 should be returned. Otherwise the
original value of attribute DISCOUNT should be used.
Note
The expression can also be assembled by dragging and dropping the expression elements from the
menus below the editor window.
g) Choose OK.
5. Optional Step: To filter and view the table data in the modeled view, which is relevant to a specific client as
specified in the table fields, such as, MANDT or CLIENT, at runtime perform the following:
1. Select the Semantics node, in the Properties panel edit the Default Client property.
Note
The default value for the property is the one that is specified as a preference. At runtime, if the property is
set to Dynamic then, the value set for the Session Client property is used to filter table data. The Session
Client property is set while creating a user.
Note
You can choose to create hierarchies in order to define relationships between attributes.
Note
You can choose to associate an attribute with another attribute, which describes it in detail. For example,
when reporting via Label Mapping (also known as Description Mapping), you can associate Region_ID
with Region_Text.
Before SP05, you could associate an attribute with another attribute in a model. In the runtime object an
<attribute>.description column is generated and is shown during data preview. Now, from SP05 onwards
the behavior is as follows:
124
PUBLI C

For an attribute (CUSTOMER) you can now maintain label mapping by selecting another attribute
(TEXT) from the same model as "Label Column" in the Semantics node. The result is "TEXT"
displaying as the label column in data preview. Note that the CUSTOMER.description column is not
generated and is not shown in data preview anymore. You can choose to rename a label column as
<attribute>.description. For example, if A1 has a label column B1, then you can rename B1 to
A1.description but not as B1.description. Once you rename a label column with .description as suffix,
the related columns appear side by side during data preview.
If you have created an object using the old editor (which supported the old style of description
mapping) and try to open it using the new editor you will see a new column CUSTOMER.description
(as an attribute) which is hidden and disabled because this column cannot be used in other places
such as parameter/variable, calculated column, restricted column and so on. You cannot maintain
properties for this attribute for example, description. CUSTOMER.description displays in the data
preview as long as you do not change it in the editor. You can change its name. After changing the
name you can maintain its properties and use it like other attributes.
You can rename the label column as <attribute>.description. For example, if A1 has a label column
A1.description, then you can rename A1.description to B1 and again as A1.description but not as
B1.description. Once you rename a label column with .description as suffix, the related columns
appear side by side during data preview.

6. Activate the view using one of the following options in the toolbar:
Save and Activate - to activate the current view and redeploy the affected objects if an active version of
the affected object exists. Otherwise only current view gets activated.
Save and Activate All - to activate the current view along with the required and affected objects.
Note
You can also activate the current view by selecting the view in the Navigator view and choosing Activate in
the context menu.
You can find the activated model in the related package. If you want to modify this model, from the context menu,
choose Open and make the necessary changes.
Restriction
The behavior of attribute views with the new editor is as follows:
When an object (a table of an attribute view) is removed and added again in an attribute view in order to
reflect the recently modified columns with its data type, it reflects the previous state of the columns alone.
For more information, see SAP Note 1783668.
When you open an attribute view and there is a missing column in the required object, an error is shown
and the editor does not open. For information regarding the solution of this issue, see SAP Note 1788552.
Related Links
Creating Hierarchies [page 148]
Activating Objects [page 153]
Generating Time Data [page 118]
For modeling a business scenario that requires time dimension, you can populate time data in the default time-
related tables present in _SYS_BI schema, for example while creating time attribute views. You can choose to
generate the time data for a given time span based on your requirements such as, calendar type and granularity.
PUBLI C

7.2.4 Analytic Views
Analytic views are used to model data that includes measures.
For example, an operational data mart representing sales order history would include measures for quantity,
price, and so on.
The data foundation of an analytic view can contain multiple tables. However, measures that are selected for
inclusion in an analytic view must originate from only one of these tables (for business requirements that include
measure sourced from multiple source tables, see calculation view ).
Analytic views can be simply a combination of tables that contain both attribute data and measure data. For
example, a report requiring the following:
<Customer_ID Order_Number Product_ID Quantity_Ordered Quantity_Shipped>
Optionally, attribute views can also be included in the analytic view definition. In this way, you can achieve
additional depth of attribute data. The analytic view inherits the definitions of any attribute views that are included
in the definition. For example:
<Customer_ID/Customer_Name Order_Number Product_ID/Product_Name/Product_Hierarchy
Quantity_Ordered Quantity_Shipped>
You can model the following elements within an analytic view:
Columns
Calculated Columns
Restricted Columns
Remember
In the Semantics node, you can classify columns and calculated columns as type attributes and measures.
The attributes you define in an analytic view are Local to that view. However, attributes coming from
attribute views in an analytic view are Shared attributes. For more information about the attributes and
measures mentioned above, see section Attributes and Measures.
Variables
Input parameters
Note
For more information about the variables and input parameters mentioned above, see sections Assigning
Variables and Creating Input Parameters.
You can choose to further fine-tune the behavior of the attributes and measures of an analytic view by setting the
properties as follows:
Filters to restrict values that are selected when using the analytic view.
Attributes can be defined as Hidden so that they are able to be used in processes but are not viewable to end
users.
Attributes can be defined as key attribute and used when joining multiple tables.
The Drill Down Enabled property can be used to indicate if an attribute is available for further drill down when
consumed.
126
PUBLI C

Aggregation type on measures
Currency and Unit of Measure parameters (you can set the Measure Type property of a measure, and also in
Calculated Column creation dialog, associate a measure with currency and unit of measure)
Tip
If there is a name conflict that is, more than one element having the same name among the local and shared
attributes, calculated columns, restricted columns, and measures of an analytic view, the activation of the view
does not go through. You can resolve such conflict using the aliases. Aliases must also have unique names. You
can assign an alias to the required element in the Column view of the Semantics node by editing its name inline.
Hereinafter, the element is referred by its alias.
If two or more shared columns have a name conflict, during save the aliases for the conflicting name columns
are proposed. You can choose to overwrite the proposed names.
In case of old models, if you find any error while opening the object due to aliasing that was caused due to
swapping of column names with the alias names, use the Quick Fix. To use the Quick Fix, select the error
message that is, the problem in the Problems view, and choose Quick Fix in the context menu. This resolves the
swapping issue by assigning right names to the column and alias.
You can choose to hide the attributes and measures that are not required for client consumption by assigning
value true to the property Hidden in the Properties pane, or selecting the Hidden checkbox in the Column view.
The attributes or measures marked as hidden are not available for input parameters, variables, consumers or
higher level views that are build on top of the analytic view. For old models (before SPS06), if the hidden attribute
is already used, you can either unhide the element or remove the references.
For an analytic view, you can set the property MultiDimensional Reporting to true or false. If the MultiDimensional
Reporting property of the analytic view is set to false, the view will not be available for multidimensional reporting
purposes. If the value is set to true, an additional column Aggregation is available to specify the aggregation type
for measures.
You can enable relational optimization for your analytic view such as, Optimize stacked SQL for example, convert
SELECT a, SUM(X) FROM ( SELECT * FROM AV) GROUP BY A
to
SELECT A, SUM(X) FROM AV GROUP BY A
by setting the property Allow Relational Optimization.
Setting this property would be effective only for analytic views having complex calculations such that deployment
of analytic view generates catalog calculation view on top of the generated catalog OLAP view.
Caution
In this case, if this flag is set counters and SELECT COUNT may deliver wrong results
7.2.5 Creating Analytic Views
Analytic views are typically defined on a fact table that contains transactional data (as measures). Using analytic
views you can create a selection of measures, add attributes and join attribute views.
PUBLI C

Analytic views leverage the computing power of SAP HANA to calculate aggregate data, e. g. the number of sold
cars per country, or the maximum power consumption per day. They are defined on only one fact table, a table
which contains for example, one row per sold car or one row per power meter reading, or some form of business
transaction records. Fact tables can be joined to allow access to more detailed data using a single analytic view as
long as the facts (measures) come from a single table. Analytic views can be defined in a single table, or in joined
tables.
Analytic views can contain two types of columns, attributes and measures, for which measures are of type simple,
calculated, and restricted. Measures are attributes for which an aggregation must be defined. If analytic views are
used in SQL statements then the measures have to be aggregated e. g. using the SQL functions SUM(<column
name>), MIN(<column name>), or MAX(<column name>). Normal columns can be handled as regular attributes.
They do not need to be aggregated.
1. Set Parameters
a) In the Modeler perspective, expand the Content node of the required system.
b) Select the package in which you want to save your information object.
c) From the context menu of the package, choose New Analytic View .
Note
Once you have an analytic view in a package, you can also create another one from the context menu
of the Analytic View node in the respective package.
d) Enter a name and description for the view.
e) If you want to create a view based on the existing one, choose Copy From option, and select the required
view.
f) Choose Finish.
The analytic view editor opens. The Scenario panel of the editor consist of the following three nodes:
Data Foundation - represents the tables used for defining the fact table of the view.
Logical Join - represents the relationship between the selected table fields (fact table) and attribute views
that is, used to create the star schema.
Semantics - represents the output structure of the view.
In the Details panel you define the relationship between data sources and output elements.
2. Define Output Structure
a) Add the tables that you want to use in any of the following ways:
Drag the required tables present in the Catalog to the Data Foundation node.
Select the Data Foundation node in the Scenario panel, and in the context menu of the Details panel,
choose Add Tables.
Note
You can choose to add the same table again in Data Foundation using table aliases in the editor. For
example, if you want to have different cardinalities from the same table.
Restriction
It is not allowed to add column views to the Data Foundation of an analytic view. However, you can add
column views in a calculation view.
128
PUBLI C

Remember
If there is more than one table, you need to specify one of the tables as the central table (fact table)
from which the measures will be derived. You can specify the central table by selecting a value in the
Central Entity property of the Data Foundation node.
b) If you want to query data from more than one table, in the Details panel context menu, choose Create
Join, and enter the required details.
Tip
After specifying the central table or choosing one of the field as measure, when you save the view
automatic assignment of attributes and measures is done. All the numeric fields from the central table
are assigned as measures and others as attributes. In this case, you will not be able to join measures
with other numeric fields. Also, if you have specified all the attributes and measures of the view before
saving it, automatic assignment of attributes and measures will not overwrite them.
Note
After creating the join, you can edit its properties such as join type, cardinality, etc in the Properties
view. You can choose to create a Text Join between table fields in order to get language specific data.
For example, consider that you have a product table that contains product IDs but no product
description, and you have a text table for products that has language-specific description for each
product. You can create a text join between the two tables to get language-specific details. In a text
join, the right table should be the text table and it is mandatory to specify the Language Column.
c) Add the table columns to the output structure (Semantics node) that you want to use to define the facts
in any of the following ways:
Select the toggle button on the left of the table field.
Right-click the table field, and choose Add to Output.
d) If you want to specify a filter condition based on which system must display data for a table field in the
output, for example to display revenue for only selected companies based on the filter value, do the
following:
1. Right-click the table field, and choose Apply Filter.
2. Select the required operator, and enter filter values.
The table fields selected above form the fact table.
e) To create a star schema that is, linking the fact table with the descriptive data (attribute views), do the
following:
1. Add the required attribute views in the Logical Join node in any of the following ways:
Drag the required attribute views present in the Content node to the Logical Join node.
Select the Logical Join node in the Scenario panel, and choose Add button to add the attribute
views.
2. Create joins between the views and fact table.
Note
In the Logical Join, you can create a temporal join between the date field of the fact table to an
interval (to and from) field of the attribute view. The temporal join has to start from the fact table
such that the single column must be in the fact table and, the to and from columns must be in the
PUBLI C

table that is directly joined to the fact table. The join type must be a referential join. The supported
data types are timestamp, date, and integer.
Restriction
While creating joins you must ensure that a table does not appear twice in any join path (in particular a self
join is not supported). A join path is the set of joins that links the fact table to other tables.
While creating joins between analytic view and attribute view the same table cannot be used both in the join
path of analytic view and attribute view. Also, in a join path of analytic view and attribute view, the table of
the attribute view which is linked to the fact table should not have an alias table.
3. Optional Step: Create Calculated Column
a) In the Output of Logical Join panel, right-click Calculated Columns.
b) In the context menu, choose New.
c) Enter a name and description (label) for the calculated column.
d) Select a data type, and enter the length and scale for the calculated column.
e) Select the Column Type to determine whether it is a calculated attribute or a calculated measure.
f) If you select Calculate Before Aggregation, select the aggregation type.
Note
If you select Calculate Before Aggregation, the calculation happens as per the expression specified and
then the results are aggregated as SUM, MIN, MAX or COUNT. If Calculate Before Aggregation is not
selected, the data is not aggregated but it gets calculated as per calculation expression (formula), and
the aggregation is shown as FORMULA. If the aggregation is not set, then it will be considered as an
attribute.
g) In the Expression Editor enter the expression. For example, you can write a formula:
if("PRODUCT" = 'ABC, "DISCOUNT" * 0.10, "DISCOUNT") which is equivalent to, if attribute PRODUCT
equals the string ABC then DISCOUNT equals to DISCOUNT multiplied by 0.10 should be returned.
Otherwise the original value of attribute DISCOUNT should be used.
Note
The expression can also be assembled by dragging and dropping the expression elements from the
menus below the editor window.
h) If you want to associate the calculated column with currency and unit of measuring quantity, select the
Advanced tab page and select the required Type.
i) Choose OK.
Remember
Calculated Columns can be created only at the Logical Join level and not at the Data Foundation level.
Note
The aggregation type for calculated columns is only taken into account, if they are calculated before
aggregation
130
PUBLI C

4. Optional Step: Create Restricted Columns
You can create restricted columns if you want to filter the value for an output field based on the user-defined
rules. For example, you can choose to restrict the value for the Revenue column only for Region = APJ, and
Year = 2012.
a) In the Output panel of the Logical Join, right-click Restricted Columns, and choose New.
b) Enter a name and description for the restricted column.
c) From the Column dropdown list, select the column for which you want to apply a restriction.
Caution
The column for which you apply a restriction must be defined as measure in the semantics node
otherwise the validation will fail.
d) Choose Add Restriction.
e) In the Parameter field, select the column that you want to create a restriction for, then select the operator
and value.
f) Choose OK.
Note
For a restricted column the aggregation type of the base column is applied.
5. Define Attributes and Measures
a) Select the Semantics node.
b) In the Column panel, select the Local tab page, and change the type as attributes and measures.
Note
The Shared tab page shows the attributes of the used attribute views. While generating the column
views, the joined private attribute name is ignored and the shared attribute name is considered,
therefore the joined private attribute is not shown on the Semantics node.
Remember
If the MultiDimensional Reporting property of the analytic view is set to false, the view will not be
available for reporting purposes. If the value is set to true, an additional column Aggregation is available
to specify the aggregation type for measures.
Note
You can choose to associate an attribute with another attribute, which describes it in detail. For
example, when reporting via Label Mapping (also known as Description Mapping), you can associate
Region_ID with Region_Text.
Before SP05, you could associate an attribute with another attribute in a model. In the runtime object
an <attribute>.description column is generated and is shown during data preview. Now, from SP05
onwards the behavior is as follows:
For an attribute (CUSTOMER) you can now maintain label mapping by selecting another attribute
(TEXT) from the same model as "Label Column" in the Semantics node. The result is "TEXT"
displaying as the label column in data preview. Note that the CUSTOMER.description column is not
generated and is not shown in data preview anymore.
PUBLI C

You can choose to rename a label column as <Attribute>.description. For example, if A1 has a label
column B1, then you can rename B1 to A1.description but not as B1.description. Once you rename
a label column with .description as suffix, the related columns appear side by side during data
preview.
If you have created an object using the old editor (which supported the old style of description
mapping) and try to open it using the new editor you will see a new column
CUSTOMER.description (as an attribute) which is hidden and disabled because this column cannot
be used in other places such as parameter/variable, calculated column, restricted column and so
on. You cannot maintain properties for this attribute for example, description.
CUSTOMER.description displays in the data preview as long as you do not change it in the editor.
You can change its name. After changing the name you can maintain its properties and use it like
other attributes.
You can rename the label column as <Attribute>.description. For example, if A1 has a label column
A1.description, then you can rename A1.description to B1 and again as A1.description but not as
B1.description. Once you rename a label column with .description as suffix, the related columns
appear side by side during data preview.

Note
You can change the type of a measure and perform currency conversion by selecting it in the Local tab
page and changing the Measure Type property in the properties panel.
6. Optional Step: You can filter and view the table data in the modeled view for a specific client as specified in the
table fields, such as MANDT or CLIENT, by doing the following:
1. Select the Semantics node, in the Properties panel, edit the Default Client property.
Note
The default value for the property is the one that is specified as a preference. If the property is set to
Dynamic, at runtime the value set for the Session Client property is considered to filter table data. The
Session Client property is set while creating a user.
7. Optional Step: Assign Variable
You assign variables to a field at design time for obtaining data based on the values you provide for the
variable. At runtime, you can provide different values to the variable to view the corresponding set of attribute
data.
a) In the Semantics node Variables/Input Parameter panel, choose the Create Variable option.
b) Enter a name and description for the variable.
c) Select the required attribute from the dropdown list.
Note
At runtime, the value for the variable is fetched from the selected attribute's data.
d) Choose the required Selection Type from the dropdown list.
Note
Single Value - Used to filter and view data based on a single attribute value. For example, to view
the sales of a product where the month is equal to January.
132
PUBLI C

Interval - Used to filter and view a specific set of data. For example, to view the expenditure of a
company from March to April.
Range - Used to filter and view data based on the conditions that involve operators such as:
"="(equal to)
">" (greater than)
"<" (less than)
">=" (greater than or equal to)
"<=" (less than or equal to)
For example, to view the sales of all products in a month where the quantity sold is >= 100.
e) Select Multiple Entries if you want to provide different values at runtime to filter data. For example, to view
the revenue from a period of 2000 to 2005 and 2012.
f) If you want to assign the variable to attribute(s), in the Apply variable filter to panel, choose Add.
g) Select the attribute(s) from the dropdown list to which you want to assign this variable.
Note
You can also assign a variable to an attribute later in the Column panel of the Semantics node Details
panel. To assign a variable to an attribute, select a variable from the variable dropdown list in the
Variable column.
h) Choose OK.
8. If you want to parameterize currency conversion and calculated columns, create input parameters.
9. In the editor toolbar, choose Save and Activate. This saves and activates the view.
Note
If an active version of the affected objects exist, activating the current view redeploys the affected objects.
You can also activate an object from the object context menu in the Navigator view.
Tip
You can choose to activate the other objects (required or impacted objects) along with the currenct object
using the Save and Activate All option in the toolbar.
Note
Restriction
The behavior of analytic views with the new editor is as follows:
When an object (a table of or an attribute view) is removed and added again in an attribute view and
analytic view editor in order to reflect the recently modified columns with its data type, it reflects the
previous state of the columns . For more information about the problem and its solution, see SAP
Note 1783668.
When you open an analytic view and there is a missing column in the required object, an error is
shown and the editor does not open. For information regarding the solution of this issue, see SAP
Note 1788552.
PUBLI C

Related Links
Using Currency and Unit of Measure [page 150]
Creating Input Parameters [page 145]
Using Temporal Joins [page 134]
A temporal join indicates the time interval mapping between the master data and the transaction data for which
you want to fetch the records.
7.2.5.1 Using Temporal Joins
A temporal join indicates the time interval mapping between the master data and the transaction data for which
you want to fetch the records.
You create a temporal join using the temporal column that specifies the time interval with the start and the end
date. The result set is fetched based on the time interval mapped using the temporal column.
A record is only included in the results set if the time interval lies within the valid time interval. A time interval is
assigned to each record in the results set. The records are valid for the duration of the interval to which they are
assigned.
You use temporal conditions to indicate whether to include or exclude the value of the FROM and TO date fields
while executing the join condition.
In the logical join, you can create a temporal join between the date field of the fact table to an interval (to and
from) field of the attribute view. The temporal join must start from the fact table such that the single column must
be in the fact table and, the to and from columns must be in the table that is directly joined to the fact table. The
join type must be Referential. The supported data types are: timestamp, date, and integer.
1. Create a referential join between the attribute of the fact table and the attribute view.
2. In the Properties panel, select:
a) The Temporal Column that indicates the single time column in the fact table.
b) The From Column and the To Column to specify the time interval from the attribute view.
c) The Temporal Condition which would be considered while executing the join.
Example
Consider an attribute view Product that contains master data about Products with attributes like, ProductID,
Validity_Date_From, Validity_Date_To, and so on.
Similarly, consider an analytic view Sales that contains transactional data corresponding to the products sales
with attributes, ProductID, Date, Revenue.
Now, to analyze sales data for products you can create a join between the two views using ProductID.
But to fetch data for a particular time period you need to assign temporal properties to the join. The temporal
column in our example would be Date field in the Analytic view, and the From date and To date would come
from the Validity_Date_From, and Validity_Date_To of the attribute view.
134
PUBLI C

7.2.6 Calculation Views
A calculation view is used to define more advanced slices on the data in SAP HANA database. Calculation views
can be simple and mirror the functionality found in both attribute views and analytic views. However, they are
typically used when the business use case requires advanced logic that is not covered in the previous types of
information views.
For example, calculation views can have layers of calculation logic, can include measures sourced from multiple
source tables, can include advanced SQL logic, and so on. The data foundation of the calculation view can include
any combination of tables, column views, attribute views and analytic views. You can create joins, unions,
projections, and aggregation levels on the sources.
You can model the following elements within a calculation view:
Attributes
Measures
Calculated measures
Counters
Hierarchies (created outside of the attribute view)
Note
For more information about the attributes, measures, counters, and hierarchies mentioned above, see
sections Attributes and Measures, and Hierarchies.
Variables
Input parameters
Note
For more information about the variables and input parameters mentioned above, see sections
AssigningVariables and Creating Input Parameters.
Calculation views can include measures and be used for multi-dimensional reporting or can contain no measures
and used for list-type of reporting. Calculation views can either be created using a graphical editor or using a SQL
Console . These various options provide maximum flexibility for the most complex and comprehensive business
requirements.
7.2.7 Creating Calculation Views
Calculation views are used to provide composites of other views. Essentially they are based on a join or union of
two or more data flows or on invoke of built-in or generic SQL functions.
Calculation views are defined as either graphical views or scripted views but not as SQLScript (for exceptions, see
below) depending on how they are created. They can be used in the same way as analytic views, however, in
contrast to analytic views it is possible to join several fact tables in a calculation view.
Graphical views can be modeled using the graphical modeling features of the SAP HANA Modeler. Scripted views
are created as sequences of SQL statements.
PUBLI C

As mentioned previously calculation views are generally not created using SQLScript. There are however,
exceptions to this rule. SQLScripts with the following properties can be used in calculation views:
- No input parameters
- Always read-only (that is, do not make changes to the database)
- Side-effect free
You can use calculation views to derive values and key performance indicators(KPIs).
Note
The terms "attribute" and "columns" are used interchangeably in this procedure. They may denote a table
column, a particular data field of a table row, or the contents of such a data field. The respective meaning
should be clear from the context.
Procedure
1. In the Modeler perspective, expand the system node from the Navigator view.
2. Expand the Content node.
3. Right-click the required package.
4. From the context menu, choose New Calculation View .
a. Enter a name and description.
b. Select the required package.
Create a Script-Based Calculation View
1. Choose SQL Script.
2. Select the required schema from the Default Schema dropdown list, for unqualified access in SQL.
Note
If you do not select a default schema, while scripting, you need to provide fully qualified names of the
objects used.
3. Select the required option from the Run With dropdown list, which identifies whose rights are to be considered
while executing the calculation scenario.
Note
Definer's right: If you want system to use the rights of the definer while executing the view or procedure for
any user.
Invoker's right: If you want system to use the rights of the current user while executing the view or
procedure .
136
PUBLI C

4. Choose Finish.
5. Define the Output Structure
a. Choose Define Output Parameter from the Output pane.
b. To add the output parameters with the required data type and length, from the Output pane toolbar,
choose Define Output Parameter button.
Note
The order and data types of the output parameters should match the table columns order and data
type which is used in the select query.
c. Choose OK.
6. Define the Function
a. Define the function using SQL Script commands.
Note
You can create variables, and bind them to attributes for filtering data. The values you provide for
the variables at runtime determine which data records are selected for consumption.
You can create input parameters for the calculation view that works as a placeholder in a query.
You can create a hierarchy between attributes of the view.
7. Save and Activate
a. To save the view, choose File Save .
b. From the context menu of the calculation view, choose Activate.
Create a Graphical Calculation View
1. In the Subtype dropdown list, choose Graphical.
2. Select the required schema from the Schema for Conversion dropdown list.
Note
The schema selected for conversion is used during the currency conversion. It list down all the schemas
which has currency related tables and the same can be changed during design time from the properties.
3. Choose Next.
4. Follow the instructions in the wizard to select the required tables and content models.
Adding Unions/Joins/ Projections/ Aggregation
a. From the Tools Palette, choose the required option as follows:
View Description
Union Used to combine the result set of two or more data
sources. For example, to show the names of all the
employees of a store which has different branches each
maintaining its own employee records table.
PUBLI C

View Description
Join Used to query data from two or more data sources,
based on a specified condition. For example, to retrieve
the sales of two stores maintaining individual table for
sales based on the customer ID.
Projection Used to filter or create a subset of the required columns
of a table or view for creating the model. For example,
selecting the employee name and sales quantity from a
table consisting of many more columns.
Aggregation Used to summarize data of a group of rows by
calculating values in a column. For example, to retrieve
total sales of a product in a month. The supported
aggregation types are sum, min, and max.
Note
The input for union, join, projection, and aggregation views can consist of data sources, union, join,
projection, or aggregation views.
You can have only one source of input for aggregation and projection views.
You can choose to create filters on projection and aggregation view attributes.
b. Map the input to the selected option.
Mapping attributes
a. To map attributes in a union view, drag and drop the required columns from Source to Target.
You can also modify the attribute mapping. For more information, see Managing Attribute Mappings
[page 142]
Tip
To create a system generated mapping, choose Auto Map By Name.
b. In case of a join view, join the columns of the source data sources.
Note
The output of a union view is the attributes that you added to the Target.
The output of a join view is the joined attributes. However, to add additional attributes to the view's
output, from the context menu, choose Add to Output.
To add attributes of projection or aggregation view to its output, choose Add to Output from the
context menu.
Optional Step: Creating Calculated Columns
The output of union, join, or projection view is stored under the Column node in the Output panel.To
perform calculations on these columns, do the following:
1. Right-click the Calculated Columns node.
2. From the context menu, choose New.
a. Enter the name.
b. Select the data type.
c. Enter the length and scale.
138
PUBLI C

d. To perform calculations on the output columns, do the following:
If you know the formula, enter the expression to perform the calculation.
From the list, select the required elements, operator, and functions.
e. Choose Add.
Optional Step: Applying Filter on Aggregation and Projection View Attributes
1. Right-click the required attribute.
2. From the context menu, choose Apply Filter.
3. Select the required operator.
4. Enter value.
5. Choose OK.
Note
You can edit a filter using filter expressions from the Output pane which provides you with more
conditions that can be used in the filter including AND, OR, and NOT. For example, to retrieve the
sales of a product where (revenue >= 100 AND region = India) OR (revenue >=50 AND region =
Germany).
Note
In order to define pattern-based filters in calculation views, you use the expression editor accessed
by double-clicking the expression icon under the filter node of the Output panel. If you type
match("ABC",'*abc*')
the equivalent SQL is
where ABC like %abc%
.
Adding Attributes and Measures to Calculation View Output
a. From the designer panel, choose the Output node.
b. To add an attribute, from the context menu, choose Add as Attribute.
c. To add a measure, from the context menu, choose Add as Measure.
Note
You can choose to associate a measure with the currency or unit of measure. To do so, select the
measure and in the Properties panel select Measure Type.
d. If you want to create calculated attributes, perform the following substeps:
a. In the Output pane, right-click Calculated Attributes.
b. From the context menu, choose New.
c. Enter a name and description.
d. On the Key tab, define the formula for the calculated attribute as follows:
a. Select a data type for the calculated attribute from the dropdown list.
b. Enter the length and scale.
PUBLI C

c. Select the required attributes, operator, and function.
d. To check the validity of the formula, choose Validate.
e. Choose Add.
e. To add a Description to the calculated attribute, write the formula as above.
e. If you want to create calculated measures, do the following:
a. In the Output pane, right-click Calculated Measures.
b. From the context menu, choose New.
b. Select the required Aggregation Type.
c. If you want to hide the measure while previewing data, choose Hidden.
d. Select the required data type.
e. Enter the length and scale.
f. Define the measure by selecting the required measures, operator, and function.
g. Choose Validate.
h. Choose OK.
Note
You can choose to rename the attributes, calculated attributes, measures, and calculated measures of
the view using the Rename button in the Output panel toolbar. However, renaming a field can impact
other objects that reuse the field as the new name is not reflected in the other objects.
Optional Step: Creating Counters
To obtain the number of distinct values of an attribute, do the following:
1. In the Output pane, right-click Counters.
3. Choose Add Attribute.
4. Choose OK.
Note
If you set the calculation view property Multidimensional Reporting as disabled, you can create a
calculation view without adding any measure to the calculation view output. A calculation view
without any measure works like an attribute view and is not available for reporting purposes.
You can choose to hide the attributes and measures that are not required for client consumption
by assigning value true to the property Hidden in the Properties pane.
You can choose to create variables, and bind them to attributes for filtering data. The values you
provide to the variables at runtime determine which data records are selected for consumption.
You can choose to create input parameters for the union, join, projection, aggregation, and
calculation view output. Also, if the calculation view data sources have input parameters, you can
map them to the calculation view input parameters.
You can choose to create hierarchy between attributes of the view.
You can specify a value for the view's Default Client property to filter the table data that is relevant
to a specific client as specified in the table fields at runtime, such as, MANDT or CLIENT. The
default value for the property is the one that is specified as a preference. If the property is set to
Dynamic, at runtime, the value set for the Session Client property is considered to filter table data.
The Session Client property is set while creating a user.
140
PUBLI C

You can set the cardinality of a join as required from the join properties. By default the cardinality
of a join is "empty". The empty cardinality can be specified in cases when you are not sure about
the right cardinality. If you specify the empty cardinatily, system during the join evaluation phase
diagnose the best suited cardinality and execute the join. Wrong selection of cardinality can lead to
erroneous data and out of memory issues.
Activate
a. To activate the view select Save and Activate from the toolbar.
Note
Activating the current view redeploys the impacted objects if an active version of the impacted
objects exist. You can also activate an object from the context menu of the object in the Navigator
view.
Tip
You can choose to activate the other objects (required or impacted objects) along with the current
object using the Save and Activate All from the toolbar.
Note
For more information about activation, see Activating Objects [page 153].
Note
1. For an active calculation view, you can preview the data of an intermediate node in a calculation view, which
helps to debug each level of a complex calculation scenario having join, union, aggregation, projection, and
output nodes. The data you preview for a node is for the active version of the calculation view. If no active
version for the object exists then you need to activate the object first. You can choose the Data Preview
option from the context menu of a node.
2. You can choose to generate documentation for the calculation view. For more information, see Generating
Object Documentation [page 167].
3. You can find the details of the functions available on content assist that is, by pressing Ctrl + Space in the
SQL Console while writing procedures in the SAP HANA SQLScript Reference.
Related Links
Assigning Variables [page 143]
Creating Input Parameters [page 145]
Creating Hierarchies [page 148]
Managing Attribute Mappings [page 142]
Using Currency and Unit of Measure [page 150]
Mapping Input Parameters [page 142]
PUBLI C

7.2.7.1 Mapping Input Parameters
You use this procedure to map the input parameters in the underlying data sources (analytic views and
calculation views) of the calculation view to the calculation view parameters. You can:
Map many data source parameters to one calculation view input parameter.
Perform a one on one mapping of the data source parameters to the calculation view parameters.
Procedure
1. In the Output pane, select an input parameter.
2. In the context menu, choose Manage Mappings.
3. In the Map Data Source Parameters dialog, map the data source input parameters with the calculation view
parameters.
Note
You can choose the Auto Map by Name option to automatically create the input parameters corresponding
to the source and perform a 1:1 mapping. You can also select a source input parameter and use the
following context menu options:
Create New Map 1:1 - to create the same input parameter for the calculation view as for the source, and
create a 1:1 mapping between them.
Map By Name - to map the source input parameter with the calculation view input parameter having
the same name.
Remove Mapping - to delete the mapping between the source and calculation view input parameter.
7.2.7.2 Managing Attribute Mappings
You use this procedure to map the source attribute to the target attribute if there are a large number of attributes,
or to assign a constant value to the target attribute.
Procedure
1. Right-click the attribute in the target list.
2. From the context menu, choose Manage Mappings.
a. To map the source to the target column, select the required source from the dropdown list.
b. To assign a default value to the constant column, enter the value in the Constant Value field. For more
information, see Constant Column [page 143].
c. Select the required data type.
d. Enter the length and scale as required.
142
PUBLI C

e. Choose OK.
7.2.7.3 Constant Column
In a union view, a Constant Column is created if there are any target or output attributes for which there are no
mappings to the source attributes. The default value for the constant column is NULL.
Note
The target attribute is mapped to all the sources.
For example, you have two tables with similar structures, Actual Sales and Planned Sales, corresponding to the
sales of products. You want to see the combined data in a single view, but differentiate between the data from the
two tables for comparison. To do so, you can create a union view between the two tables and have a constant
column indicating constant values like A & P, as shown below:
Actual Sales
Sales Product
5000 A1
2000 B1
Planned Sales
Sales Product
3000 A1
6000 B1
The result of this query can be as follows:
Actual Planned Indicator Sales Product
A 5000 A1
P 3000 A1
A 2000 B1
P 6000 B1
7.2.8 Assigning Variables
You use this procedure to assign variables to a filter at design time for obtaining data based on the values you
provide for the variable. At runtime, you can provide different values to the variable to view the corresponding set
of attribute data. You provide values to the variables either by entering the values manually, or by selecting them
from the Value Help dialog.
Note
PUBLI C

You can apply variables to attributes of analytic and calculation views.
The following types of variables are supported:
Type Description
Single Value Use this to filter and view data based on a single attribute
value. For example, to view the sales of a product where the
month is equal to January.
Interval Use this to filter and view a specific set of data. For example,
to view the expenditure of a company from March to April.
Range Use this to filter and view data based on the conditions that
involve operators such as "="(equal to), ">" (greater than),
"<" (less than), ">=" (greater than or equal to), and "<=" (less
than or equal to). For example, to view the sales of all
products in a month where the quantity sold is >= 100..
Each type of variable can be either mandatory or non-mandatory. For a mandatory variable, you need to provide a
value at runtime. For a non-mandatory variable, if you have not specified a value at runtime, you view unfiltered
data.
Note
You can check whether or not a variable is mandatory from the properties of the variable in the Properties pane.
In Analytic View
1. In the Scenario panel, select the Semantics node.
2. In the Variables/Input Parameters panel, select the Create Variable option.
a. Enter a name and description (label).
b. Select the required attribute from the dropdown list.
Note
c. Choose the required Selection Type from the dropdown list.
d. Select Multiple Entries if you want to provide different values at runtime to filter data. For example, to view
the Revenue from a period 2000 to 2005 and 2012.
e. If you want to assign the variable to attribute(s), in the Attribute Assignment panel, choose Add.
f. Select the attribute from the dropdown to which you want to assign this variable.
Note
You can also assign a variable to an attribute later in the Column panel of the Semantics node. To
assign a variable for an attribute, select the variable from the variable dropdown in the Variable
column.
g. Choose OK.
144
PUBLI C

In Calculation View
1. Create a Variable
1. In the Output pane, right-click the Variables node.
2. From the context menu, choose New and do the following:
b. Select the required attribute from the dropdown list.
Note
c. Choose the required Selection Type from the dropdown list.
d. If you want to specify a default value that is used to filter attribute data, enter the value in the Default
Value field.
e. Choose OK.
Note
You can also choose to create a variable using the Create Variable option from the context menu of an
attribute. In this case, the details of the variable are pre-filled.
2. Assign a Variable to a Filter
1. On the Output panel, right-click the attribute.
2. In the context menu, choose Apply Filter.
3. In the Operator dropdown list, choose Variable.
4. In the Variable dropdown list, choose the required variable.
5. Choose OK.
Note
You can also choose to create a variable and apply a filter using the Create Variable - Apply Filter option from
the context menu of an attribute.
7.2.9 Creating Input Parameters
Use this procedure to allow you to provide input for the parameters within stored procedures, in order to obtain a
desired functionality when the procedure is executed.
In an analytic view you use input parameters as placeholders during currency conversion and formulas like
calculated columns. When used in formulas, the calculation of the formula is based on the input that you provide
PUBLI C

at runtime during data preview. Input parameters are not used for filtering attribute data in analytic views that can
be achieved using variables.
In calculation views you can use input parameter during currency conversion, calculated measures, input
parameters of the script node, and to filter data as well.
You can apply input parameters in analytic and calculation views. If a calculation view is created using an analytic
view with input parameters, those input parameters are also available in the calculation view but you cannot edit
them.
The following types of input parameters are supported:
Type Description
Attribute Value/ Column Use this when the value of a parameter comes from an
attribute.
Currency (available in Calculation View only) Use this when the value of a parameter is in a currency
format, for example, to specify the target currency during
currency conversion.
Date (available in Calculation View only) Use this when the value of a parameter is in a date format, for
example, to specify the date during currency conversion.
Static List Use this when the value of a parameter comes from a user-
defined list of values.
Derived From Table (available in Analytic View and Graphical
Calculation View)
Use this when the value of a parameter comes from a table
column based on some filter conditions and you do not need
to provide any input at runtime.
Empty Use this when the value of a parameter could be anything
from the selected data type.
Direct Type (available in Analytic View) Use this to specify an input parameter as currency and date
during currency conversion.

In the case of analytic views, all input parameters are mandatory. However, in the case of calculation views, each
type of input parameter can be either mandatory or non-mandatory. For a mandatory input parameter, it is
necessary to provide a value at runtime. However, for a non-mandatory input parameter, if you do not specify a
value at runtime, the data for the column in which the input parameter is used remains blank.
Note
You can check whether an input parameter is mandatory or not from the properties of the input parameter in
the Properties pane.
Example
If you want to create a formula to analyze the annual sales of a product in various regions, you can use Year
and Region as input parameters.
If you want to preview a sales report with data for various countries in their respective currency for a
particular date for correct currency conversion, you can use Currency and Date as input parameters.
146
PUBLI C

Procedure
Analytic View
1. In the Output panel of the Data Foundation or Logical Join node, right-click the Input Parameters node.
Note
You can also create input parameters at the Semantics node level, using the Create Input Parameter option
in the Variables/Input Parameters panel.
b. Select the type of input parameter from the Parameter Type dropdown list.
Note
For the Column type of input parameter, you need to select the attribute from the dropdown list. At
runtime the value for the input parameter is fetched from the selected attribute's data.
For an input parameter of type Derived from Table, you need to select a table and one of it's
column as Return Column whose value is then used as input for the formula calculation. You can
also define conditions to filter the values of Return Column in the Filters panel. For example, to
calculate Discount for specific clients, you can create an input parameter based on Sales table and
return column Revenue with filter set on the Client_ID.
For Direct Type input parameter, specify the Semantic Type that describes the use parameter as a
currency or date, for example, to specify the target currency during currency conversion.
c. If required, select a data type.
d. Enter length and scale for the input parameter.
e. Choose OK.
Calculation View
1. In the Output panel, right-click the Input Parameters node.
b. Select the type of input parameter from the dropdown list.
Note
For the Attribute Value type of input parameter, you need to select the attribute from the
dropdown list. At runtime the value for the input parameter is fetched from the selected attribute's
data.
For an input parameter of type Derived from Table, you need to select a table and one of it's
column as Return Column whose value is then used as input for the formula calculation. You can
also define conditions to filter the values of Return Column in the Filters panel. For example, to
PUBLI C

calculate Discount for specific clients, you can create an input parameter based on Sales table and
return column Revenue with filter set on the Client_ID.
c. Select a data type.
d. Enter length and scale for the input parameter.
e. Choose OK.
7.2.10 Creating Hierarchies
You use this procedure to create hierarchies between attributes to enhance analysis by displaying attributes
according to their defined hierarchical relationships. Hierarchies can exist cross-attributes (that is, Country -
State - City) or within the values of a single attribute (that is, employee manager - employee direct report).
You structure and define relationships between attributes in the attribute view and calculation view using the
following hierarchy types:
Level Hierarchy
A level hierarchy is rigid in nature, and the root and the child nodes can be accessed only in the defined order.
Level hierarchies consist of one or more levels of aggregation. Attributes roll up into the next higher level in a
many-to-one relationship, and members at this higher level roll up into the next higher level, and so on to the
top level.
For example: an address hierarchy comprised of region, country, state, and so on.
Parent/Child Hierarchy
A parent/child hierarchy is a hierarchy in a standard view that contains a parent attribute. A parent attribute
describes a self-referencing relationship, or self-join, within the main table. Parent-child hierarchies are
constructed from a single parent attribute.
For example: a bill of materials hierarchy (parent and child) or an employee master (employee and manager)
hierarchy.
Procedure
Creating a Level Hierarchy
1. Select the Semantics node.
2. In the Hierarchies panel, choose Create option .
3. Enter a name and description for the hierarchy.
4. In the Hierarchy Type dropdown, select Level Hierarchy.
5. In the Node tab page do the following:
a. Select the required value from the Node Style dropdown list.
Note
Node style determines the composition of a unique node ID. The different values for node styles are as:
Level Name - the unique node ID is composed of level name and node name, for example "[Level
2].[B2]".
148
PUBLI C

Name Only - the unique node ID is composed of level name, for example "B2".
Name Path - the unique node ID is composed of the result node name and the names of all
ancestors apart from the (single physical) root node. For example "[A1].[B2]".
b. Add the required columns as levels from the drop-down list.
Note
You can select columns from the required table fields in the drop-down list to add to the view.
c. Select the required Level Type.
Note
The level type is used to specify formatting instructions for the level attributes.
For example, a level of the type LEVEL_TYPE_TIME_MONTHS can indicate that the attributes of the
level should have a text format such as "January", and LEVEL_TYPE_REGULAR indicates that a level
does not require any special formatting.
d. To control how the memebers of the hierarchy are ordered, select the required column in the OrderBy
drop-down list.
e. To sort the display of the hierarchy members in the ascending or descending order, select the required
option from the Sort Direction drop-down list.
6. In the Advanced tab page do the following:
a. Select the required value in the Aggregate All Nodes.
Note
This option indicates that data is posted on aggregate nodes and should be shown in the user interface.
For example, if you have the members A with value 100, A1 with value 10, and A2 with value 20 where
A1 and A2 are children of A. By default the value is set to false, and you will see a value of 30 for A.
With the value set to true, you will count the posted value 100 for A as well and see a result of 130.
If you are sure that there is no data posted on aggregate nodes you should set the option to false. The
engine will then calculate the hierarchy faster as when the option is set.
Note that this flag is only interpreted by the SAP HANA MDX engine. In the BW OLAP engine the node
values are always counted.
b. Enter a value for the default member.
c. Select the With Root Node checkbox if required.
Note
If a hierarchy does not have a root node but needs one for reporting use case, set the option to true.
This will create a root node.
d. If the level hierarchy needs to support mulitple parents for its elements for example, Country 'Turkey' to
be assigned to two regions 'Europe' and 'Asia', select the Mulitple Parent checkbox.
7. Choose OK.
PUBLI C

Creating a Parent/Child Hierarchy
1. Select the Semantics node.
2. In the Hierarchies panel, choose Create option .
3. Enter a name and description for the hierarchy.
4. In the Hierarchy Type dropdown, choose Parent Child Hierarchy.
5. In the Node tab page, add the parent and child nodes by selecting the Parent Node and Child Node from the
drop-down list.
6. In the Advanced tab page, do the following:
a. Select the required value in the Aggregate All Nodes.
Note
This option indicates that data is posted on aggregate nodes and should be shown in the user interface.
For example, if you have the members A with value 100, A1 with value 10, and A2 with value 20 where
A1 and A2 are children of A. By default the value is set to false, and you will see a value of 30 for A.
With the value set to true, you will count the posted value 100 for A as well and see a result of 130.
If you are sure that there is no data posted on aggregate nodes you should set the option to false. The
engine will then calculate the hierarchy faster as when the option is set.
Note that this flag is only interpreted by the SAP HANA MDX engine. In the BW OLAP engine the node
values are always counted.
b. Enter a value for the default member.
c. Select the With Root Node checkbox if required.
Note
If a hierarchy does not have a root node but needs one for reporting use case, set the option to true.
This will create a root node.
d. If the level hierarchy needs to support mulitple parents for its elements for example, Country 'Turkey' to
be assigned to two regions 'Europe' and 'Asia', select the Mulitple Parent checkbox.
7. Choose OK.
Note
The hierarchies belonging to an attribute view are available in an analytic view that reuses the attribute view, in
read-only mode. However, the hierarchies belonging to an attribute view are not available in a calculation view
that reuses the attribute view.
7.2.11 Using Currency and Unit of Measure
Use this procedure to define a measure as an amount or weight in the analytical space and to perform currency
conversion.
To simplify the process of currency conversion, the system provides a list of currencies, and exchange rates
based on the tables imported for currency. Currency conversion is performed based on source currency, target
150
PUBLI C

currency, exchange rate, and date of conversion. You can also select currency from the attribute data used in the
view.
For example, you need to generate a sales report for a region in a particular currency, and you have sales data in
database tables in a different currency. You can create an analytic view by selecting the table column containing
the sales data in this other currency as a measure, and perform currency conversion. Once you activate the view,
you can use it to generate reports.
Note
Currency conversion is enabled for analytic views and base measures of calculation views.
Prerequisites
You have imported tables T006 and T006A for Unit of Measure.
You have imported TCURC, TCURF, TCURN, TCURR, TCURT, TCURV, TCURW, and TCURX for currency.
Procedure
1. Select a measure.
2. In the Properties pane, select Measure Type.
3. If you want to associate the measure with a currency, do the following:
a. In the Measure Type dropdown list, select the value Amount with Currency.
b. In the Currency Dialog, select the required Type as follows:
Type Purpose
Fixed To select currency from the currency table TCURC.
Attribute To select currency from one of the attributes used in the
view.
c. Select the required value, and choose OK.
d. If you want to convert the value to another currency, choose Enable for Conversion.
a. To select the source currency, choose Currency.
b. Select the target currency.
Note
For currency conversion, in addition to the types Fixed and Attribute, you can select an Input
Parameter to provide target currency at runtime. If you select an input parameter for specifying
target currency and deselect Enable for Conversion checkbox, the target currency field is cleared
because input parameters can be used only for currency conversion.
c. To specify exchange rate type, in the Exchange Rate Types dialog, select the Type as follows:
PUBLI C

Type Purpose
Fixed To select exchange rate from the currency table
TCURW.
Input Parameter To provide exchange rate input at runtime as input
parameter.
d. To specify the date for currency conversion, in the Conversion Date dialog, select the Type as follows:
Type Purpose
Fixed To select conversion date from the calendar.
Attribute To select conversion date from one of the attributes
used in the view.
Input Parameter To provide conversion date input at runtime as input
parameter.
e. To specify the schema where currency tables are located for conversion, in the Schema for currency
conversion, select the required schema.
f. To specify the client for which the conversion rates to be looked for, in the Client for currency
conversion, select the required option.
e. From the dropdown list, select the required value that is used to populate data if the conversion fails:
Option Result
Fail In data preview, the system displays an error for
conversion failure.
Set to NULL In data preview, the value for the corresponding records
is set to NULL.
Ignore In data preview, you view the unconverted value for the
corresponding records.
4. If you want to associate a measure with a unit of measure other than currency, do the following:
a. Select the value Quantity with Unit of Measure in the Measure Type dropdown list.
b. In the Quantity Units dialog , select the required Type as follows:
Type Purpose
Fixed To select a unit of measure from the unit tables T006
and T006A.
Attribute To select a unit of measure from one of the attributes
used in the view.
c. Select the required value, and choose OK.
5. Choose OK.
Note
You can associate currency or unit of measure with a calculated measure, and perform currency conversion for
a calculated measure by editing it.
152
PUBLI C

7.2.12 Activating Objects
You activate objects available in your workspace to expose the objects for reporting and analysis. Based on your
requirements, you can do the following:
Activate - Deploys the inactive objects.
Redeploy - Deploys the active objects in one of the following scenarios:
If your runtime object gets corrupted or deleted, and you want to create it again.
If an object goes through client-level activation and server-level activation but fails at MDX, and the object
status is still active.
The following activation modes are supported:
Activate and ignore the inconsistencies in impacted objects - To activate the selected objects even if it results
in inconsistent impacted objects. For example, if you choose to activate an object A that is used by B and C,
and it causes inconsistencies in B and C but you can choose to go ahead with the activation of A. This is the
default activation mode.
Stop activation in case of inconsistencies in impacted objects - To activate the selected objects only if there
are no inconsistent impacted objects.
Note
Irrespective of the activation mode, if even one of the selected objects fails (either during validation or during
activation), the complete activation job will fail and none of the selected objects will be activated.
Depending on where you invoke the activation, redeployment or cascade activation, the behavior is as follows:
Context Activate Redeploy
Quick Launch tab page A dialog box appears with a preselected
list of all your inactive objects.
A dialog box appears with a list of active
objects in your workspace.
Package context menu A dialog box appears with a preselected
A dialog box appears with a list of active
objects in your workspace.
Content context menu A dialog box appears with a preselected
Not applicable
Editor
If you select Save and Activate,
current object is activated and the
impacted objects are redeployed if
an active version for the impacted
objects exist.
If you select Save and Activate All, a
dialog box appears with a
preselected list of the selected
object along with all the required
and impacted objects.
Not applicable
Object context menu A dialog box appears with a preselected
list of the selected object along with all
the required objects.
A redeployment job is submitted for the
selected object.
Note
PUBLI C

If an object is the only inactive object in the workspace, the activation dialog box is skipped and the
activation job is submitted.
If an object is inactive and you want to revert back to the active version, from the editor or object context
menu, choose Revert To Active.
In the Activate dialog, you can select the Bypass validation checkbox in order to skip validation before
activation to improve the activation time. For example, if you have imported a number of objects and want
to activate them without spending time on validation.
Note
During delivery unit import, full server side activation is enabled, activation of objects after import is done. In
this case all the imported objects are activated (moved to active table), even if there are errors in activated or
impacted objects. But the objects for which activation results in error are considered as broken or inconsistent
objects which means that the current runtime representation of these objects is not in sync with the active
design time version. The broken objects are shown in the Navigator view with an x along side.
Note
The behavior of activation job is as follows:
The status (completed, completed with warnings, and completed with errors) of the activation job
indicates whether the activation of the objects is successful or failed.
In case of failure that is when the status is completed with errors, the process is rolled back. This means,
even if there are individual objects successfully activated, since the activation job is rolled back, none of
the objects are activated.
Even in case of failure, the job log shows success in the summary part. This is to help the user to indicate
that those objects were successfully activated without any issues. But because the entire job is a failure,
none of the objects are activated and all of them are rolled back.
When you open the job log, the summary list shows only those objects that are submitted for activation. It
does not list all the affected objects. They are listed only in detail section.
7.3 Creating Decision Tables
You use this procedure to create a decision table to model related business rules in a tabular format for decision
automation. Using decision tables you can manage business rules, data validation, data quality rules without any
IT knowledge on technical languages like, SQL Script, MDX . A data architect or a developer creates the decision
table and activates it. The active version of the decision table can be used in applications.
154
PUBLI C

Procedure
1. Set Parameters
1. In the Modeler perspective, expand <System Name> Content <Package Name> .
2. In the context menu of the package, choose New Decision Table .
3. In the New Decision Table dialog box, enter a name and description for the decision table.
4. To create data foundation for the decision table, perform substeps of the required scenario given in the table
below:
Scenario Substeps
Create a decision table from scratch.
1. Choose Create New.
2. Choose Next.
Note
If you launch the New Decision Table dialog from
the Quick Launch tab page, specify the package
where you want to save the decision table.
3. Add the required tables, table type or an information
view to the Selected list.
Note
You can choose to add the required data sources to
the decision table later by dragging them from the
Catalog node in the Navigator view to the Data
Foundation panel.
4. Choose Finish.
Create a decision table from an existing decision table.
1. Choose Copy From.
Note
If you launch the New Decision Table dialog from
the Quick Launch page, specify the package where
you want to save the decision table.
2. Select the required decision table.
3. Choose Finish.
Note
Only an active version of an information view can be used. Any changes made to the information view
are not reflected in the decision table.
You can create a decision table using an analytic view only if it has a calculated attribute.
If you create a decision table using an analytic view, the analytic view must have a calculated attribute.
If you choose to create a decision table based on a table type or an information view, you cannot add
any other data source. This implies that a decision table can be based on multiple tables or a table type
or an information view.
PUBLI C

You can add only one table type or information view to the data foundation.
You can mark table type columns and information view columns only as conditions.
While designing decision table using information view, you can only use view's attributes as conditions
and not the other attributes. In addition, you can use only parameters as actions.
2. Create Joins
1. If you want to define a relationship between tables to query data from two or more tables, do the following:
a. In the editor pane, from the context menu, choose Create Join.
b. In the Create Join dialog, select the required tables, columns, join type, and cardinality.
c. Choose Create Join.
Note
You can also create a join between table columns by dragging it from one table column to another table
column. The supported join types are inner, left outer and right outer.
3. Add Conditions and Actions
1. In the Data Foundation view, select the required column, and perform substeps of the required scenario given
in the table below:
Scenario Substeps
Include table field in the output structure. From the context menu, choose Add as Attribute.
Note
Attributes contains a subset of columns that you use to
derive conditions and actions.
Add conditions based on which you want to derive actions.
1. In the Output view, select the required attributes .
2. From the context menu, choose Add as Conditions.
Add actions for the selected conditions.
1. In the Output view, select the required attributes .
2. From the context menu, choose Add as Actions.
2. To add condition values, do the following:
a. In the Decision Table view, right-click a condition, and choose Add Conditions Values.
Note
The supported data types for an operator are:
156
PUBLI C

Operator Supported Data Types Syntax
Not Equal To Number & CHAR-based != ABC
In Number & Char-based In ABC;CDA
Not In Number & Char-based Not In A;B;C
Like CHAR-based Like Abc*
Not Like CHAR-based Not Like Abc*
Greater Than
Greater Than or Equals
Number & CHAR-based >20
>=20
Less Than
Less Than or Equals
Number & CHAR-based <10
<=10
Between Number Between 20 and 30
Before Date Dates Before 2012-12-12
Or
< 2012-12-12
After Date Dates After 2012-12-12
Or
> 2012-12-12
Between Date Dates Between 2012-12-12 and
2012-12-25
b. Enter a value, and choose OK.
Note
If a database table column is used as condition, you can use the value help dialog to select the
condition values. You can select multiple values at one time. You can edit a condition value by
selecting the condition, and entering a value.
You can enter a pattern for the condition values having data type as VARCHAR. The pattern must
be prefixed with the LIKE and NOT LIKE operators. For example, LIKE a*b or NOT LIKE a?b. If the
LIKE or NOT LIKE operator is not present the pattern is treated as a string
3. To set action values, right-click an action cell, and enter value.
4. If you want to write a complex expression as action value, do the following:
a. Right-click the action field.
b. From the context menu, choose Set Dynamic Value.
c. Write the expression for example, PRICE-(PRICE*0.1).
d. To edit a value you need to select that value.
Note
PUBLI C

You can use parameters and table columns of the same data type as that of the action in expressions.
5. To assign a value to a condition or an action based on the table data, choose Open Value Help Dialog, and do
the following:
a. In the Value Help for Column dialog, enter the search string, and choose Find.
Note
If you do not provide a value for search and choose Find, all the data corresponding to the selected
column is shown.
b. Select a value, and choose OK.
Remember
You can provide an alias name to a condition or an action by editing the value of Alias name property.
You can choose to create parameters and use them as conditions or actions. The values you provide to the
parameters at the runtime determine which data records are selected for consumption. For more
information regarding how to use parameters, Using Parameters in a Decision Table [page 161].
You can export decision table data to an excel sheet using context menu option Export Data to Excel in the
Decision Table view. You can also import decision table data from an excel using context menu option
Import Data from Excel in the Decision Table view.
You can arrange the condition and action columns of the decision table depending on how you want them
to appear. For more information, see Changing the Layout of a Decision Table [page 160].
6. Optional Step: Validate Decision Table
1. To set the rules that you want to use for validation do the following:
a. Choose Window Preferences .
b. In the Preferences dialog box, expand Modeler Validation Rules .
c. In the Validation Rules view, select Decision Table checkbox to check for all the rules during validation.
d. If you want to check for individual rules, select the required rules.
e. Choose OK.
2. In the decision table editor, choose Validate in the editor toolbar .
Note
In the Job Log section, you can see the validation status and detailed report of the decision table
7. Activate Decision Table
1. Choose File Save .
2. From the context menu of the decision table, choose Activate.
158
PUBLI C

Note
You can choose to save and activate the view from the editor using .
Result: On successful activation, a procedure corresponding to the decision table is created in _SYS_BIC schema.
The name of the procedure is in the format, <package name>/<decision table name>. In addition, if a parameter
is used as an action in the decision table, the corresponding table type is created in _SYS_BIC schema. The name
of the table type is in the format, <package name>/<decision table name>/TT.
Remember
If in a decision table, parameters are used as conditions then corresponding IN parameters are generated.
Also, if the parameters are used as actions then an OUT parameter is generated.
8. Execute Decision Table Procedure
1. To execute the decision table procedure perform the following steps as required:
Data Source Condition Action Script
Physical tables Physical table
column
Physical table
column
call "<schema name>"."<procedure
name>";
Physical tables Parameters Physical table
column
call "<schema name>"."<procedure
name>"(<IN parameter>,,<IN
parameter>);
Physical tables Physical table
column
Parameters call "<schema name>"."<procedure
name>"(?);
Physical tables Parameters Parameters call "<schema name>"."<procedure
parameter>,?);
Information
View
View attributes Parameters call "<schema name>"."<procedure
name>"(?);
Information
View
Parameters Parameters call "<schema name>"."<procedure
parameter>,?);
Table Type Table Type
column
Parameters call "<schema name>"."<procedure
name>"(?);
Table Type Parameters Parameters call "<schema name>"."<procedure
parameter>,?);
Remember
PUBLI C

The order of the parameter while executing the procedure must be same as in the Output panel, and not as
used in the Decision Table.
Tip
You can view the procedure name using the Open Definition context menu option on the selected
procedure.
Result: On execution of the procedure, the physical table data is updated (if no parameters are used) based on the
data that you enter in the form of condition values and action values.
Remember
If parameters are being used as actions in a decision table, the physical table is not updated.
9. Data Preview
To preview the populated data in the decision table, in the Decision Table editor, from the toolbar, choose Open
Data Preview.
Restriction
Data preview is supported only if:
A decision table is based on physical tables and has at least one parameter as action.
A decision table is based on information views and parameter(s) as action.
7.3.1 Changing the Layout of a Decision Table
Use this procedure to change the decision table layout by arranging the condition and action columns. By default,
all the conditions appear as vertical columns in the decision table. You can choose to mark a condition as a
horizontal condition, and view the corresponding values in a row. The evaluation order of the conditions is such
that first the horizontal condition is evaluated and then the vertical ones.
Note
You can only change the layout of a decision table if it has more than one condition. You can select only one
condition as horizontal condition.
Procedure
1. In the context menu of the Decision Table editor, choose Change Layout.
160
PUBLI C

2. If you want to view a condition as a horizontal condition, in the Change Decision Table Layout dialog, select
Table has Horizontal Condition (HC) checkbox.
Note
By default the first condition in the list of conditions is marked as horizontal.
3. In the Conditions and Actions sections, choose options on the right-hand side of the dialog box to arrange the
conditions and actions in the desired sequence.
Note
The available options to arrange the conditions in a sequence are:
Move Condition to Top
Move Condition Up
Move Condition Down
Move Condition to Bottom
4. Choose OK.
5. Save the changes.
Note
You can also set a condition as horizontal from the context menu of the condition in the Output view. You can
also arrange the conditions and actions in the desired sequence in the Output view using the respective
buttons.
7.3.2 Using Parameters in a Decision Table
You use this procedure to create a parameter that can be used to simulate a business scenario. You can use
parameters as conditions and actions in the decision table at design time. Parameters used as conditions,
determine the set of physical table rows to be updated based on the condition value that you provide at runtime
during procedure call. Parameters used as actions, simulate the physical table without updating to it.
The following parameter types are supported:
Type Description
Static List Use this when the value of a parameter comes from a user-
defined list of values.
Empty Use this when the value of a parameter could be anything of
the selected data type.
Example
Consider a sales order physical table with column headers as follows:
ID Name Supplier Model Price Quantity
PUBLI C

If you want to evaluate Discount based on the Quantity and Order Amount, you can create two parameters,
Order Amount and Discount. Use Quantity and Order Amount as the condition, and Discount as the action. The
sample decision table could be:
Quantity Order Amount Discount
>5 50000 10
>=10 100000 15
Procedure
1. Create a Parameter
1. In the Output pane, right-click the Parameters node.
2. From the context menu, choose New and do the following:
b. Select the required data type from the dropdown list.
c. Enter the length and scale as required.
d. Choose the required Type from the dropdown list.
Note
If you have selectedStatic List for Type, choose Add in the List of Values section to add values. You can
also provide an alias for the enumeration value.
e. Choose OK.
2. Use Parameter as Condition or Action
1. In the Output pane, expand the Parameters node.
2. Right-click a parameter, choose Add As Conditions/ Add as Actions.
7.4 Managing Object Versions
7.4.1 Switching Ownership of Inactive Objects
Use this procedure to take over the ownership of the inactive version of an object from other users' workspace.
Objects in edit mode in other workspaces are not available for modification. In order to modify such objects you
need to own the inactive object.
162
PUBLI C

The options available for changing the inactive object ownership are as follows:
Option Purpose
Switch Ownership To take over multiple inactive objects from other users.
Inactive objects that do not have an active version are also
available for take over using this option
Take Over To take a single inactive object from another workspace that
you wish to edit using the editor.
Note
Using this functionality you can only own the inactive version of the object. The active version is owned by the
user who created and activated the object.
Prerequisite
You have obtained the Work in Foreign Workspace authorization.
Procedure
1. If you want to own multiple inactive objects from other workspaces, do the following:
a. In the Quick Launch page, choose Switch Ownership.
b. In the Source User field, select the user who owns the inactive objects.
c. Add the required inactive objects to the Selected Models section.
d. Choose OK.
2. If an object opens in read-only mode and you want to edit it, do the following:
a. In the editor toolbar, select Switch Version.
b. Choose Take Over.
Note
You can choose to save the changes made by the other user (previous owner of the inactive version) to
the inactive version of the object.
7.4.2 Toggling Versions of Content Objects
You use this procedure to view the active version of an information object while working with its inactive version
for example, to view the changes made to the active version.
PUBLI C

Procedure
2. Select the required object from a package.
3. From the context menu, choose Open.
4. In the editor pane, choose Show Active Version.
5. Compare the inactive and active versions of the object.
6. Choose OK.
7.4.3 Viewing Version History of Content Objects
You use this procedure to view the version details of an information model for tracking purposes.
Procedure
2. Select the required object from a package.
3. From the context menu, choose History.
7.5 Working with Objects
7.5.1 Managing Layout
You use this procedure to adjust the data foundation and logical view layout comprising user interface controls
like, tables and attribute views in a more readable manner. This functionality is supported for attribute views and
analytic views.
The options available are as follows:
Option Purpose Substeps
Auto Arrange Use this option to arrange the user
interface elements automatically.
In the editor tool bar, choose .
Show outline Use this option to view an outline of the
elements arranged so that , you do not
have to navigate in the editor using
horizontal and vertical scrollbars.
In the editor tool bar, choose .
Highlight related tables Use this option if you want to view only
those tables that are related to a table
selected in the editor.
1. In the editor, right-click the
selected table.
164
PUBLI C

Option Purpose Substeps
2. From the context menu, choose
Highlight related tables.
Display Use this option if you have a table with a
large number of columns in the editor,
and you want to view them in a way that
meets your needs: for example, only the
table name, or only joined columns, or
the expanded form with all the columns.
1. In the editor, right-click the relevant
table.
2. From the context menu, choose
Display.
3. If you want to view only the table
name, choose Collapsed.
4. If you want to view all the columns
of the table, choose Expanded.
5. If you want to view only the joined
columns of the table, choose Joins
only.
7.5.2 Filtering Packages and Objects
You use this procedure to filter the content, and view packages and objects that you want to work with. If you
apply a filter at the package level, all the packages including subpackages that satisfies the search criteria are
shown. You can apply a filter for packages only on the Content node in the Navigator view. You can apply a filter for
objects at the package level including subpackages.
Applying Filter for Packages
1. In the Navigator view, select the Content node.
2. In the context menu, choose Filter Packages....
3. In the Filter Packages dialog, enter the filter text.
4. If you want to search for the exact word written in the filter text, select Match whole word checkbox.
5. Choose OK.
Note
If a filter already exists on the Content node, the new filter will overwrite the existing one. You can also apply the
previous filter on the Content using Apply Filter '<filter text>' option.
Applying Filter for Objects
1. In the Navigator view, expand the Content node.
2. In the context menu of a package, choose Filter Objects....
3. In the Filter Objects dialog, enter the filter text.
PUBLI C

4. If you want to search for the exact word written in the filter text, select Match whole word checkbox.
5. If you want to apply the filter on the subpackages, choose Apply filter to sub packages checkbox.
If you want to apply a filter only if no filter already exists on the selected package or otherwise retain the
existing filter, choose Apply only if no filter already exists.
If you want to apply filter even if there is an existing filter, choose Apply to all and overwrite existing.
6. Choose OK.
Note
If a filter already exists on the package, the new filter will overwrite the existing one.
7.5.3 Refactoring Objects
You use this procedure to restructure your content objects in the Navigator view without changing their behavior.
As you move the objects, their references are automatically adjusted. Objects available for refactoring are,
packages, attribute views, analytic views, graphical calculation views, and analytic privileges.
The activation state of the objects from source package to target package is maintained as follows:
At Source At Target
Base Object- active
Impacted Object- active
Base Object- active
Base Object- inactive
Impacted Object- inactive
Base Object- active
Base Object- active
Note
An impacted object (also known as affected object) is the one that uses the base object. For example, an
analytic view using an attribute view is called impacted object for that attribute view.
1. Open the Modeler perspective.
3. Select the required objects, in the context menu, choose Refactor Move .
4. In the Move dialog, select the target package where you want to move the package/objects, and choose Next.
5. If you want to skip the movement of objects/packages, in the Changes to be performed panel, deselect them.
6. Choose Finish.
166
PUBLI C

7.5.4 Validating Models
You use this procedure to check if there are any errors in an information object and if the object is based on the
rules that you specified as part of preferences. For example, the "Check join: SQL" rule checks that the join is
correctly formed.
For more information about setting preferences, see Setting Preferences for Modeler .
Procedure
If you are in the Modeler perspective, do one of the following:
1. On the Quick Launch page, choose Validate.
2. From the Available list, select the required models that system must validate.
3. Choose Add.
4. Choose Validate.
To validate an individual view from the editor choose the validate button in the toolbar.
Note
The validation in the modeler perspective checks for both the client-side and server-side validation rules.
If you are in the SAP HANA Development perspective, perform one of the following:
To validate against the server-side rules, perform the following:
1. In the Project Explorer select the required object.
2. In the context-menu, choose Team Check .
To validate against the client-side rules, choose the Save and Validate option from the editor toolbar.
7.5.5 Generating Object Documentation
Use this procedure to capture the details of an information model in a single document. This helps you view the
necessary details from the document, instead of referring to multiple tables. The following table specifies the
details that you can view from the document.
Type Description
Attribute View General object properties, attributes, calculated attributes
(that is, calculated columns of type attribute), data
foundation joins, and cross references
Analytic View General object properties, private attributes, calculated
attributes (that is, calculated columns of type attribute),
attribute views, measures, calculated measures (that is,
calculated columns of type measure), restricted measures
(that is, restricted columns), variables, input parameters,
data foundation joins, logical view joins, and cross references
PUBLI C

Type Description
Calculation View General object properties, attributes, calculated attributes,
measures, calculated measures, counters, variables, input
parameters, calculation view SQL script, and cross
references
Procedure
1. In the eidtor, choose Auto Documentation option from the editor toolbar.
2. In the pop-up dialog, browse the location where you want to save the file.
3. Choose Finish.
Note
The generated document is placed at the specified location under the SAP HANA system folder. For
example, if you choose to generate the pdf for an attribute view, it is placed at <specified location>
<SAP HANA system> <package_name> ATTRIBUTE_VIEW <package_name>.<view_name>.pdf
7.5.6 Enabling Multilanguage Support for Objects
You use this procedure to enable translation of text pertaining to objects and their elements that did not have
multilanguage support.
For example, you can enable multilanguage support for models along with their elements like attributes and
measures in different languages.
Procedure
1. From the Quick Launch tab page, choose Migrate.
2. Select the required objects.
3. Choose Add.
4. Choose OK.
Result
Object texts along with the corresponding elements are flagged for translation. These objects can be viewed in
multiple languages provided that the object texts are translated.
168
PUBLI C

7.5.7 Checking Model References
You use this procedure to identify whether an information model is referenced by any other information model(s).
Procedure
1. In the Modeler perspective, expand the system node in the Navigator view.
3. Expand the required package node.
4. Select the required object.
5. From the context menu, choose Where Used.
7.5.8 Viewing the Job Log
The job log displays information related to requests entered for a job. A job log consists of two tab pages as
follows:
Current: Lists all waiting, running, and last five jobs.
History: Lists all the jobs.
Note
You can terminate the job only if it is in the waiting state.
You can perform the following operations using the job log:
Open Job Details: Use this to view the job summary in the current tab page.
Open Job Log File: Use this to view the information pertaining to a job in detail using the internal browser.
Clear Log Viewer: Use this to delete all the job from the current tab page.
Export Log File: Use this to export the log file to a target location other than the default location for further
reference.
Delete Job: Use this to delete single job from the current tab page.
7.5.9 Maintaining Search Attributes
You use this procedure to enable an attribute search for an attribute used in a view. Various properties related to
attribute search are as follows:
Freestyle Search: Set to True if you want to enable the freestyle search for an attribute. You can exclude
attributes from freestyle search by setting the property to False.
Weights for Ranking: To influence the relevancy of items in the search results list, you can vary the weighting
of the attribute. You can assign a higher or lower weighting (range 0.0 to 1.0). The higher the weighting of the
PUBLI C

attribute, the more influence it has in the calculation of the relevance of an item. Items with a higher relevance
are located higher up the search results list. Default value: 0.5.
Note
To use this setting the property Freestyle Search must be set to True.
Fuzziness Threshold: This parameter is reserved for the future usage of the fault-tolerant search..
Note
It is recommended to use the default values for ranking and fuzziness at the beginning. You can fine-tune the
search settings based on the experience gained using the search. You can also fine-tune the search using
feedback collected from your users.
7.5.10 Previewing Data of Content Objects
You use this procedure to preview the content of content models for analysis purposes.
2. Select the object from a package for which you want to view the content.
3. From the context menu, choose Data Preview.
The system displays the content in different formats as shown in the table below.
Tab Page Displays
Raw Data All attributes along with data in a table format.
Distinct values All attributes along with data in a graphical format.
Analysis All attributes and measures in a graphical format.
Tip
If there are inconsistencies in runtime information (that is, calculation views in catalog or in tables related
to runtime) of a view, you might get invalidated view error. In such cases, you need to redeploy the view in
order to correct the inconsistencies with runtime information.
Note
If you refresh the Analysis or Distinct values view, the behavior is as follows:
The selection of the columns in Lables Axis and Values Axis panel of the Analysis view, and in the select
objects panel of the Distinct values view gets cleared.
The data in the Raw Data view gets cleared and you need to perform refresh in the Raw Data view to
fetch the latest data.
4. Navigate to the required tab page and view the content.
170
PUBLI C

7.5.11 Functions used in Expressions
This topic covers the functions that you can use while creating expressions like, calculated attributes and
calculated measures.
Conversion Functions
Function Syntax Purpose Example
int int int(arg) convert arg to int type int(2)
float
float float(arg)
convert arg to float type float(3.0)
double
double double (arg) convert arg to double type double(3)
sdfloat
sdfloat sdfloat (arg)
convert arg to sdfloat type
decfloat
decfloat decfloat (arg)
convert arg to decfloat type
fixed fixed fixed (arg, int, int) arg2 and arg3 are the
intDigits and fractdigits
parameters, respectively.
Convert arg to a fixed type of
either 8, 12, or 16 byte length,
depending on intDigits and
fractDigits
fixed(3.2, 8, 2) + fixed(2.3, 8,
3)
string
string string (arg) convert arg to string type
raw
raw raw (arg) convert arg to raw type
date
date date(stringarg)
date date(fixedarg)
date date(int, int)
date date(int, int, int)
date date(int, int, int, int)
date date(int, int, int, int, int)
date date(int, int, int, int, int,
int)
convert arg to date type. The
first version parses a string in
the format "yyyy-mm-dd
hh:mi:ss" where trailing
components except for the
year may be omitted. The
version with one fixed
number arg strips digits
behind the comma and tries
to make a date from the rest.
The other versions accept the
individual components to be
set.
date(2009) -> date('2009')
date(2009, 1, 2) ->
date('2009-01-02')
date(fixed(20000203135026
.1234567, 10, 4)) ->
date('2000-02-03 13:50:26')
longdate
longdate longdate(stringarg)
longdate longdate(fixedarg)
longdate longdate(int, int, int)
longdate longdate(int, int, int,
int, int)
int, int, int)
convert arg to longdate type,
similar to date function
above.
longdate(fixed(2000020313
5026.1234567, 10, 5)) ->
longdate('2000-02-03
13:50:26.1234500')
longdate(2011, 3, 16, 9, 48,
12, 1234567) ->
longdate('2011-03-16
09:48:12.1234567')
PUBLI C

int, int, int, int)
time
time time(stringarg)
time time(fixedarg)
time time(int, int)
time time(int, int, int)
convert arg to time type,
similar to date function above
String Functions
Function Syntax Purpose
strlen
int strlen(string) returns the length of a string in bytes, as
an integer number.
midstr
string midstr(string, int, int)
returns a part of the string starting at
arg2, arg3 bytes long.
arg2 is counted from 1 (not 0)
leftstr
string leftstr(string, int)
returns arg2 bytes from the left of the
arg1. If arg1 is shorter
than the value of arg2, the complete
string will be returned.
rightstr
string rightstr(string, int)
returns arg2 bytes from the right of the
arg1. If arg1 is shorter
than the value of arg2, the complete
string will be returned.
instr
int instr(string, string)
returns the position of the first
occurrence of the second string
within the first string (>= 1) or 0, if the
second string is not
contained in the first.
hextoraw
string hextoraw(string)
convert a hexadecimal representation of
bytes to a string of
bytes. The hexadecimal string may
contain 0-9, upper or lowercase a-f and
no spaces between the two digits of a
byte; spaces between bytes are allowed.
rawtohex
string rawtohex(string)
convert a string of bytes to its
hexadecimal representation.
172
PUBLI C

The output will contain only 0-9 and
(upper case) A-F, no spaces and is twice
as many bytes as the original string.
ltrim string ltrim(string)
string ltrim(string, string)
removes a whitespace prefix from a
string. The Whitespace characters may
be specified in an optional argument.
This functions operates on raw bytes of
the UTF8-string and has no knowledge
of multi byte codes (you may not
specify multi byte whitespace
characters).
rtrim string rtrim(string)
string rtrim(string, string)
removes trailing whitespace from a
string. The Whitespace characters may
be specified in an optional argument.
This functions operates on raw bytes of
the UTF8-string and has no knowledge
of multi byte codes (you may not
specify multi byte whitespace
characters).
trim string trim(string)
string trim(string, string)
removes whitespace from the beginning
and end of a string.
lpad string lpad(string, int)
string lpad(string, int, string)
add whitespace to the left of a string. A
second string argument specifies the
whitespace which will be added
repeatedly until the string has reached
the intended length. If no second string
argument is specified, chr(32) (' ') will
be added. This function operated on
UTF-8 bytes and has no knowledge of
unicode characters (neither for the
whitespace string nor for length
computation).
rpad string rpad(string, int)
string rpad(string, int, string)
add whitespace to the end of a string. A
second string argument specifies the
whitespace which will be added
repeatedly until the string has reached
the intended length. If no second string
argument is specified, chr(32) (' ') will
be added. This function operated on
UTF-8 bytes and has no knowledge of
unicode characters (neither for the
whitespace string nor for length
computation).
replace
string replace(string, string, string) replace every occurrence of arg2 in arg1
with arg3 and return the resulting string
Mathematical Functions
PUBLI C

sign int sign(double)
int sign(time)
int sign(date)
Sign returns -1, 0 or 1
depending on the sign of its
argument. Sign is
implemented for all numeric
types, date and time.
abs double abs(double)
decfloat abs(decfloat)
decfloat abs(decfloat)
time abs(time)
Abs returns arg, if arg is
positive or zero, -arg else.
Abs is implemented for all
numeric types and time.
round
.
double round(double, int) round does rounding of
absolute values toward zer
while the sign is retained
round(123.456, 0) = 123
round(123.456, 1) = 123.5
round(-123.456, 1) = -123.5
round(123.456, -1) = 120
rounddown
double rounddown(double,
int)
rounddown rounds toward
negative infinity making
rounddown(-1.1, 0) = -2
rounddown(123.456, -1) =
120
rounddown(-123.456, -1) =
-130
Date Functions
utctolocal
utctolocal(datearg, timezonearg)
interprets datearg (a date, without
timezone) as utc and convert it to the
timezone named by timezonearg (a
string)
localtoutc
localtoutc(datearg, timezonearg) converts the local datetime datearg to
the timezone specified by the string
timezonearg, return as a date
weekday
weekday(date) returns the weekday as an integer in the
range 0..6, 0 is monday.
now
now() returns the current date and time
(localtime of the server timezone) as
date
daysbetween daysbetween(date1, date2)
daysbetween(daydate1, daydate2)
daysbetween(seconddate1,
seconddate2)
daysbetween(longdate1, longdate2)
returns the number of days (integer)
between date1 and date2. The first
version is an alternative to date2 - date1.
Instead of rounding or checking for
exactly 24 hours distance, this will
truncate both date values today
174
PUBLI C

precision and subtract the resulting day
numbers, meaning that if arg2 is not the
calendar day following arg1,
daysbetween will return 1 regardless of
the time components of arg1 and arg2.
secondsbetween secondsbetween(seconddate1,
seconddate2)
secondsbetween(longdate1, longdate2)
returns the number of seconds the first
to the second arg, as a fixed point
number. The returned value is positive if
the first argument is less than the
second. The return values are fixed18.0
in both cases (note that it may prove
more useful to use fixed11.7 in case of
longdate arguments).
component
component(date, int) the int argument may be int the range
1..6, the values mean year, day, month,
hour, minute, second, respectively. If a
component is not set in the date, the
component function will return a default
value, 1 for the month or the day, 0 for
other components. The component
function may also be applied to
longdate and time types.
addseconds addseconds(date, int)
addseconds(seconddate, decfloat)
addseconds(longdate, decfloat)
Return a date plus a number of seconds.
Fractional seconds will also be used in
case of longdate. Null handling is (in
opposition to the default done with
adds) to return null if any argument is
null.
adddays adddays(date, int)
adddays(daydate, int)
adddays(seconddate, int)
adddays(longdate, int)
Return a date plus a number of days.
Null handling is (in opposition to the
default done with adds) to return null if
any argument is null.
Misc Functions
if
if(intarg, arg2, arg3) return arg2 if intarg is
considered true (not equal to
zero), else return arg3.
Currently, no shortcut
evaluation is implemented,
meaning that both arg2 and
arg3 are evaluated in any
case. This means you cannot
use if to avoid a divide by zero
if("NETWR"<=500000,'A',
if("NETWR"<=1000000,'B','C
') )
PUBLI C

error which has the side
effect of terminating
expression evaluation when it
occurs.
in
in(arg1, ...) return 1 (= true) if arg1 is
equal to any of the remaining
args, return 0 else
case
case(arg1, default) case(arg1,
cmp1, value1, cmp2,
value2, ..., default)
return value1 if arg1 == cmp1,
value2 if arg1 == cmp2 etc,
default if there no match
case("CATEGORY", 'A', 'LV',
'B', 'MV', 'HV')
isnull
isnull(arg1) return 1 (= true), if arg1 is set
to null and null checking is on
during Evaluator run
(EVALUATOR_MAY_RETURN
_NULL)
7.5.12 Searching Tables, Models and Column Views
You use this feature to search for tables, models and column views in a system. You can search these objects in
any system from the list of systems added in the Navigator view that has all services started, and is up and
running.
In the search results, for a matching object you can perform the following actions:
Tables - you can open the table definition and add the table in the analytic and attribute view editor.
Note
You can add a table only if the editor is open otherwise it results in error.
Models - you can open the models in the editor and in case of attribute view, you can add it in the view editor
of an analytic view.
Note
You can add a model only if the editor is open otherwise it results in error.
Column Views - you can only open and view the definition.
1. Enter the object that you want to search in the Modeler toolbar search field.
2. Select the system in which you want to search the object from the dropdown option next to the Search
button.
3. To execute the search, choose Search.
The matching objects are listed in the expanded results pane with three tab pages, Tables, Models, and
Column Views. You can select each to view the matching objects and corresponding actions that you can
perform on them, for example, Open and Add.
176
PUBLI C

7.5.13 Setting Keyboard Shortcuts
You can enable keyboard shortcuts for modeling actions such as, activate and validate.
The supported commands with the default keyboard shortcuts are as follows:
Command Binding When Category
Activate Ctrl+Shift+A Navigator Modeler Keys
Activate Ctrl+Shift+A In Windows Modeler Keys
Add Table/Model Ctrl+Shift+= In Windows Modeler Keys
Auto Arrange Ctrl+L In Windows Modeler Keys
Data Preview Ctlr+Shift+P Navigator Modeler Keys
Data Preview Ctlr+Shift+P In Windows Modeler Keys
Display XML Alt+D In Windows Modeler Keys
Find Ctrl+F Navigator Modeler Navigator
Fit to Window Ctrl+0 In Windows Modeler Keys
Move Element in Output
Pane (Direction: Down)
Ctrl+] In Windows Modeler Keys
Move Element in Output
Pane (Direction: Up)
Ctrl+[ In Windows Modeler Keys
Open Ctrl+O Navigator Modeler Keys
Show View (View: History) Alt+Shift+H In Windows Views
Show View (View: Job
Log)
Alt+Shift+L In Windows Views
Show View (View: Where-
Used List)
Alt+Shift+U In Windows Views
Validate Ctrl+Shift+V In Windows Modeler Keys
Validate Ctrl+Shift+V Navigator Modeler Keys
Zoom (Type: In) Ctrl+= In Windows Modeler Keys
Zoom (Type: Out) Ctrl+- In Windows Modeler Keys
Zoom (Type: Reset) Alt+Shift+0 In Windows Modeler Keys
1. Choose Window Preferences General Keys .
2. In the Keys panel, choose Modeler as a Scheme.
3. If you only want to see the keyboard shortcuts for the Modeler , enter Modeler Keys in the text field.
Note
You cannot add new commands, but you can customize the commands as follows:
PUBLI C

Copy Command - to provide a different keyboard shortcut for an existing command.
Unbind Command - to clear the key bindings with the command and provide a new keyboard shortcut
for an existing command.
Restore Command - to restore the default key bindings provided by the Modeler for an existing
command.
4. Choose Apply.
7.5.14 Copying an Object
You can choose to copy an object in the Navigator view and paste it to a required package. You must have write
permissions on the target package where you are pasting the object. The copy paste feature is supported for all
Modeler objects that is, attribute view, analytic view, calculation view, procedure and analytic privilege. The object
that is copied to the target package is always inactive, even if in the source package it is in active state.
By default the keyboard shortcut for copy and paste is CTRL + C and CTRL + V respectively. To enable keyboard
shortcut for copy and paste you must apply the Modeler keyboard shortcuts from the Window Preferences
General Keys and select Modeler as scheme.
Note
Copy paste is supported only for a single object.
1. In the Navigator view, select an object and in the contect menu, choose Copy.
Note
If you have applied the Keyboard shortcuts then you can also press CTRL + C to copy an object.
2. Navigate to the package where you want to paste the object, and choose Paste.
Note
If you have applied the Keyboard shortcuts then you can also press CTRL + V to paste an object.
7.6 Importing BW Objects
You have implemented SAP Notes 1703061, 1759172, 1752384, 1733519, 1769374, and 1790333.
You have upgraded your Modeler to SAP HANA 1.0 SP05 Revision 58.
You have added BW schema in the SQL privileges for the Modeler user to import BW models.
_SYS_REPO user has SELECT with GRANT privileges on the schema that contains the BW tables.
You use this feature to import SAP NetWeaver Business Warehouse (SAP NetWeaver BW) models that are SAP
HANA-optimized InfoCubes, SAP HANA-optimized DataStore objects, Standard DataStore Objects, and Query
178
PUBLI C

Snapshot InfoProviders to the SAP HANA modeling environment. These imported objects are exposed as SAP
HANA information models and can be consumed for reporting using client tools such as, SAP BusinessObjects
Explorer, SAP BusinessObjects BI 4.0 Suite (Web Intelligence via Universes, Dashboards, Crystal Reports),
Microsoft Office and so on. The model properties are set based on the SAP NetWeaver BW models metadata.
Note
You can only import those Standard DataStore objects that have SID Generation set to During Activation.
If you select a DataStore object, the resultant SAP HANA model is an analytic view with the same name as that of
the DataStore object. If you select an InfoCube, two objects- analytic view and calculation view are created. In this
case, the name of calculation view and analytic view is same as that of the InfoCube and the name of the analytic
view is suffixed with _INTERNAL. The analytic view generated in the case of an InfoCube is used internally for the
generation of the respective calculation view and is not available for client consumption. If you select a
QuerySnapshot InfoProvider, the resultant SAP HANA model is an analytic view.
Conversion of BW analysis authorizations to analytic privileges
When you import the SAP NetWeaver BW InfoProviders, you can choose to import the relevant BW analysis
authorizations. The analysis authorizations are imported as analytic privileges in the SAP HANA Modeler. These
analysis authorizations could be associated with the InfoProviders or roles. You can import the analysis
authorizations in the following way:
You can choose to import only InfoProvider specific analysis authorizations. In this case, for all the
authorization objects specific to the InfoProvider having 0CTAIPROV = <InfoProvider name>
corresponding analytic privileges are generated. The name of the analytic privilege is the same as that of
the BW analysis authorization object.
You can choose to import analysis authorizations associated with the BW roles for the InfoProviders. In
this case, all the analysis authorizations assigned to the selected roles are merged as one or more
analytic privileges. The name of the generated analytic privilege is <InfoProvider
name>_BWROLE_<number>, such as, MyCube_BWROLE_1.
These analysis authorizations set on the InfoProviders are applicable at runtime for reporting. For example,
consider that a user has the following authorizations in BW:
Table 8: AO1
0CUSTOMER 1000 - 2000
0PRODUCT ABC*
Table 9: AO2
0CTAIPROV CUBE1, CUBE2
0CUSTOMER 3000 - 4000
0CTAACTVT 03 (display)
Now, if you choose to import only InfoProvider specific authorization, in the SAP HANA side, user would only
be able to see 0CUSTOMER from 3000 to 4000.
However, if you choose to import also role based authorizations, in the SAP HANA side, user would be able to
see 0CUSTOMER from 1000 to 4000, and 0PRODUCT = ABC*.
Note
PUBLI C

In the case of Query Snapshot, all the BW Analysis Authorization objects which are applicable on the
underlying InfoProvider of the query, will also be applicable on the Query Snapshot.
These BW analysis authorization objects will be imported as analytic privileges when importing the query
snapshot.
You can choose to place the generated models and analytic privileges in any of the user-defined packages in the
import wizard and enhance the generated models. However, with the subsequent import of the same objects, the
changes are overridden. Also, changes made to the models on BW side are not automatically reflected in the
generated models. This may lead to inconsistent generated models based on the changes made to the physical
tables. To avoid this, you need to re-import the models.
Caution
The calculated key figures (CKFs) and restricted key figures (RKFs) defined on the SAP BW models are not
created for the generated SAP HANA models. In this case, you can create an RKF as restricted measure in
the generated analytic view. For CKF you can create calculated measures in the generated calculation view
or analytic view. These CKFs and RKFs are retained during subsequent import. Additionally, the calculated
attributes created on the generated analytic views (in case of InfoCubes and DSOs) are also retained
during subsequent import. If a change is made to the characteristics or key figures based on which these
restricted measures and calculated measures are created, this may lead to inconsistency in the generated
models. In this case, you need to manually adjust these restricted measures and calculated measures.
The restricted measures and calculated measures that you define for the analytic view that correspond to a
query snapshot, is overwritten with the subsequent import.
The BW analysis authorization objects are not always mapped 1:1 with the generated analytic privileges on
the SAP HANA Modeler side. If the BW Analysis Authorization object does not include 0TCAIPROV, the
authorization is not moved to SAP HANA. Also, restrictions created in the BW analysis authorization are
skipped if they do not match with the restrictions supported by the SAP HANA Modeler. In such cases, the
data available for reporting for a SAP HANA Modeler user differs from the SAP NetWeaver BW user with
the assigned restrictions.
For reporting purposes, data which is visible for a user is:
For a DSO generated analytic view, all the data in the active table is available for reporting.
For an InfoCube generated calculation view, only successfully loaded requests are available for reporting
(these are the green requests in manage InfoCube section).
Restriction
The following features are not supported on the generated SAP HANA models:
DSO without any key figure
Currency and unit of measure conversion
Note
Only currency mapping is supported and not the conversion.
Time dependent text and attributes
Non-cumulative key figures
Conversion routines in the BW system
Hierarchies
180
PUBLI C

The following features are not supported on generated analytic privileges:
Exclude operator
Aggregated value operator :
Variables, User exits
Authorization on Key Figures
Authorization on hierarchy node
Exception aggregation such as, average, counter, first value, last value, no aggregation, standard
deviation is not supported for generated measures.
The query name for the Query Snapshot should not be the same as the BW InfoProvider name (this results
in conflict on the SAP HANA side).
Query Snapshot InfoProvider for BOE supports only key figures with aggregation types MIN, MAX, SUM,
and COUNT.
1. Open the Modeler perspective.
2. In the main menu, choose File Import .
3. Expand the SAP HANA Content node.
4. Choose Import SAP NetWeaver BW Models, and choose Next.
5. To establish a connection with the SAP NetWeaver BW system (underlying BW Application Server), in the
Source System page, enter the SAP NetWeaver BW system credentials and choose Next.
Note
To add new connection details, select New Connection option from the Connection dropdown list. The
connection details are saved and are available as dropdown options on subsequent logons.
6. Select the target system (an SAP NetWeaver BW on SAP HANA) to which you want to import the models, and
choose Next.
7. Select the BW InfoProviders that you want to import and expose as SAP HANA information models.
Remember
In order to import the QuerySnapshot InfoProvider, make sure that the BW Query is unlocked in
transaction RSDDB, and an index is created via the same transaction before it can be used as
InfoProviders.
8. Select the target package where you want to place the generated models, and analytic privileges.
Note
Your package selection is saved during the subsequent import. Hence, the next time you visit the same
wizard you get to view the package that was selected previous time. You can though change the package
where you want to import objects.
9. If you want import the selected models along with the display attributes for IMO Cube and IMO DSO, select
Include display attributes.
10. If you want to replace previously imported models in the target system with a new version, select Overwrite
existing models.
11. If you do not want to import the analysis authorizations associated with the selected InfoProviders, deselect
Generate InfoProvider based analytic privileges.
PUBLI C

12. If you want to import the role based analysis authorizations as analytic privileges, select Generate Role based
analytic privileges, and choose Next.
13. Select the roles to import the related analysis authorizations.
14. Choose Finish.
The generated information models and analytic privileges are placed in the package selected above. In order to
view the data of generated models, you need to assign the associated analytic privileges that are generated as
part of the model import to the user. If these privileges are not assigned, user is not authorized to view the data.
182
PUBLI C

8 Developing Procedures
SQL in SAP HANA includes extensions for creating procedures, which enables you to embed data-intensive
application logic into the database, where it can be optimized for performance (since there are no large data
transfers to the application and features such as parallel execution is possible). Procedures are used when other
modeling objects, such as analytic or attribute views, are not sufficient.
Some of the reasons to use procedures instead of standard SQL:
SQL is not designed for complex calculations, such as for financials.
SQL does not provide for imperative logic.
Complex SQL statements can be hard to understand and maintain.
SQL queries return one result set. Procedures can return multiple result sets.
Procedures can have local variables, eliminating the need to explicitly create temporary tables for
intermediate results.
Procedures can be written in the following languages:
SQLScript: The language that SAP HANA provides for writing procedures.
R: An open-source programming language for statistical computing and graphics, which can be installed and
integrated with SAP HANA.
There are additional libraries of procedures, called Business Function Library and Predictive Analysis Library, that
can be called via SQL or from within another procedure.
SQL Extensions for Procedures
SQL includes the following statements for enabling procedures:
CREATE TYPE: Creates a table types, which are used to define parameters for a procedure that represent
tabular results. For example:
CREATE TYPE tt_publishers AS TABLE (
publisher INTEGER,
name VARCHAR(50),
price DECIMAL,
cnt INTEGER);
CREATE PROCEDURE: Creates a procedure. The LANGUAGE clause specifies the language you are using to
code the procedure. For example:
CREATE PROCEDURE ProcWithResultView(IN id INT, OUT o1 CUSTOMER)
LANGUAGE SQLSCRIPT READS SQL DATA WITH RESULT VIEW ProcView AS
BEGIN
o1 = SELECT * FROM CUSTOMER WHERE CUST_ID = :id;
END;
CALL: Calls a procedure. For example:
CALL getOutput (1000, 'EUR', NULL, NULL);
Developing Procedures
PUBLI C

Tools for Writing Procedures
Use the SQLScript editor, which includes debugging capabilities, to build SQLScript procedures.
You can also use the Navigator view in the Modeler perspective to build procedures, but there are no debugging
capabilities. You should only use this method:
If you need to develop a procedure using a local table type as an input or output parameter. A local table type
is created within the SAP HANA Systems procedure tool and for only the current procedure. If you can use a
global table type, then use the SQLScript Editor.
If you need to edit a procedure previously created in the Navigator view that contains table type parameters.
Related Links
SAP HANA SQL Reference
SAP HANA SQLScript Reference
SAP HANA R Integration Guide
SAP HANA Business Function Library (BFL) Reference
SAP HANA Predictive Analysis Library (PAL) Reference
Editing SQLScript [page 184]
The SAP HANA SQLScript editor allows you to create, edit and activate procedures.
8.1 Editing SQLScript
The SAP HANA SQLScript editor allows you to create, edit and activate procedures.
Before you begin working in the SAP HANA SQLScript editor, open the SAP HANA Development perspective and
do the following:
Create a development workspace. For more information, see Creating a Repository Workspace [page 43].
Checkout a package. For more information, see Working with the Repository [page 30].
Note
After checking out a package that contains active procedures, you can modify and debug the procedures.
Create and share a project. For more information, see Using SAP HANA Projects [page 42].
.
Note
You can also share your project after you create your procedure.
To write and edit a procedure in the SAP HANA SQLScript editor, perform the following steps:
1. After you have created your workspace and your project, go to the Project Explorer view in the SAP HANA
Development perspective, right-click on the file name, select New, and select File. The New File wizard will
appear.
184
PUBLI C

2. Enter or select the parent folder and enter the file name using the following naming convention
<filename>.procedure. Choose Finish. The icon shows that your procedure is created locally. Choose
Save.
Your procedure will open containing the default Create Procedure template. In the Properties view, you will
see the properties of your procedure, such as Access Mode, Name and Language. You can also change the
default schema that this procedure is using.
Note
You can also create a folder first and add a file. Right-click on the project name, select New, and select
Folder. The New Folder wizard will appear. Enter or select the project, enter the folder name, and choose
Finish.
3. To share your project, right-click on the project name, select Team, and select Share Project. The Share
Project wizard will appear. Choose Finish. The icon shows that your procedure is not committed and not
activated.
4. You can begin writing your code inside your new procedure and save it locally. The syntax is checked
simultaneously and is highlighted. Auto-completion of the syntax appears as you type; press Ctrl + Spacebar
to get a list of relevant SQLScript statements.
Note
You can only write one stored procedure per file. The file name and the procedure name must be the same.
5. To commit your new procedure or make changes to an existing one, first save it, right-click on the procedure,
select Team, and select Commit. Your procedure is now synchronized to the repository as a design-time
object and the icon shows that your procedure is committed.
Caution
The design-time presentation of the procedure is currently in XML format that you must not edit.
6. When you have finished writing your procedure and you are ready to activate it, right-click on the procedure,
select Team, and select Activate. Your procedure is created in the catalog as a runtime object and the icon
shows that your procedure is activated . This will allow you and other users to call the procedure and debug it.
If an error is detected during activation, an error message will appear in the Problems view.
Note
You can also activate your procedure at the project and folder level.
Related Links
SAP HANA Development Perspective [page 28]
SAP HANA Repository Packages and Namespaces [page 48]
In SAP HANA, a package typically consists of a collection of repository objects, which can be transported between
systems. Multiple packages can be combined in a delivery unit (DU).
About SAP HANA SQLScript
Defining Local Table Types in Procedures [page 186]
PUBLI C

You can use table types to define parameters for a procedure that represent tabular results. These parameters
have a type and are either based on a global table (with a reference to a catalog table) or a local table type.
http://help.sap.com/hana/html/sqlmain.html
http://help.sap.com/hana/html/monitor_views.html
8.1.1 Defining Local Table Types in Procedures
You can use table types to define parameters for a procedure that represent tabular results. These parameters
have a type and are either based on a global table (with a reference to a catalog table) or a local table type.
Before you define local table types in the SAP HANA SQLScript editor, you must create or open a procedure. For
more information, see Editing SQLScript.
To define local table types in a procedure in the SAP HANA SQLScript editor, perform the following steps:
1. Choose the Local Table Types tab.
2. Define your local table type structure using a standard SQL create statement. The local table type is specified
using a list of attribute names and primitive data types. For example:
CREATE TYPE <type_name> AS TABLE (<column_definition>[{,<column_definition>}...])
Note
You can create multiple CREATE TYPE statements.
Caution
You can only use this local table type in the procedure you defined them in.
3. Use the local table table type as input and output parameters of the procedure, for example:
CREATE PROCEDURE <procedure_name> ( IN|OUT|INOUT <param_name> <type_name>, ... )
Caution
You can only use this tab to define local table types and not for other SQL statements.
4. Click the Save button. Commit and activate your procedure to create a local table types in the catalog. For
more information about committing and activating a procedure, see Editing SQLScript.
Note
The local table types are bound to the procedure artifact, so if the procedure is committed, activated, or
deleted, then the same applies to the local table type. For example, if you delete a procedure the local table
type will be automatically deleted (similar to a drop statement) from the catalog.
Related Links
CREATE TYPE
Table Types
186
PUBLI C

8.2 Debugging SQLScript
The SAP HANA SQLScript debugger allows you to debug and analyze procedures. In a debug session, your
procedures will be executed in a serial mode, not in parallel (not-optimized). This allows you to test the
correctness of the procedure logic and is not for evaluating the performance.
Before you begin using the SAP HANA SQLScript debugger, do the following:
Activate your procedures and they must belong to a project on your local work station. For more information,
see Editing SQLScript [page 184].
Grant debugger privileges to your user:
Note
Contact your System Administrator if you are not authorized to modify your user.
1. Go to the Navigator view in the SAP HANA Development perspective and open Security Users .
2. Double-click your user ID. Your system privilege information will appear. Choose the SQL Privileges tab.
3. Select the _SYS_BIC schema and select all of the privileges.
4. Select the DEBUG (SYS) procedure and select the EXECUTE privilege.
5. Choose the Deploy button (F8).
To debug a procedure in the SAP HANA SQLScript debugger, perform the following steps:
1. Open the Debug perspective in the SAP HANA studio and select the procedure you want to debug by clicking
on the relevant tab in the Editor view.
2. Double-click on the left vertical ruler to add breakpoints to your procedure. You can see a list of all of the
breakpoints in the Breakpoints view.
From the Breakpoints view, you can:
Deselect specific breakpoints or skip all of them.
Delete a specific breakpoint or delete all.
Double-click on a breakpoint to see which line it belongs to in the Editor view.
See the status of the breakpoint:
Pending
Valid
PUBLI C

Invalid
3. To create a debug session you must first create a debug configuration. Choose and select Debug
Configurations.... The Debug Configurations wizard will appear.
Note
The current debug configuration allows you to debug any stored procedure that is opened in the editor.
Therefore, you do not have to create a new debug configuration for every procedure you want to debug.
4. Double-click SQLScript Procedure, enter a name, choose Apply, and choose Close.
Note
If you want to start debugging your procedure, choose Apply and choose Debug.
5. To start your debug session, choose and select your debug configuration. The debug session will begin
and you will see the status of the session in the Debug view. The status of a breakpoint may change after the
server has validated it. The breakpoint position might change if it is placed on an invalid line where the
debugger cannot stop. The debugger will stop at the first breakpoint and the session will be suspended until
you resume it.
If your run is successful, the valid status will appear next to the breakpoints in the Breakpoints view.
Note
You must set breakpoints in the lines you want to break at and resume the session again.
You can evaluate your local scalar and table variables in the Variable view. The view shows the values of the
scalar variables and the number of rows in each table.
6. To view the content of the tables listed in the Variable view, right-click on the table name and select Open Data
Preview. The results will appear in the Preview view. This view will automatically close when you resume your
debug session.
The session will be terminated when you reach the end of the procedure run.
8.3 Developing Procedures in the Modeler Editor
To create procedures, use the SQLScript Editor, as described in Editing SQLScript [page 184].
If you need to create procedures with local table types, that is, table types created only for the procedure, perform
the steps described in this section.
1. On the Quick Launch tab page, choose Procedure.
If the Quick Launch page is not open, go to Help Quick Launch .
2. Enter a name and description for the procedure.
3. For unqualified access in SQL, select the required schema from the Default Schema dropdown list.
Note
188
PUBLI C

If you do not select a default schema, while scripting you need to provide fully qualified names of the
catalog objects that include the schema.
If you specify a default schema, and write SQL such as select * from myTable, the specified
default schema is used at runtime to refer to the table.
4. Select the package in which you want to save the procedure.
5. Select the required option from the Run With dropdown list to select which privileges are to be considered
while executing the procedure.
Note
There are two types of rights, as follows:
Definer's right: If you want the system to use the rights of the definer while executing the procedure for any
user.
Invoker's right: If you want the system to use the rights of the current user while executing the procedure.
6. Select the required access mode as follows:
Access Mode Purpose
Read Only Use this mode to create procedures for fetching table data.
Read Write Use this mode to create procedures for fetching and
updating table data.
7. Select the language in which you are writing the procedure.
Note
You can choose to create procedures in Read Write mode and make use of L- Lang and R-lang languages
only if you have done the repository configuration for the field sqlscript_mode . Two values for
sqlscript_mode field exist, DEFAULT, and UNSECURE. By default DEFAULT is assigned which means Read
Only mode with non-modifiable access mode and SQLScript as language. To change the configuration, go
to administration console -> Configuration tab -> indexserver.ini -> repository -> sqlscript_mode, and
assign the required value.
8. Choose Finish.
9. In the function editor pane, write a script for the function using the following data types:
Table or scalar data types for input parameters.
Table data types for output parameters.
Note
You can only write one function in the function body. However, you can refer to other functions.
10. Choose File Save .
11. Activate the procedure using one of the following options in the toolbar:
Save and Activate: Activate the current procedure and redeploy the affected objects if an active version of
the affected object exists. Otherwise only the current procedure gets activated.
Save and Activate All: Activate the current procedure along with the required and affected objects.
Note
PUBLI C

You can also activate the current procedure by selecting the procedure in the Navigator view and choosing
Activate in the context menu. For more information about activation, see Activating Objects [page 153].
190
PUBLI C

9 Defining Web-based Data Access
SAP HANA extended application services (SAP HANA XS) provide applications and application developers with
access to the SAP HANA database using a consumption model that is exposed via HTTP.
In addition to providing application-specific consumption models, SAP HANA XS also host system services that
are part of the SAP HANA database, for example: search services and a built-in Web server that provides access
to static content stored in the SAP HANA repository.
The consumption model provided by SAP HANA XS focuses on server-side applications written in JavaScript and
making use of a powerful set of specially developed API functions. However, you can use other methods to
provide access to the data you want to expose in SAP HANA. For example, you can set up an ODATA service or
use the XML for Analysis (XMLA) interface to send a Multi-dimensional Expressions (MDX) query. This section
describes how to set up a service that enables you to expose data using OData or XMLA.
9.1 Data Access with OData in SAP HANA XS
In SAP HANA Extended Application Services (SAP HANA XS), the persistence model (for example, tables, views,
and stored procedures) is mapped to the consumption model that is exposed to clients - the applications you
write to extract data from the SAP HANA database.
You can map the persistence and consumption models with OData, a resource-based Web protocol for querying
and updating data. An OData application running in SAP HANA XS is used to provide the consumption model for
client applications exchanging OData queries with the SAP HANA database.
You can use OData to enable clients to consume authorized data stored in the SAP HANA database. OData
defines operations on resources using RESTful HTTP commands (for example, GET, PUT, POST, and DELETE) and
specifies the URI syntax for identifying the resources. Data is transferred over HTTP using either the Atom (XML)
or the JSON (JavaScript) format.
Applications running in SAP HANA XS enable accurate control of the flow of data between the presentational
layer, for example, in the Browser, and the data-processing layer in SAP HANA itself, where the calculations are
performed, for example, in SQL or SQLScript. If you develop and deploy an OData service running in SAP HANA
XS, you can take advantage of the embedded access to SAP HANA that SAP HANA XS provides; the embedded
access greatly improves end-to-end performance.
9.1.1 OData in SAP HANA XS
OData is a resource-based web protocol for querying and updating data. OData defines operations on resources
using HTTP commands (for example, GET, PUT, POST, and DELETE) and specifies the uniform resource indicator
(URI) syntax to use to identify the resources.
Data is transferred over HTTP using the Atom or JSON format:
Note
Defining Web-based Data Access
PUBLI C

OData makes it easier for SAP, for partners, and for customers to build standards-based applications for many
different devices and on various platforms, for example, applications that are based on a lightweight
consumption of SAP and non-SAP business application data.
The main aim of OData is to define an abstract data model and a protocol which, combined, enable any client to
access data exposed by any data source. Clients might include Web browsers, mobile devices, business-
intelligence tools, and custom applications (for example, written in programming languages such as PHP or Java);
data sources can include databases, content-management systems, the Cloud, or custom applications (for
example, written in Java).
The OData approach to data exchange involves the following elements:
OData data model
Provides a generic way to organize and describe data. OData uses the Entity 1 Data Model (EDM).
OData protocol
Enables a client to query an OData service. The OData protocol is a set of interactions, which includes the
usual REST-based create, read, update, and delete operations along with an OData-defined query language.
The OData service sends data in either of the following ways:
XML-based format defined by Atom/AtomPub
JavaScript Object Notation (JSON)
OData client libraries
Enables access to data via the OData protocol. Since most OData clients are applications, pre-built libraries
for making OData requests and getting results reduces and simplifies work for the developers who create
those applications.
A broad selection of OData client libraries are already widely available, for example: Android, Java, JavaScript,
PHP, Ruby, and the best known mobile platforms.
OData services
Exposes an end point that allows access to data in the SAP HANA database. The OData service implements
the OData protocol (using the OData Data Services runtime) and uses the Data Access layer to map data
between its underlying form (database tables, spreadsheet lists, and so on) and a format that the requesting
client can understand.
9.1.2 Defining the Data an OData Service Exposes
An OData service exposes data stored in database tables or views as OData collections for analysis and display by
client applications. However, first of all, you need to ensure that the tables and views to expose as an OData
collection actually exist.
To define the data to expose using an OData service, you must perform at least the following tasks:
1. Create a database schema.
2. Create a simple database table to expose with an OData service.
3. Create a simple database view to expose with an OData service.
This step is optional; you can expose tables directly. In addition, you can create a modeling view, for example,
analytic, attribute, or calculation.
4. Grant select privileges to the tables and views to be exposed with the OData service.
192
PUBLI C

After activation in the repository, schema and tables objects are only visible in the catalog to the _SYS_REPO
user. To enable other users, for example the schema owner, to view the newly created schema in the SAP
HANA studio's Modeler perspective, you must grant the user the required SELECT privilege.
call
_SYS_REPO.GRANT_SCHEMA_PRIVILEGE_ON_ACTIVATED_CONTENT('select','<SCHEMANAME>','<u
sername>');
9.1.3 OData Service Definitions
The OData service definition is the mechanism you use to define what data to expose with OData, how, and to
whom. Data exposed as an OData collection is available for analysis and display by client applications, for
example, a browser that uses functions provided by an OData client library running on the client system.
To expose information by means of OData to applications using SAP HANA XS, you must define database views
that provide the data with the required granularity. Then you create an OData service definition, which is a file you
use to specify which database views or tables are exposed as OData collections.
Note
SAP HANA XS supports OData version 2.0, which you can use to send OData queries (for example, using the
HTTP GET method). Language encoding is restricted to UTF-8.
An OData service for SAP HANA XS is defined in a text file with the file suffix .xsodata, for example,
OdataSrvDef.xsodata. The file must contain at least the entry service {}, which would generate a
completely operational OData service with an empty service catalog and an empty metadata file. However, usually
you use the service definition to expose objects in the database catalog, for example: tables, SQL views, or
calculation rules.
In the OData service-definition file, you can use the following ways to name the SAP HANA objects you want to
expose by OData:
Note
The syntax to use in the OData service-definition file to reference objects depends on the object type, for
example, repository (design-time) or database catalog (runtime).
Repository objects
Expose an object using the object's repository (design-time) name in the OData service-definition file. This
method of exposing database objects using OData enables the OData service to be automatically updated if
the underlying repository object changes. Note that a design-time name can be used to reference analytic and
calculation views; it cannot be used to reference SQL views. The following example shows how to include a
reference to a table in an OData service definition using the table's design-time name.
service {
"acme.com.odata::myTable" as "myTable"
}
Note
PUBLI C

Calculation views are only accessible from within xsodata files by referring to the design-time name.
However, it is recommended to use design-time names whenever possible for calculation views or
common tables. With design-time names, the cross references are recreated during activation (for
example, for where-used), which means changes are visible automatically.
Database objects
Expose an object using the object's database catalog (runtime) name. The support for database objects is
mainly intended for existing or replicated objects that do not have a repository design-time representation.
The following example shows how to include a reference to a table in an OData service definition using the
table's runtime name.
service {
"mySchema"."myTable" as "MyTable"
}
Note
It is strongly recommended not to use catalog (runtime) names in an OData service-definition. The use of
catalog object names is only enabled in a service-definition because some objects do not have a design-
time name. If at all possible, use the design-time name to reference objects in an OData service-definition
file.
9.1.3.1 OData Service-Definition Type Mapping
During the activation of the OData service definition, SQL types defined in the service definition are mapped to
EDM types according to a mapping table.
For example, the SQL type "Time" is mapped to the EDM type "EDM.Time"; the SQL type "Decimal" is mapped to
the EDM type "EDM.Decimal"; the SQL types "Real" and "Float" are mapped to the EDM type "EDM.Single".
Note
The OData implementation in SAP HANA Extended Application Services (SAP HANA XS) does not support all
SQL types.
In the following example, the SQL types of columns in a table are mapped to the EDM types in the properties of an
entity type.
{name = "ID"; sqlType = INTEGER; nullable = false;}, {name = "RefereeID"; sqlType =
VARCHAR; nullable = true;}
<Property Name="ID" Type="Edm.Int32" Nullable="false"/> <Property Name="RefereeID"
Type="Edm.String" Nullable="true"/>
Related Links
OData Service Definition: SQL-EDM Type Mapping [page 214]
During the activation of the OData service definition, the SAP HANA SQL types are mapped to the required OData
EDM types according to the rules specified in a mapping table.
OData Service Definitions [page 193]
194
PUBLI C

9.1.3.2 OData Service-Definition Features
The OData service definition provides a list of keywords that you use in the OData service-definition file to enable
important features. For example, the following list illustrates the most-commonly used features used in an OData
service-definition and, where appropriate, indicates the keyword to use to enable the feature:
Aggregation
The results of aggregations on columns change dynamically, depending on the grouping conditions. As a
result, aggregation cannot be done in SQL views; it needs to be specified in the OData service definition itself.
Depending on the type of object you want to expose with OData, the columns to aggregate and the function
used must be specified explicitly (explicit aggregation) or derived from metadata in the database (derived
aggregation). Note that aggregated columns cannot be used in combination with the $filter query
parameter, and aggregation is only possible with generated keys.
Association
Define associations between entities to express relationships between entities. With associations it is possible
to reflect foreign key constraints on database tables, hierarchies and other relations between database
objects.
Key Specification
The OData specification requires an EntityType to denote a set of properties forming a unique key. In SAP
HANA, only tables can have a unique key, the primary key. All other (mostly view) objects require you to
specify a key for the entity. The OData service definition language (OSDL) enables you to do this by denoting a
set of existing columns or by generating a local key. Bear in mind that local keys are transient; they exist only
for the duration of the current session and cannot be dereferenced.
Note
OSDL is the language used to define a service definition; the language includes a list of keywords that you
use in the OData service-definition file to enable the required features.
Parameter Entity Sets
You can use a special parameter entity set to enter input parameters for SAP HANA calculation views and
analytic views. During activation of the entity set, the specified parameters are retrieved from the metadata of
the calculation (or analytical) view and exposed as a new EntitySet with the name suffix "Parameters", for
example "CalcViewParameters".
Projection
If the object you want to expose with an OData service has more columns than you actually want to expose,
you can use SQL views to restrict the number of selected columns in the SELECT. However, for those cases
where SQL views are not appropriate, you can use the with or without keywords in the OData service
definition to include or exclude a list of columns.
Related Links
OData Service-Definition Examples [page 199]
The OData service definition describes how data exposed in an end point can be accessed by clients using the
OData protocol.
PUBLI C

9.1.4 Creating an OData Service Definition
The OData service definition is a configuration file you use to specify which data (for example, views or tables) is
exposed as an OData collection for analysis and display by client applications.
An OData service for SAP HANA XS is defined in a text file with the file suffix .xsodata, for example,
OdataSrvDef.xsodata. The file resides in the package hierarchy of the OData application and must contain at
least the entry service {}, which would generate an operational OData service with an empty service catalog and
an empty metadata file.
Prerequisites for the creation of an OData service definition:
SAP HANA studio/client installed and configured
SAP HANA database user available with repository privileges (for example, to add packages)
A SAP HANA development system added to (and available in) SAP HANA studio, for example, in either the
Navigator view or the SAP HANA Repositories view
A working development environment including a repository workspace, a package structure for your OData
application, and a shared project to enable you to synchronize changes to the OData project files in the local
file system with the repository
Defined data to expose with the OData application, for example, at least the following:
A database schema
A database table
1. In the shared project you are using for your OData application, use the Project Explorer view to locate the
package where you want to create the new OData service definition.
Note
The file containing the OData service definition must be placed in the root package of the OData application
for which the service is intended.
2. Create the file that will contain your OData service definition.
In the Project Explorer view, right-click the folder where you want to create the new OData service-definition
file and choose New File in the context-sensitive popup menu.
3. Define the OData service.
The OData service definition uses the OData Service Definition Language (OSDL), which includes a list of
keywords that you specify in the OData service-definition file to enable important features.
The following example shows a simple OData service definition exposing a simple table:
service namespace "my.namespace" {
"sample.odata::table" as "MyTable";
}
This service definition exposes a table defined in the file sample.odata:table.hdbtable and creates an
EntitySet for this entity named MyTable. The specification of an alias is optional. If omitted, the default name
of the EntitySet is the name of the repository object file, in this example, table.
4. Place the valid OData service definition in the root package of the OData application to which it applies.
5. Save, commit, and activate the OData service definition in the SAP HANA repository.
Related Links
OData Service Definitions [page 193]
196
PUBLI C

9.1.5 Tutorial: Using the SAP HANA OData Interface
The package you put together to test the SAP HANA OData interface includes all the artifacts you need to use SAP
HANA Extended Application Services (SAP HANA XS) to expose an OData collection for analysis and display by
client applications.
Since the artifacts required to get a simple OData application up and running are stored in the repository, it is
assumed that you have already performed the following tasks:
Create a development workspace in the SAP HANA repository
Create a project in the workspace
Share the new project
To create a simple OData application, perform the following steps:
1. Create a root package for your OData application, for example, helloodata and save and activate it in the
repository.
Note
2. Create a schema, for example, HELLO_ODATA.hdbschema.
The schema is required for the table that contains the data to be exposed by your OData service-definition.
The schema is defined in a flat file with the file extension .hdbschema that you save in the repository and
which you must activate.
Enter the following code in the HELLO_ODATA.hdbschema file:
schema_name="HELLO_ODATA";
3. Create the database table that contains the data to be exposed by your OData service definition, for example,
otable.hdbtable.
The database table is a flat file with the file extension .hdbtable that you save in the repository and which
you must activate.
Enter the following code in the otable.hdbtable file:
table.schemaName = "HELLO_ODATA";
table.columns = [
"dummy comment";},
{name = "Col3"; sqlType = NVARCHAR; nullable = true; length = 20;
defaultValue = "Defaultvalue";},
3;}];
PUBLI C

4. Grant SELECT privileges to the owner of the new schema.
After activation in the repository, the schema object is only visible in the catalog to the _SYS_REPO user. To
enable other users, for example the schema owner, to view the newly created schema in the SAP HANA
studio's Modeler perspective, you must grant the user the required SELECT privilege.
a) In the SAP HANA studio Navigator view, right-click the SAP HANA system hosting the repository where
the schema was activated and choose SQL Console in the context-sensitive popup menu.
b) In the SQL Console, execute the statement illustrated in the following example, where <SCHEMANAME> is
the name of the newly activated schema, and <username> is the database user ID of the schema owner:
call
_SYS_REPO.GRANT_SCHEMA_PRIVILEGE_ON_ACTIVATED_CONTENT('select','<SCHEMANAME>',
'<username>');
5. Create an application descriptor for your new OData application in your root OData package helloodata.
The application descriptor (.xsapp) is the core file that you use to define an application's availability within
SAP HANA application. The .xsapp file sets the point in the application-package structure from which
content will be served to the requesting clients.
Note
The application-descriptor file has no content and no name; it only has the extension .xsapp.
6. Save, commit, and activate the application-descriptor file in the repository.
7. Create an application-access file for your new OData application and place it in your root OData package
helloodata.
the application.
Note
The application-access file has no name; it only has the extension .xsaccess.
Enter the following content in the .xsaccess file for your new OData application:
{
"exposed" : true
}
8. Save, commit, and activate the application-access file in the repository.
9. Create an OData service-definition file and place it in your root OData package helloodata.
The Odata service-definition file has the file extension .xsodata, for example, hello.xsodata and must be
located in the root package of the OData application:
Enter the following content in the hello.xsodata OData service-definition file:
service {
"helloodata::otable";
}
10. Save, commit, and activate the OData service-definition file in the repository.
11. Open a browser and enter the following URL.
http://<hana.server.name>:80<HANA_instance_number>/helloodata/hello.xsodata
198
PUBLI C

9.1.6 OData Service-Definition Examples
The OData service definition describes how data exposed in an end point can be accessed by clients using the
OData protocol.
Each of the examples listed below is explained in a separate section. The examples show how to use the OData
Service Definition Language (OSDL) in the OData service-definition file to generate an operational OData service
that enables clients to use SAP HANA XS to access the OData end point you set up.
Empty Service
Namespace Definition
Object Exposure
Property Projection
Key Specification
Associations
Aggregation
Parameter Entity Sets
9.1.6.1 OData Empty Service
An OData service for SAP HANA XS is defined by a text file containing at least the following line:
Service definition sample.odata:empty.xsodata
service {}
A service file with the minimal content generates an empty, completely operational OData service with an empty
service catalog and an empty metadata file:
http://localhost:8002/sample/odata/empty.xsodata
http://localhost:8002/sample/odata/empty.xsodata/$metadata
PUBLI C

An empty service metadata document consists of one Schema containing an empty EntityContainer. The
name of the EntityContainer is the name of the .xsodata file, in this example "empty".
9.1.6.2 OData Namespace Definition
By default, as shown in the metadata file of an empty OData service, the namespace of the generated Schema is
created by concatenating the package name of the .xsodata file with the file name, and separating the
concatenated names with a dot. You can specify your own namespace by using the namespace keyword:
Service definition sample.odata:namespace.xsodata
service namespace "my.namespace" {}
The resulting service metadata document has the specified schema namespace:
http://localhost:8002/sample/odata/namespace.xsodata/$metadata
9.1.6.3 OData Object Exposure
There are two ways of exposing SAP HANA objects. You can either specify the repository design-time name or the
database-object runtime name (with database schema). Although both variants are supported, the preferred
200
PUBLI C

method is the exposure via the repository design-time name. It has the advantage that the OData service is
automatically updated, if the underlying repository object changes. The support for database objects is mainly
intended for existing or replicated objects that do not have a related repository design-time object.
In the examples provided to illustrate object exposure, the following definition of a table applies:
Table definition sample.odata:table.hdbtable
table.schemaName = "ODATASAMPLES";
table.columns = [
{name = "ID"; sqlType = INTEGER;},
{name = "Text"; sqlType = NVARCHAR; length=1000;},
{name = "Time"; sqlType = TIMESTAMP;}
];
table.primaryKey.pkcolumns = ["ID"];
Repository Objects
If the object to expose via an OData service is created during an activation in the repository, then it has a
repository design-time representation. Examples for those objects are tables, SQL views and calculation views. An
example for exposing the table above is shown in the next service defintion.
Service definition sample.odata:repo.xsodata
service {
"sample.odata::table" as "MyTable";
}
This service definition exposes a table defined in the .hdbtable file sample.odata:table.hdbtable and
creates an EntitySet for this entity named "MyTable". The specification of an alias is optional. If omitted the
default name of the EntitySet is the name of the repository object file, here "table".
http://localhost:8002/sample/odata/repo.xsodata
http://localhost:8002/sample/odata/repo.xsodata/$metadata
PUBLI C

All information about the table, for example, the properties, the data types, and the primary key, is gathered from
the database catalog.
Database Objects
Caution
Do not use catalog objects if a repository design-time object is available, as changes in the catalog object are
not automatically reflected in the OData service.
Similar to the exposure of an object by using the repository design-time name is the exposure by the database
name:
Service definition sample.odata:db.xsodata
service {
"ODATASAMPLES"."sample.odata::table" as "MyTable";
}
The service exposes the same table by using the database catalog name of the object and the name of the
schema where the table is created in. The collection in the service catalog and the EntityType that is created in the
metadata document is exactly the same as for repository objects.
202
PUBLI C

9.1.6.4 OData Property Projection
If the object you want to expose with an OData service has more columns than you actually want to expose, you
can use SQL views to restrict the number of selected columns in the SELECT.
Nevertheless, SQL views are sometimes not appropriate, for example with calculation views, and for these cases
we provide the possibility to restrict the properties in the OData service definition in two ways. By providing an
including or an excluding list of columns.
Including Properties
You can specify the columns of an object that have to be exposed in the OData service by using the with keyword.
Key fields of tables must not be omitted.
Service definition sample.odata:with.xsodata
service {
"sample.odata::table" as "MyTable" with ("ID","Text");
}
The resulting EntityType then contains only the properties derived from the specified columns:
http://localhost:8002/sample/odata/with.xsodata/$metadata
Excluding Properties
The opposite of the with keyword is the without keyword, which enables you to specify which columns you do
NOT want to expose in the OData service:
Service definition sample.odata:without.xsodata
service {
"sample.odata::table" as "MyTable" without ("Text","Time");
}
The generated EntityType then does NOT contain the properties derived from the specified columns:
http://localhost:8002/sample/odata/without.xsodata/$metadata
PUBLI C

9.1.6.5 OData Key Specification
The OData specification requires an EntityType to denote a set properties forming a unique key. In HANA only
tables may have a unique key, the primary key. For all other (mostly view) objects you need to specify a key for the
entity.
In OSDL, you can specify a key for an entity by denoting a set of existing columns or by generating a key.
For the examples illustrating key specification, we use the following SQL view, which selects all data from the
specified table.
View definition sample.odata:view.hdbview
{
"name": "view",
"schema": "ODATASAMPLES",
"query": "SELECT * FROM \"ODATASAMPLES\".\"sample.odata::table\""
}
Existing Key Properties
If the object has set of columns that may form a unique key, you can specify them as key for the entity. These key
properties are always selected from the database, no matter if they are omitted in the $select query option.
Therefore explicit keys are not suitable for calculation views and analytic views as the selection has an impact on
the result.
Service definition sample.odata:explicitkeys.xsodata/$metadata
service {
"sample.odata::view" as "MyView" keys ("ID","Text");
}
The created metadata document for the exposure of the view above is almost equal to the metadata document for
repository objects. Only the key is different and consists now of two columns:
http://localhost:8002/sample/odata/explicitkeys.xsodata/$metadata
204
PUBLI C

Caution
The OData infrastructure cannot check whether your specified keys are unique. So be careful when choosing
keys.
Generated Local Key
For objects that do not have a unique key in their results, for example, calculation views or aggregated tables, you
can generate a locally valid key. This key value numbers the results starting with 1 and is not meant for
dereferencing the entity; you cannot use this key to retrieve the entity. The key is valid only for the duration of the
current session and is used only to satisfy OData's need for a unique ID in the results. The property type of a
generated local key is Edm.String and cannot be changed.
Service definition sample.odata:generatedkeys.xsodata
service {
"sample.odata::view" as "MyView" keys generate local "GenID";
}
http://localhost:8002/sample/odata/generatedkeys.xsodata/$metadata
PUBLI C

As a consequence of the transient nature of generated local keys, it is not possible to define navigation properties
on these entities or use them in filter or order by conditions.
9.1.6.6 OData Associations
You can define associations between entities to express relationships between entities. With associations it is
possible to reflect foreign key constraints on database tables, hierarchies and other relations between database
objects. OSDL supports simple associations, where the information about the relationship is stored in one of the
participating entities, and complex associations, where the relationship information is stored in a separate
association table.
Associations themselves are freestanding. On top of them you can specify which of the entities participating in the
relationship can navigate over the association to the other entity by creating NavigationPropertys.
For the examples used to illustrate OData associations, we use the tables customer and order:
Table definition: sample.odata:customer.hdbtable
table.columns = [
{name = "ID"; sqlType = INTEGER; nullable = false;},
{name = "RecruitID"; sqlType = VARCHAR; nullable = true;}
];
Table definition: sample.odata:order.hdbtable
table.columns = [
{name = "CustomerID"; sqlType = INTEGER; nullable = false;}
];
There is one relationship order.CustomerID to customer.ID and one relationship customer.RecruitID to
customer.ID.
206
PUBLI C

Simple Associations
The definition of an association requires you to specify a name, which references two exposed entities and whose
columns keep the relationship information. To distinguish the ends of the association, you must use the keywords
principal and dependent. In addition, it is necessary to denote the multiplicity for each end of the association.
Service definition: sample.odata:assocsimple.xsodata
service {
"sample.odata::customer" as "Customers";
"sample.odata::order" as "Orders";
association "Customer_Orders" principal "Customers"("ID") multiplicity "1"
dependent "Orders"("CustomerID") multiplicity "*";
}
The association in the example above with the name Customer_Orders defines a relationship between the table
customer, identified by its EntitySet name Customers, on the principal end, and the table order, identified
by its entity set name Orders, on the dependent end. Involved columns of both tables are denoted in braces ({})
after the name of the corresponding entity set. The multiplicity keyword on each end of the association
specifies their cardinality - in this example, one-to-many.
The number of columns involved in the relationship must be equal for both ends of the association, and their order
in the list is important. The order specifies which column in one table is compared to which column in the other
table. In this simple example, the column customer.ID is compared to order.CustomerID in the generated
table join.
As a result of the generation of the service definition above, an AssociationSet named Customer_Orders and
an Association with name Customer_OrdersType are generated:
http://localhost:8002/sample/odata/assocsimple.xsodata/$metadata
PUBLI C

The second association is similar to the first one and is shown in the following listing:
association "Customer_Recruit" principal "Customers"("ID") multiplicity "1"
dependent "Customers"("RecruitID") multiplicity "*";
Complex Associations
For the following example of a complex association, an additional table named knows is introduced that contains a
relationship between customers.
Table definition: sample.odata:knows.hdbtable
table.columns = [
{name = "KnowingCustomerID"; sqlType = INTEGER; nullable = false;},
{name = "KnownCustomerID"; sqlType = INTEGER; nullable = false;}
208
PUBLI C

];
table.primaryKey.pkcolumns = ["KnowingCustomerID","KnownCustomerID"];
Relationships that are stored in association tables such as knows can be similarly defined as simple associations.
Use the keyword over to specify the additional table and any required columns.
Service definition: sample.odata:assoccomplex.xsodata
service {
"sample.odata::customer" as "Customers";
association "Customer_Orders"
principal "Customers"("ID") multiplicity "*"
dependent "Customers"("ID") multiplicity "*"
over "sample.odata::knows" principal ("KnowingCustomerID") dependent
("KnownCustomerID");
}
With the keywords principal and dependent after over you can specify which columns from the association
table are joined with the principal respectively dependent columns of the related entities. The number of
columns must be equal in pairs, and their order in the list is important.
The generated Association in the metadata document is similar to the one created for a simple association
except that the ReferentialConstraint is missing:
tp://localhost:8002/sample/odata/assoccomplex.xsodata/$metadata
Navigation Properties
By only defining an association, it is not possible to navigate from one entity to another. Associations need to be
bound to entities by a NavigationProperty. You can create them by using the keyword navigates:
Service definition: sample.odata:assocnav.xsodata
service {
"sample.odata::customer" as "Customers" navigates ("Customer_Orders" as
"HisOrders");
}
PUBLI C

The example above says that it is possible to navigate from Customers over the association Customer_Order
via the NavigationProperty named "HisOrders".
The right association end is determined automatically by the entity set name. But if both ends are bound to the
same entity, it is necessary to specify the starting end for the navigation. This is done by specifying either from
principal or from dependent which refer to the principal and dependent ends in the association.
Service definition: sample.odata:assocnavself.xsodata
service {
"sample.odata::customer" as "Customers"
navigates ("Customer_Orders" as "HisOrders","Customer_Recruit" as "Recruit"
from principal);
association "Customer_Recruit" principal "Customers"("ID") multiplicity "1"
dependent "Customers"("RecruitID") multiplicity "*";
}
In both cases a NavigationProperty is added to the EntityType.
http://localhost:8002/sample/odata/assocnavself.xsodata/$metadata
9.1.6.7 OData Aggregation
The results of aggregations on columns change dynamically depending on the grouping conditions. This means
that aggregation cannot be performed in SQL views; it needs to be specified in the OData service definition itself.
Depending on the type of object to expose, you need to explicitly specify the columns to aggregate and the
function to use or derived them from metadata in the database.
In general, aggregations do not have consequences for the metadata document. It just effects the semantics of
the concerning properties during runtime. The grouping condition for the aggregation contain all selected non-
aggregated properties. Furthermore, aggregated columns cannot be used in $filter, and aggregation is only
possible with generated keys.
210
PUBLI C

Derived Aggregation
The simplest way to define aggregations of columns in an object is to derive this information from metadata in the
database. The only objects with this information are calculation views and analytic views. For all other object
types, for example, tables and SQL views, the activation will not work. To cause the service to use derived
information, you must specify the keywords aggregates always, as illustrated in the following example:
service {
"sample.odata::calc" as "CalcView"
keys generate local "ID"
aggregates always;
}
Explicit Aggregation
The example for the explicit aggregation is based on the following table definition:
sample.odata:revenues.hdbtable
table.columns = [
{name = "Month"; sqlType = INTEGER; nullable = false;},
{name = "Year"; sqlType = INTEGER; nullable = false;},
{name = "Amount"; sqlType = INTEGER; nullable = true;}
];
table.primaryKey.pkcolumns = ["Month","Year"];
You can aggregate the columns of objects (without metadata) that are necessary for the derivation of aggregation
by explicitly denoting the column names and the functions to use, as illustrated in the following example of a
service definition: sample.odata:aggrexpl.xsodata
service {
"sample.odata::revenues" as "Revenues"
aggregates always (SUM of "Amount");
}
The results of the entity set Revenues always contain the aggregated value of the column Amount. To extract the
aggregated revenue amount per year, add $select=Year,Amount to your requested URI.
9.1.6.8 OData Parameter Entity Sets
SAP HANA calculation views and analytic views can interpret input parameters. For OData, these parameters can
be entered by using a special parameter entity set.
Parameter entity sets can be generated for both calculation views and analytic views by adding parameters via
entity to the entity, as illustrated in the following service-definition example:
service {
PUBLI C

parameters via entity;
}
During activation, the parameters specified in sample.odata/calc.calculationview are retrieved from the
metadata of the calculation view and exposed as a new EntitySet named after the entity set name and the suffix
Parameters, for example, CalcViewParameters. A NavigationProperty named Results is generated to
retrieve the results from the parameterized call.
The name of the generated parameter entity set and the navigation property can be customized, as illustrated in
the following service-definition example:
service {
parameters via entity "CVParams" results property "Execute";
}
With the definition above, the name of the parameter entity set is CVParams, and the name of the
NavigationProperty for the results is Execute.
9.1.7 OData Service Definition Language Syntax
The OData Service Definition Language (OSDL) provides a set of keywords that enable you to set up an ODATA
service definition file that specifies what data to expose, in what way, and to whom.
The following list shows the syntax of the OData Service Definition Language (OSDL) in an EBNF-like format;
conditions that apply for usage are listed after the table.
definition :='service' [namespace] body
namespace :='namespace' quotedstring
quotedstring :=quote string quote
string :=UTF8
quote :='"'
body :='{' content '}'
content :=entry [content]
entry :=( entity | association ) ';'
entity :=object [entityset] [with] [keys] [navigates] [aggregates]
[parameters]
object :=['entity'] ( repoobject | catalogobject )
repoobject :=quote repopackage '/' reponame '.' repoextension quote
repopackage :=string
reponame :=string
repoextension :=string
catalogobject :=catalogobjectschema '.' catalogobjectname
catalogobjectschema :=quotedstring
catalogobjectname :=quotedstring
entityset :='as' entitysetname
entitysetname :=quotedstring
with :=( 'with' | 'without' ) propertylist
propertylist :='(' columnlist ')'
columnlist :=columnname [',' columnlist]
columnname :=quotedstring
keys :='keys' ( keylist | keygenerated )
keylist :=propertylist
keygenerated :='generate' ( keygenlocal )
keygenlocal :='local' columnname
navigates :='navigates' '(' navlist ')'
navlist :=naventry [',' navlist]
212
PUBLI C

naventry :=assocname 'as' navpropname [fromend]
assocname :=quotedstring
navpropname :=quotedstring
fromend :='from' ( 'principal' | 'dependent' )
aggregates :='aggregates' 'always' [aggregatestuple]
aggregatestuple :='(' aggregateslist ')'
aggregateslist :=aggregate [',' aggregateslist]
aggregate :=aggregatefunction 'of' columnname
aggregatefunction :=( 'SUM' | 'AVG' | 'MIN' | 'MAX' )
parameters :='parameters' 'via' 'entity' [parameterentitysetname]
[parametersresultsprop]
parameterentitysetname :=quotedstring
parametersresultsprop :='results' 'property' quotedstring
association :=associationdef principalend dependentend [assoctable]
associationdef :='association' assocname
principalend :='principal' end
dependentend :='dependent' end
end :=endref multiplicity
endref :=endtype '(' joinproperties ')'
endtype :=entitysetname
joinproperties :=columnlist
multiplicity :='multiplicity' quote multiplicityvalue quote
multiplicityvalue :=( '1' | '0..1' | '1..*' | '*' )
assoctable :='over' repoobject overprincipalend overdependentend
overprincipalend :='principal' overend
overdependentend :='dependent' overend
overend :=propertylist
Conditions
The following conditions apply when using the listed keywords:
1. If the namespace is not specified, the schema namespace in the EDMX metadata document will be the
repository package of the service definition file concatenated with the repository object name. E.g. if the
repository design time name of the .xsodata file is sap.hana.xs.doc/hello.xsodata the namespace
will implicitly be sap.hana.xs.doc.hello.
2. keyslist must not be specified for objects of type 'table'. They must only be applied to objects referring a
view type. keygenerated in turn, can be applied to table objects.
3. If the entityset is not specified in an entity, the EntitySet for this object is named after the repository object
name or the catalogobjectname. E.g. if object is "sap.hana.xs.doc/odata_docu" the
entitysetname is implicitly set to odata_docu which then can also be referenced in associations.
4. The fromend in a naventry must be specified if the endtype is the same for both the principalend and
the dependentend of an association.
5. The number of joinproperties in the principalend must be the same as in the dependentend.
6. Ordering in the joinproperties of ends is relevant. The first columnname in the joinproperties of the
principalend is compared with the first columnname of the dependentend, the second with the second,
and so on.
7. The overprincipalend corresponds to the principalend. The number of properties in the
joinproperties and the overproperties must be the same and their ordering is relevant. The same
holds for the dependent end.
8. aggregates can only be applied in combination with keygenerated.
9. If aggregatestuple is omitted, the aggregation functions are derived from the database. This is only
possible for calculation views and analytic views.
PUBLI C

10. Specifying parameters is only possible for calculation views and analytic views.
11. The default parameterentitysetname is the entitysetname of the entity concatenated with the suffix
"Parameters".
12. If the parametersresultsprop is omitted, the navigation property from the parameter entity set to the
entity is called "Results".
9.1.8 OData Service Definition: SQL-EDM Type Mapping
During the activation of the OData service definition, the SAP HANA SQL types are mapped to the required OData
EDM types according to the rules specified in a mapping table.
The following mapping table lists how SAP HANA SQL types are mapped to OData EDM types during the
activation of an OData service definition.
Note
The OData implementation in SAP HANA XS supports only those SQL types listed in the following table.
Table 10: SAP HANA SQL to OData EDM Type Mapping
SAP HANA SQL Type OData EDM Type
Time Edm.Time
Date Edm.DateTime
SecondDate Edm.DateTime
LongDate Edm.DateTime
Timestamp Edm.DateTime
TinyInt Edm.Byte
SmallInt Edm.Int16
Integer Edm.Int32
BigInt Edm.Int64
SmallDecimal Edm.Decimal
Decimal Edm.Decimal
Real Edm.Single
Float Edm.Single
Double Edm.Double
Varchar Edm.String
NVarchar Edm.String
Char Edm.String
NChar Edm.String
214
PUBLI C

SAP HANA SQL Type OData EDM Type
Binary Edm.Binary
Varbinary Edm.Binary
Example SQL Type Mapping
The following examples shows how SAP HANA SQL types (name, integer, Varchar) of columns in a table are
mapped to the OData EDM types in the properties of an entity type.
SAP HANA SQL:
{name = "RefereeID"; sqlType = VARCHAR; nullable = true;}
The following example illustrates how the SAP HANA SQL types illustrated in the previous example are mapped to
EDM types:
<Property Name="ID" Type="Edm.Int32" Nullable="false"/>
<Property Name="RefereeID" Type="Edm.String" Nullable="true"/>
9.1.9 OData URI Parameters, Query Options, and Features
Since OData fully embraces Web technologies, and more specifically URI concepts, all the actions that manage
the interaction with the service at runtime are controlled by HTTP methods, the URI structure, or URI query
parameters.
The examples listed here illustrate how to use the OData service-definition file to specify what data to extract and
what to do with it. You can find examples of the following features:
HTTP Methods
For example, HTTPGET
Resource types
For example: Service documents, entity sets, $metadata, and so on
Query options
For example, $format, $filter, $select, and so on
HTTP Methods
Since the current OData Infrastructure only supports read services (no Create/Update/Delete), the only
supported HTTP Method is GET, although the POST method is used for the $batch resource, which you can use to
send a batch of requests in one multipart request (containing only GET requests). The result is that only read
requests are possible.
PUBLI C

Resource Types
You can use the resource types (or constructs) described and explained here to define how the service definition
exposes data. For example, the Service Document, Resource Type, or Entity sets enable you to select and view
the contents of the entity sets that are queried.
Service Document
The most basic resource type is the call to the .xsodata resource itself. This supplies information about the
entity set within this resource; otherwise known as the Service Document.
Sample XSODATA Service:
service namespace "SAMPLE" {
"sample.hello::hello_tbl" as "Entries";
}
URI Construct:
/<service>.xsodata/
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/
Sample XML Output:
Sample JSON Output:
216
PUBLI C

Entity Sets
This construct allows you to view the entire content of the queried Entity Set.
URI Construct
/<service>.xsodata/<Entity Set Name>
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries
Sample ATOM Output:
Sample JSON Output:
PUBLI C

Single Entity
Any single entity (record) can be accessed by supplying the key field values after the Entity Set.
URI Construct:
Single Key: /<service>.xsodata/<Entity Set Name>(<KeyValue>)
Multiple Keys: /<service>.xsodata/<Entity Set Name>(<KeyName>=<KeyValue>,<Repeat>)/
Sample URI:
Single Key: /<service>.xsodata/<Entity Set Name>(<KeyValue>)
Multiple Keys: http://hanaxs:8000/sap./sflight/spfli.xsodata/
FLIGHTS(CARRID='AA',CONNID='0017')#
Sample ATOM Output:
Sample JSON Output:
218
PUBLI C

Single Property
Likewise, any single property of an entry can be directly accessed as well.
URI Construct:
Single Key: /<service>.xsodata/<Entity Set Name>(<KeyValue>)/<Property Name>
Multiple Keys: http://hanaxs:8000/sap./sflight/spfli.xsodata/
FLIGHTS(CARRID='AA',CONNID='0017')/CITYFROM
Sample URI:
Single Key: http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries(1)/TEXT
Multiple Key: http://hanaxs:8000/sap./sflight/spfli.xsodata/
FLIGHTS(CARRID='AA',CONNID='0017')/CITYFROM
Sample ATOM Output:
Sample JSON Output:
PUBLI C

$metadata
This parameter requests that the OData service return a special document that describes the structure of the
service itself. All properties, their data types, and all associations and relationships are described in this
document.
URI Construct:
/<service>.xsodata/$metadata
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/$metadata
Sample Output:
$count
This request allows you to return the number of records in an Entity Set instead of the details of the Entity Set. It is
basically the same as using a SELECT COUNT(*) in an SQL Statement without top limitation.
URI Construct:
/<service>.xsodata/<Entity Set Name>/$count
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries/$count
Sample Output:
220
PUBLI C

$value
By specifying this segment after a single property, the service returns the value of the property as plain text.
URI Construct:
/<service>.xsodata/<Entity Set Name>/$value
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries(1)/Text/$value
Sample Output:
$batch
The $batch resource allows you to send multiple requests in one multipart request by using the POST http
method.
Note
You cannot use the $batch resource with a GET request.
URI Construct:
/<service>.xsodata/$batch
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/$batch
Sample Request:
PUBLI C

Sample ATOM and JSON Output:
222
PUBLI C

$format
This parameter controls the output format. The default format is ATOM. However, the other supported option is
JSON, which can be triggered by setting the $format=json.
URI Construct:
/<service>.xsodata/<Entity Set Name>?$format=<json|atom(default)>
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries/?$format=json
Sample Output:
$filter
A parameter that allows you to restrict the entities in the entity set. This option builds a dynamic WHERE condition
into the underlying query.
URI Construct:
/<service>.xsodata/<Entity Set Name>?$filter <Property> eq '<value>'
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries/?$filter=CREATED%20eq
%20datetime'2012-05-14T08:28:44.405'
Sample ATOM Output:
PUBLI C

Sample JSON Output:
$orderby
This is a parameter that controls how the entities in the entity set are sorted.
URI Construct:
/<service>.xsodata/<Entity Set Name>?$orderby=<Property> asc|desc
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries/?$orderby=CREATED
%20desc
Sample ATOM Output:
224
PUBLI C

Sample JSON Output:
$skip
This is a parameter that allows you to skip ahead in an entity set only. It is useful when enabling "paging" through
an entity set.
URI Construct:
/<service>.xsodata/<Entity Set Name>?$skip=n
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries/?$skip=2
Sample ATOM Output:
PUBLI C

Sample JSON Output:
$top
This option allows you to limit the number of records returned regardless of how many records exist in the Entity
Set. It is basically the same as using a SELECT ... UP TO n ROWS in an SQL Statement.
URI Construct:
/<service>.xsodata/<Entity Set Name>?$top=n
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries/?$top=3
226
PUBLI C

Sample ATOM Output:
Sample JSON Output:
$inlinecount
This parameter triggers the OData service to embed an extra property with the total count in the complete entity
set. This is the count of all entities after the $filter option has been applied.
URI Construct:
/<service>.xsodata/<Entity Set Name>?$inlinecount=allpages
Sample URI:
PUBLI C

http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries/?$inlinecount=allpages
Sample ATOM Output:
Sample JSON Output:
$select
This property allows you to control which properties for each entity are returned in the entity set.
URI Construct:
/<service>.xsodata/<Entity Set Name>?$select=<property>,<property>,<...>
Sample URI:
http://hanaxs:8000/sample/hello/hello_odata.xsodata/Entries/?$select=TEXT
Sample ATOM Output:
228
PUBLI C

Sample JSON Output:
9.1.10 OData Security Considerations
Enabling access to data by means of OData can create some security-related issues that you need to consider and
address, for example, the data you want to expose, who can start the OData service, and so on.
PUBLI C

If you want to use OData to expose data to users and clients in SAP HANA application services, you need to bear
in mind the security considerations described in the following list:
Data Access
Restrict user select authorization for tables/views exposed by the OData service
OData Service
Restrict authorization rights to start the OData service
OData Statistical content
Restrict access to the URL/Path used to expose OData content in the Web browser
9.2 Data Access with XMLA in SAP HANA XS
In SAP HANA Extended Application Services, the persistence model (for example, tables, views and stored
procedures) is mapped to the consumption model that is exposed to clients - the applications you write to extract
data from the SAP HANA database.
You can map the persistence and consumption models with XML for Analysis (XMLA). With XMLA, you write
multi-dimensional -expressions (MDX) queries wrapped in an XMLA document. An XML for Analysis (XMLA)
application running in SAP HANA application services is used to provide the consumption model for client
applications exchanging MDX queries (wrapped in XMLA documents) with the SAP HANA database.
XMLA uses Web-based services to enable platform-independent access to XMLA-compliant data sources for
Online Analytical Processing (OLAP). XMLA enables the exchange of analytical data between a client application
and a multi-dimensional data provider working over the Web, using a Simple Object Access Protocol (SOAP)-
based XML communication application-programming interface (API).
Applications running in SAP HANA XS enable very accurate control of the flow of data between the presentational
layer, for example, in the Browser, and the data-processing layer in SAP HANA itself, where the calculations are
performed, for example in SQL or SqlScript. If you develop and deploy an XMLA service running in SAP HANA XS,
you can take advantage of the embedded access to SAP HANA that SAP HANA XS provides; the embedded
access greatly improves end-to-end performance.
9.2.1 XML for Analysis (XMLA)
XML for Analysis (XMLA) uses Web-based services to enable platform-independent access to XMLA-compliant
data sources for Online Analytical Processing (OLAP).
XMLA enables the exchange of analytical data between a client application and a multi-dimensional data provider
working over the Web, using a Simple Object Access Protocol (SOAP)-based XML communication application-
programming interface (API).
Implementing XMLA in SAP HANA enables third-party reporting tools that are connected to the SAP HANA
database to communicate directly with the MDX interface. The XMLA API provides universal data access to a
particular source over the Internet, without the client having to set up a special component. XML for Analysis is
optimized for the Internet in the following ways:
Query performance
Time spent on queries to the server is kept to a minimum
230
PUBLI C

Query type
Client queries are stateless by default; after the client has received the requested data, the client is
disconnected from the Web server.
In this way, tolerance to errors and the scalability of a source (the maximum permitted number of users) is
maximized.
XMLA Methods
The specification defined in XML for Analysis Version 1.1 from Microsoft forms the basis for the implementation of
XML for Analysis in SAP HANA.
The following list describes the methods that determine the specification for a stateless data request and
provides a brief explanation of the method's scope:
Discover
Use this method to query metadata and master data; the result of the discover method is a rowset. You can
specify options, for example, to define the query type, any data-filtering restrictions, and any required XMLA
properties for data formatting.
Execute
Use this method to execute MDX commands and receive the corresponding result set; the result of the
Execute command could be a mult-idimensional dataset or a tabular rowset. You can set options to specify
any required XMLA properties, for example, to define the format of the returned result set or any local
properties to use to determine how to format the returned data.
9.2.2 XMLA Service Definition
The XMLA service definition is a file you use to specify which data is exposed as XMLA collections. Exposed data is
available for analysis and display by client applications, for example, a browser that uses functions provided either
by the XMLA service running in SAP HANA XS or by an XMLA client library running on the client system.
To expose information via XMLA to applications using SAP HANA Extended Application Services (SAP HANA XS),
you define database views that provide the data with the required granularity and you use the XMLA service
definition to control access to the exposed data.
Note
SAP HANA XS supports XMLA version 1.1, which you can use to send MDX queries.
An XMLA service for SAP HANA XS is defined in a text file with the file suffix .xsxmla, for example,
XMLASrvDef.xsxmla. The file must contain only the entry {*}, which would generate a completely operational
XMLA service.
PUBLI C

XMLA Service-Definition Keywords
Currently, the XMLA service-definition file enables you to specify only that all authorized data is exposed to XMLA
requests, as illustrated in the following example:
Service {*}
9.2.3 XMLA Security Considerations
Enabling access to data by means of XMLA opens up some security considerations that you need to address, for
example, the data you want to expose, who can start the XMLA service, and so on.
If you want to use XMLA to expose data to users and clients in SAP HANA XS, you need to bear in mind the
security considerations described in the following list:
Data Access
Restrict user select authorization for data exposed by the XMLA service
XMLA Statistical content
Restrict access to the URL/Path used to expose XMLA content in the Web browser, for example, using the
application-access file (.xsaccess)
9.2.4 Multidimensional Expressions (MDX)
Multidimensional Expressions (MDX) is a language you can use to query multidimensional data stored in OLAP
cubes.
MDX uses a multidimensional data model to enable navigation in multiple dimensions, levels, and up and down a
hierarchy. With MDX, you can access pre-computed aggregates at specified positions (levels or members) in a
hierarchy.
Note
MDX is an open standard. However, SAP has developed extensions to MDX that are designed to enable faster
and more efficient access to multidimensional data, for example, to serve specific SAP HANA application
requirements and to optimize the resultset for SAP HANA clients.
MDX is implicitly a hierarchy-based paradigm. All members of all dimensions must belong to a hierarchy. Even if
you do not explicitly create hierarchies in your SAP HANA data model, the SAP HANA modeler implicitly
generates default hierarchies for each dimension. All identifiers that are used to uniquely identify hierarchies,
levels and members in MDX statements (and metadata requests) embed the hierarchy name within the identifier.
In SAP HANA, the standard use of MDX is to access SAP HANA models (for example, analytical and attribute
views) that have been designed, validated and activated in the modeler in the SAP HANA studio. The studio
provides a graphical design environment that enables detailed control over all aspects of the model and its
language-context-sensitive runtime representation to users.
MDX in SAP HANA uses a runtime cube model, which usually consists of an analytical (or calculation) view that
represents data in which dimensions are modeled as attribute views. You can use the analytical view to specify
232
PUBLI C

whether a given attribute is intended for display purposes only or for aggregation. The key attributes of attribute
views are linked to private attributes in an analytical view in order to connect the entities. One benefit of MDX in
SAP HANA is the native support of hierarchies defined for attribute views.
Note
MDX in SAP HANA includes native support of hierarchies defined for attribute views. SAP HANA supports level-
based and parent-child hierarchies and both types of hierarchies are accessible with MDX.
SAP HANA supports the use of variables in MDX queries; the variables are a SAP-specific enhancement to
standard MDX syntax. You can specify values for all mandatory variables that are defined in SAP HANA studio to
various modeling entities. The following example illustrates how to declare SAP HANA variables and their values:
MDX
Select
From [MINI_C1_VAR]
Where [Measures].[M2_1_M3_CONV]
SAP VARIABLES [VAR_VAT] including 10,
[VAR_K2] including 112,
[VAR_TARGET_CURRENCY] including 'EUR',
9.2.5 MDX Functions
MDX in SAP HANA supports a variety of standard MDX functions.
Table 11: Supported MDX Functions
Function Description
Aggregate Returns a calculated value using the appropriate aggregate function, based on the
aggregation type of the member.
Ancestor Returns the ancestor of a member at a specified level or at a specific distance away
in the hierarchy.
Ancestors Returns a set of all ancestors of a member at a specified level or at a specific
distance away in the hierarchy.
Ascendants Returns the set of the ascendants of the member.
Avg Returns the average value of a numeric expression evaluated over a set.
BottomCount Returns a specified number of items from the bottom of a set, optionally sorting
the set first.
Children Returns the children of a member.
ClosingPeriod Returns the last sibling among the descendants of a member at a specified level.
Count Counts the number of members in the tuple.
Cousin Returns the child member with the same relative position under a parent member
as the specified child member.
Crossjoin Returns the cross product of two sets.
PUBLI C

CurrentMember Returns the current member along a hierarchy.
DefaultMember Returns the default member of a hierarchy.
Descendants Returns the set of descendants of a member at a specified level or at a specific
Dimension Returns the hierarchy that contains a specified member or level.
Dimensions Returns a hierarchy specified by a numeric or string expression.
Distinct Returns a set, removing duplicate tuples from a specified set.
DistinctCount Returns the number of distinct tuples in a set.
DrillDownLevel Drills down the members of a set one level below the lowest level represented in the
set, or to one level below an optional level of a member represented in the set.
DrillDownLevelBottom Drills down the members of a specified count of bottom members of a set, at a
specified level, to one level below.
DrillDownLevelTop Drills down a specified count of top members of a set, at a specified level, to one
level below.
DrillDownMember Drills down the members in a specified set that are present in a second specified
set.
DrillDownMemberBottom Drills down the members in a specified set that are present in a second specified
set, limiting the result set to a specified number of bottommost members.
DrillDownMemberTop Drills down the members in a specified set that are present in a second specified
set, limiting the result set to a specified number of topmost members.
DrillUpLevel Drills up the members of a set that are below a specified level.
DrillUpmember Drills up the members in a specified set that are present in a second specified set.
Except Finds the difference between two sets, optionally retaining duplicates.
Filter Returns the set resulting from filtering a set based on a search condition.
FirstChild Returns the first child of a specified member.
FirstSibling Returns the first child of the parent of a specified member.
Generate Applies a set to each member of another set, and then joins the resulting sets by
union. Alternatively, this function returns a concatenated string created by
evaluating a string expression over a set.
Head Returns the first specified number of elements in a set.
Hierarchize Orders the members of a specified set in a hierarchy in natural or, optionally, post-
natural order.
Hierarchy
Instr The InStr function finds the starting location of a substring within a specified string.
Intersect Returns the intersection of two sets, optionally retaining duplicates.
234
PUBLI C

IsAncestor Returns true if the first member specified is an ancestor of the second member
specified, else returns false.
IsGeneration Returns true if the member specified is a leaf, else returns false.
IsLeaf Returns true if the first member specified is an ancestor of the second member
IsSibling Returns true if the first member specified is an sibling of the second member
Item If an integer is specified, the Item function returns the tuple that is in the zero-
based position specified by Index.
IIF Returns one of values determined by a logical test.
Lag Returns the member that is a specified number of positions prior to a specified
member along the dimension of the member.
LastChild Returns the last child of a specified member.
LastPeriods Returns a set of members prior to and including a specified member.
LastSibling Returns the last child of the parent of a specified member.
Lead Returns the member that is a specified number of positions following a specified
Leaves If a dimension name is specified, returns a set that contains the leaf members of
the key attribute for the specified dimension.If a dimension name is not specified,
the function returns a set that contains the leaf members of the entire cube.
Left The Left function returns a string of a specified number of characters from the left
side (beginning) of a specified string.
Level Returns the level of a member.
Levels Returns the level whose zero-based position in a dimension is specified by a
numeric expression.
Max Returns the maximum value of a numeric expression evaluated over a set.
Member_caption Returns the caption of a member
Members Returns the set of all members in a specified hierarchy.
MembersAscendantsDesc
endants
Returns the set of specified members in a given hierarchy.
Mid The Mid function returns a substring of a string argument.
Min Returns the minimum value of a numeric expression evaluated over a set
MTD Returns a set of members from the Month level in a Time dimension starting with
the first period and ending with a specified member.
Name Returns the name of a specified hierarchy or member.
NextMember Returns the next member in the level that contains a specified member.
PUBLI C

NOT Performs a logical negation on a numeric expression.
OpeningPeriod Returns the first sibling among the descendants of a specified level, optionally at a
specified member.
OR Performs a logical disjunction on two numeric expressions.
Ordinal Returns the zero-based ordinal value associated with a specified level.
ParallelPeriod Returns a member from a prior period in the same relative position as a specified
member.
Parent Returns the parent of a specified member.
PeriodsToDate Returns a set of members (periods) from a specified level starting with the first
member and ending with a specified member.
PrevMember Returns the previous member in the level that contains a specified member.
Properties Returns a string containing the value of the specified member property.
QTD Returns a set of members from the Quarter level in a Time dimension starting with
Range Performs a set operation that returns a naturally ordered set, with the two
specified members as endpoints, and all members between the two specified
members included as members of the set
Right The Right function returns a string of a specified number of characters from the
right side (end) of a specified string.
Siblings Returns the set of siblings of a specified member, including the member itself.
StrToMember Returns a member from a string expression in MDX format.
StrToSet Constructs a set from a specified string expression in MDX format.
StrToTuple Constructs a tuple from a specified string expression in MDX format.
StrToValue Returns a value from a string expression
Subset Returns a subset of tuples from a specified set.
Sum Returns the sum of a numeric expression evaluated over a specified set.
Tail Returns the last specified number of elements in a set.
TopCount Returns a specified number of items from the topmost members of a specified set,
optionally ordering the set first.
Union Performs a set operation that returns a union of two sets, removing duplicate
members.
UniqueName Returns the unique name of a specified hierarchy.
WTD Returns a set of members from the Week level in a Time dimension starting with
YTD Returns a set of members from the Year level in a Time dimension starting with the
first period and ending with a specified member.
236
PUBLI C

9.2.6 MDX Extensions
SAP HANA supports several extensions to the MDX language, for example, additional predefined functions and
support for variables.
9.2.6.1 Sibling_Ordinal Intrinsic Property
The object Member includes a property called Sibling_Ordinal, that is equal to the 0-based position of the member
within its siblings.
Example
WITH
MEMBER [Measures].[Termination Rate] AS
[Measures].[NET_SALES] / [Measures].[BILLED_QUANTITY]
SELECT
{
[Measures].[NET_SALES],
[Measures].[BILLED_QUANTITY],
[Measures].[Termination Rate]
} ON COLUMNS,
Descendants
(
[MDX_TEST_10_DISTRIBUTION_CHANNEL].[MDX_TEST_10_DISTRIBUTION_CHANNEL].[All].
[(all)],
1,
SELF_AND_BEFORE
)
DIMENSION PROPERTIES SIBLING_ORDINAL ON ROWS
FROM MDX_TEST_10_ITELO_SALES_DATA
9.2.6.2 MembersAscendantsDescendants Function
SAP HANA includes a new function called MembersAscendantsDescendants that enables you to get, for
example, all ascendants and descendants of a specific member. This function improves on the standard MDX
functions Ascendants and Descendants.
The function can be called as follows:
MembersAscendantsDescendants (<set>, <flag>)
set: A set of members from a single hierarchy
flag: Indicates which related members to return, and can be one of the following:
MEMBERS_AND_ASCENDANTS_AND_DESCENDANTS
MEMBERS_AND_ASCENDANTS
MEMBERS_AND_DESCENDANTS
ASCENDANTS_AND_DESCENDANTS
PUBLI C

ONLY_ASCENDANTS
ONLY_DESCENDANTS
Example
SELECT
{ [Measures].[SALES] }
ON COLUMNS,
NON EMPTY
{ Hierarchize( MembersAscendantsDescendants([JUICE_TIME].[TimeHier].[QUARTER].[3]:
[JUICE_TIME].[TimeHier].[QUARTER].[4], MEMBERS_AND_ASCENDANTS_AND_DESCENDANTS )) }
ON ROWS
FROM [JUICE]
Example
SELECT
ON COLUMNS,
NON EMPTY
[JUICE_TIME].[TimeHier].[QUARTER].[4], ONLY_ASCENDANTS )) }
ON ROWS
FROM [JUICE]
9.2.6.3 Variables
An MDX SELECT statement in SAP HANA enables you to send values for variables defined within modeling views.
Analytic and calculation views can contain variables, that can be bound to specific attributes. When calling the
view, you can send values for those variables. These variables can be used, for example, to filter the results.
SAP HANA supports an extension to MDX whereby you can pass values for variables defined in views by adding a
SAP Variables clause in your select statement. Here is the syntax for a Select statement:
<select_statement>: [WITH <formula_specification> ]
SELECT [<axis_specification>[,<axis_specification>...]]
FROM <cube_specification>
[WHERE <slicer_specification>
SAP VARIABLES: <sap_variable> [[,] <sap_variable>]]
<sap_variable>: <variable_name> <sign> [<option>] <variable_value>
<sign>: INCLUDING | EXCLUDING
<option>: = | > | >= | < | <= | <>
<variable_value>: <unique_member_name>
| <unsigned_numeric_literal>
| <string_value_expression>
| <member> : <member>
| <character_string_literal> : <character_string_literal>
| <unsigned_numeric_literal> : <unsigned_numeric_literal>
Example
The following specifies a single value for variables VAR_KAT, VAR_K2, and VAR_TARGET_CURRENCY.
SELECT
FROM [MINI_C1_VAR]
238
PUBLI C

WHERE [Measures].[M2_1_M3_CONV]
[VAR_TARGET_CURRENCY] including 'EUR'
Example
The following specifies an interval for variable VAR_K2.
SELECT NON EMPTY
{
[K2].[K2].Members
}ON ROWS
FROM [MINI_C1_VAR_SIMPLE]
WHERE [Measures].[M3_CONV]
SAP VARIABLES [VAR_K2] including [K2].[K2].&[122]:[K2].[K2].&[221]
Metadata on Variables in Views
SAP HANA includes the following set of tables that contain information about the variables defined for views:
BIMC_VARIABLE
BIMC_VARIABLE_ASSIGNMENT
BIMC_VARIABLE_VALUE
BIMC_VARIABLE_ODBO (virtual table)
The tables enable, for example, an application to retrieve the variables defined for a view and create a user
interface so the user can enter values.
9.2.7 Defining the Data an XMLA Service Exposes
An XMLA service exposes data stored in database tables for analysis and display by client applications. However,
first of all, you need to ensure that the tables and views to expose as an XMLA service actually exist and are
accessible.
To define the data to expose using an XMLA service, you must perform at least the following tasks:
1. Create a simple database schema.
2. Create a simple database table to expose with an XMLA service.
3. If required, create a simple database view to expose with an XMLA service.
4. Grant select privileges to the tables and views to be exposed with the XMLA service.
PUBLI C

9.2.8 Creating an XMLA Service Definition
The XMLA service definition is a file you use to specify which data is exposed as XMLA/MDX collections for
analysis and display by client applications.
An XMLA service for SAP HANA XS is defined in a text file with the file suffix .xsxmla, for example,
XMLASrvDef.xsxmla. The file resides in the package hierarchy of the XMLA application and must contain the
entry service {*}, which generates an operational XMLA service.
Prerequisities for the creation of an XMLA service definition:
SAP HANA studio and client installed and configured
SAP HANA database user available with repository privileges (for example, to add packages)
A SAP HANA development system added to (and available in) SAP HANA studio, for example, in either the
Navigator view or the SAP HANA Repositories view
A working development environment including a repository workspace, a package structure for your XMLA
application, and a shared project to enable you to synchronize changes to the XMLA project files in the local
file system with the repository
Data is available to expose using the XMLA interface
1. In the shared project you are using for your XMLA application, use the Project Explorer view to locate the
package where you want to create the new XMLA service definition.
Note
The file containing the XMLA service definition must be placed in the root package of the XMLA application
for which the service is intended.
2. Create the file that will contain your XMLA service definition.
In the Project Explorer view, right-click the folder where you want to create the new XMLA service-definition
file and choose New File in the context-sensitive popup menu displayed.
3. Create the XMLA service definition.
The XMLA service definition is a configuration file that you use to specify which data is to be exposed as an
XMLA collection.
The following code is an example of a valid XMLA service definition, which exposes all authorized data to
XMLA requests:
service{*}
4. Place the valid XMLA service definition in the root package of the XMLA application.
5. Save, commit, and activate the XMLA service definition in the SAP HANA repository.
9.2.9 Tutorial: Using the SAP HANA XMLA Interface
You can use the XML for Analysis (XMLA) interface included in SAP HANA Extended Application Services (SAP
HANA XS) to provide a service that enables XMLA-capable clients to query multidimensional cubes in SAP HANA.
Since the artifacts required to get a simple XMLA service up and running are stored in the repository, make sure
that you read through and comply with the following prerequisites:
240
PUBLI C

You have a development workspace in the SAP HANA repository
You have created a dedicated project in the repository workspace
You have shared the new project
A multidimensional data cube is available in SAP HANA, for example, in the form of a calculation view, an
analytic view, or an attribute view
An XMLA client is available
To send an XMLA query to SAP using the XMLA interface provided by SAP HANA XS, perform the following steps:
1. Create a root package for your XMLA interface test, for example, helloxmla and save and activate it in the
repository.
Note
2. Create an application descriptor for your new XMLA test in your root XMLA package helloxmla.
The application descriptor (.xsapp) is the core file that you use to define an application's availability within
SAP HANA. The .xsapp file sets the point in the application-package structure from which content will be
served to the requesting clients.
Note
The application-descriptor file has no content and no name; it only has the extension .xsapp.
3. Save, commit, and activate the application-descriptor file in the repository.
4. Create an application-access file for your new XMLA test and place it in your root XMLA package helloxmla.
the application.
Note
The application-access file has no name; it only has the extension .xsaccess.
Ensure the application content is exposed to HTTP requests by entering the following command in
the .xsaccess file for your new XMLA test:
{
"exposed" : true
}
5. Save, commit, and activate the application-access file in the repository.
6. Create an XMLA service-definition file and place it in your root XMLA package helloxmla.
The XMLA service-definition file has the file extension .xsxmla, for example, hello.xsxmla and must be
located in the root package of the XMLA application:
Enter the following content in the hello.xsxmla XMLA service-definition file:
service {*}
7. Save, commit, and activate the XMLA service-definition file in the repository.
PUBLI C

8. Test the connection to the SAP HANA XS Web server.
http://<hana.server.name>:80<HANA_instance_number>/helloxmla/hello.xsxmla
Note
You have successfully completed this step if you see a 404 Error page; the page indicates that the SAP
HANA XS Web server has responded.
9. Connect your XMLA client application to the inbuilt XMLA interface in SAP HANA XS.
To connect an XMLA-capable client (for example, Microsoft Excel) with the XMLA interface in SAP HANA XS,
you will need a product (for example, a plug-in for Microsoft Excel) that can transfer the XMLA message that
the SAP HANA XS XMLA interface can understand.
10. Configure your client to send an XMLA query to SAP HANA.
242
PUBLI C

10 Writing Server-Side JavaScript Code
SAP HANA Extended Application Services (SAP HANA XS) provide applications and application developers with
In addition to providing application-specific consumption models, SAP HANA XS also host system services that
are part of the SAP HANA database, for example: search services and a built-in Web server that provides access
to static content stored in the SAP HANA repository.
The consumption model provided by SAP HANA XS focuses on server-side applications written in JavaScript.
Applications written in server-side JavaScript can make use of a powerful set of specially developed API functions,
for example, to enable access to the current request session or the database. This section describes how to write
server-side JavaScript code that enables you to expose data, for example, using a Web Browser or any other
HTTP client.
10.1 Data Access with JavaScript in SAP HANA XS
In SAP HANA Extended Application Services, the persistence model (for example, tables, views and stored
procedures) is mapped to the consumption model that is exposed via HTTP to clients - the applications you write
to extract data from SAP HANA.
You can map the persistence and consumption models in the following way:
Application-specific code
Write code that runs in SAP HANA application services. Application-specific code (for example, server-side
JavaScript) is used in SAP HANA application services to provide the consumption model for client
applications.
Applications running in SAP HANA XS enable you to accurately control the flow of data between the
presentational layer, for example, in the Browser, and the data-processing layer in SAP HANA itself, where the
calculations are performed, for example in SQL or SQLScript. If you develop and deploy a server-side JavaScript
application running in SAP HANA XS, you can take advantage of the embedded access to SAP HANA that SAP
HANA XS provides; the embedded access greatly improves end-to-end performance.
10.2 Server-Side JavaScript in SAP HANA XS
SAP HANA application services (XS server) supports server-side application programming in JavaScript. The
server-side application you develop can use a collection of JavaScript APIs to expose authorized data to client
requests, for example, to be consumed by a client GUI such as a Web browser or any other HTTP client.
The functions provided by the JavaScript APIs enable server-side JavaScript applications to perform the following
actions:
Interact with the SAP HANA XS runtime environment
Directly access SAP HANA database capabilities
Note
Writing Server-Side JavaScript Code
PUBLI C

Unlike OData and XMLA, the JavaScript API enables you not only to expose data but to update, insert, and
delete data, too.
JavaScript programs are stored in the repository along with all the other development resources. When the
programs are activated, the code is stored in the repository as a runtime object.
10.2.1 JavaScript Editor
You can write server-side JavaScript using the SAP HANA studio JavaScript editor, which provides syntax
validation, code highlighting and code completion.
When you edit a JavaScript file within an XS Project, the editor automatically has a reference to the SAP HANA XS
JavaScript API. You can access the APIs objects by entering $, following by a period (.), and the objects are
available, as in the following example.
JSLint Validations
The JavaScript editor includes the JSLint open-source library, which helps to validate JavaScript code. The editor
highlights any code that does not conform to the JSLint standards.
To configure the JSLint library and determine which validations are performed, go to: Windows Preferences
JSLint . In the preferences window, each JSLint setting is followed by the corresponding JSLint command name,
which you can use to lookup more information on the JSLint Web site.
You can disable all JSLint validations for files in a specific project by right-clicking the project and choosing
Disable JSLint.
Related Links
http://www.jslint.com/lint.html
244
PUBLI C

10.2.2 Server-Side JavaScript Security Considerations
for (and risk of) external attacks such as cross-site scripting and forgery, and insufficient authentication.
The following list illustrates the areas where special attention is required to avoid security-related problems when
writing server-side JavaScript. Each of the problems highlighted in the list is described in detail in its own
dedicated section:
SSL/HTTPS
Enable secure HTTP (HTTPS) for inbound communication required by an SAP HANA application.
Injection flaws
In the context of SAP HANA Extended Application Services (SAP HANA XS) injection flaws concern SQL
injection that modifies the URL to expand the scope of the original request.
Cross-site scripting (XSS)
Web-based vulnerability that involves an attacker injecting JavaScript into a link with the intention of running
the injected code on the target computer.
Broken authentication and session management
Leaks or flaws in the authentication or session management functions allow attackers to impersonate users
and gain access to unauthorized systems and data.
Insecure direct object references
An application lacks the proper authentication mechanism for target objects.
Cross-site request forgery (XSRF)
Exploits the trust boundaries that exist between different Web sites running in the same web browser session.
Incorrect security configuration
Attacks against the security configuration in place, for example, authentication mechanisms and
authorization processes.
Insecure cryptographic storage
Sensitive information such as logon credentials is not securely stored, for example, with encryption tools.
Missing restrictions on URL Access
Sensitive information such as logon credentials is exposed.
Insufficient transport layer protection
Network traffic can be monitored, and attackers can steal sensitive information such as logon credentials or
credit-card data.
Invalid redirects and forwards
Web applications redirect users to other pages or use internal forwards in a similar manner.
XML processing issues
Potential security issues related to processing XML as input or to generating XML as output
Related Links
SAP HANA Security Guide
SAP HANA SQL Reference
SSL/HTTPS [page 246]
for (and risk of) external attacks such as cross-site scripting and forgery, and insufficient authentication. You can
set up SAP HANA to use secure HTTP (HTTPS).
Injection flaws [page 247]
PUBLI C

for (and risk of) injection flaws. Typically, injection flaws concern SQL injection and involve modifying the URL to
expand the scope of the original request.
Cross-site scripting (XSS) [page 248]
for (and risk of) cross-site scripting (XSS) attacks. Cross-site scripting is a web-based vulnerability that involves
an attacker injecting JavaScript into a link with the intention of running the injected code on the target computer.
Broken authentication and session management [page 249]
for (and risk of) attack against authentication infrastructure. Leaks or flaws in the authentication or session
management functions allow attackers to impersonate users and gain access to unauthorized systems and data.
Insecure direct object references [page 249]
for (and risk of) attacks using insecure references to objects.
Cross-site request forgery (XSRF) [page 250]
for (and risk of) cross-site request forgery (XSRF). Cross-site scripting is a web-based vulnerability that exploits
the trust boundaries that exist between different websites running in the same web browser session.
Incorrect security configuration [page 252]
for (and risk of) attacks against the security configuration in place, for example, authentication mechanisms and
Insecure cryptographic storage [page 252]
for (and risk of) attacks against the insecure or lack of encryption of data assets.
Missing restrictions on URL Access [page 253]
for (and risk of) unauthorized access to URLs.
Insufficient transport layer protection [page 253]
for (and risk of) insufficient protection of the transport layer.
XML processing issues [page 254]
for (and risk of) attacks aimed at the process used to parse XML input and generate the XML output.
10.2.2.1 Server-Side JavaScript: SSL/HTTPS
for (and risk of) external attacks such as cross-site scripting and forgery, and insufficient authentication. You can
set up SAP HANA to use secure HTTP (HTTPS).
246
PUBLI C

SSL/HTTPS Problem
Incoming requests for data from client applications use secure HTTP (HTTPS), but the SAP HANA system is not
configured to accept the HTTPS requests.
SSL/HTTPS Recommendation
Ensure the SAP Web Dispatcher is configured to accept incoming HTTPS requests. For more information, see the
SAP HANA Security Guide.
Note
The HTTPS requests are forwarded internally from the SAP Web Dispatcher to SAP HANA XS as HTTP (clear
text).
Related Links
10.2.2.2 Server-Side JavaScript: Injection Flaws
for (and risk of) injection flaws. Typically, injection flaws concern SQL injection and involve modifying the URL to
expand the scope of the original request.
Injection Flaws Problem
In the context of SAP HANA XS, injection flaws mostly concern SQL injection, which can occur in the SAP HANA
XS JavaScript API or SQL script itself (both standard and dynamic). For example, the URL http://xsengine/
customer.xsjs?id=3 runs the code in the JavaScript file customer.xsjs shown below:
var pstmt = conn.prepareStatement( " SELECT * FROM accounts WHERE custID='" +
$.request.parameters.get("id"));
By modifying the URL, for example, to http://xsengine/customer.xsjs?id=3 'OR 1=1', an attacker can
view not just one account but all the accounts in the database.
Note
SAP HANA XS applications rely on the authorization provided by the underlying SAP HANA database. Users
accessing an SAP HANA XS based application require the appropriate privileges on the database objects to
execute database queries. The SAP HANA authorization system will enforce the appropriate authorizations.
This means that in those cases, even if the user can manipulate a query, he will not gain more access than is
PUBLI C

assigned to him through roles or privileges. Definer mode SQL script procedures are an exception to this rule
that you need to take into consideration.
Injection Flaws Recommendation
To prevent injection flaws in the JavaScript API, use prepared statements to create a query and place-holders to
fill with results of function calls to the prepared-statement object; to prevent injection flaws in standard SQL
Script, use stored procedures that run in caller mode; in caller mode, the stored procedures are executed with the
credentials of the logged-on HANA user. Avoid using dynamic SQL if possible. For example, to guard against the
SQL-injection attack illustrated in the problem example, you could use the following code:
var pstmt = conn.prepareStatement( " SELECT * FROM accounts WHERE custID=?' );
pstmt.setInt(1, $.request.parameters.get("id"), 10);
Prepared statements enable you to create the actual query you want to run and then create several placeholders
for the query parameters. The placeholders are replaced with the proper function calls to the prepared statement
object. The calls are specific for each type in such a way that the SAP HANA XS JavaScript API is able to properly
escape the input data. For example, to escape a string, you can use the setString function.
Related Links
http://help.sap.com/hana/html/SAP_HANA_SQL_Reference_en.html
10.2.2.3 Server-Side JavaScript: Cross-Site Scripting
for (and risk of) cross-site scripting (XSS) attacks. Cross-site scripting is a web-based vulnerability that involves
an attacker injecting JavaScript into a link with the intention of running the injected code on the target computer.
Cross-Site Scripting Problem
The vulnerability to cross-site scripting attacks comes in the following forms:
Reflected (non-persistent)
Code affects individual users in their local Web browser
Stored (persistent)
Code is stored on a server and affects all users who visit the served page
248
PUBLI C

Cross-Site Scripting Recommendation
Since there are currently no libraries provided by the standard SAP HANA XS API to provide proper escaping, the
best solution for generating HTML on SAP HANA XS is to use the ESAPI JavaScript libraries as a starting point. In
addition, we recommend not to write custom interfaces but to rely on well-tested technologies supplied by SAP,
for example, OData or JSON together with SAPUI5 libraries.
Related Links
10.2.2.4 Server-Side JavaScript: Broken Authentication
for (and risk of) attack against authentication infrastructure. Leaks or flaws in the authentication or session
management functions allow attackers to impersonate users and gain access to unauthorized systems and data.
Authentication Problem
Leaks or flaws in the authentication or session management functions allow attackers to impersonate users; the
attackers can be external as well as users with their own accounts to obtain the privileges of those users they
impersonate.
Authentication Recommendation
Use the built-in SAP HANA XS authentication mechanism and session management (cookies). For example, use
the "authentication" keyword to enable an authentication method and set it according to the authentication
method you want implement, for example: SAP logon ticket, form-based, or basic (user name and password) in
the application's .xsaccess file, which ensures that all objects in the application path are available only to
authenticated users.
Related Links
10.2.2.5 Server-Side JavaScript: Insecure Object Reference
for (and risk of) attacks using insecure references to objects.
PUBLI C

Object Reference Problem
An SAP HANA XS application is vulnerable to insecure direct object reference if the application lacks the proper
authentication mechanism for target objects.
Object Reference Recommendation
Make sure that only authenticated users are allowed to access a particular object. In the context of SAP HANA XS,
use the "authentication" keyword to enable an authentication method and set it according to the
authentication method you implement, for example: SAP logon ticket, form-based, or basic (user name and
password) in the application's .xsaccess file, which ensures that all objects in the application path are available
only to authenticated users.
Related Links
10.2.2.6 Server-Side JavaScript: Cross-Site Request Forgery
for (and risk of) cross-site request forgery (XSRF). Cross-site scripting is a web-based vulnerability that exploits
the trust boundaries that exist between different websites running in the same web browser session.
Cross-Site Request-Forgery Problem
Since there are no clear trust boundaries between different Web sites running in the same Web-browser session,
an attacker can trick users (for example, by luring them to a popular Web site that is under the attacker's control)
into clicking a specific hyperlink. The hyperlink displays a Web site that performs actions on the visitor's behalf, for
example, in a hidden iframe. If the targeted end user is logged in and browsing using an account with elevated
privileges, the XSRF attack can compromise the entire Web application.
Cross-Site Request-Forgery Recommendation
SAP HANA XS provides a way to include a random token in the POST submission which is validated on the server-
side. Only if this token is non-predictable for attackers can one prevent cross-site, request-forgery attacks. The
easiest way to prevent cross-site, request-forgery attacks is by using the standard SAP HANA XS cookie. This
cookie is randomly and securely generated and provides a good random token which is unpredictable by an
attacker ($.session.getSecurityToken()).
250
PUBLI C

To protect SAP HANA XS applications from cross-site request-forgery (XSRF) attacks, make sure you always set
the prevent_xsrf keyword in the application-acess (.xsaccess) file to true, as illustrated in the following example:
{
"prevent_xsrf" : true
}
The prevent_xsrf keyword prevents the XSRF attacks by ensuring that checks are performed to establish that a
valid security token is available for given Browser session. The existence of a valid security token determines if an
application responds to the client's request to display content. A security token is considered to be valid if it
matches the token that SAP HANA XS generates in the backend for the corresponding session.
Note
The default setting is false, which means there is no automatic prevention of XSRF attacks. If no value is
assigned to the prevent_xsrf keyword, the default setting (false) applies.
The following server-side JavaScript code snippet show how to use the HTTP request header to fetch, check, and
apply the XSRF security token required to protect against XSRF attacks.
<html>
<head>
<title>Example</title>
<script id="sap-ui-bootstrap" type="text/javascript"
src="/sap/ui5/1/resources/sap-ui-core.js"
data-sap-ui-language="en"
data-sap-ui-theme="sap_goldreflection"
data-sap-ui-libs="sap.ui.core,sap.ui.commons,sap.ui.ux3,sap.ui.table">
</script>
<script type="text/javascript" src="/sap/ui5/1/resources/jquery-sap.js"></
script>
<script>
function doSomething() {
$.ajax({
url: "logic.xsjs",
type: "GET",
beforeSend: function(xhr) {
xhr.setRequestHeader("X-CSRF-Token", "Fetch");
},
success: function(data, textStatus, XMLHttpRequest) {
var token = XMLHttpRequest.getResponseHeader('X-CSRF-Token');
var data = "somePayLoad";
$.ajax({
url: "logic.xsjs",
type: "POST",
data: data,
beforeSend: function(xhr) {
xhr.setRequestHeader("X-CSRF-Token", token);
},
success: function() {
alert("works");
},
error: function() {
alert("works not");
}
});
}
});
}
</script>
</head>
<body>
<div>
PUBLI C

<a href="#" onClick="doSomething();">Do something</a>
</div>
</body>
</html>
Related Links
10.2.2.7 Server-Side JavaScript: Security Misconfiguration
for (and risk of) attacks against the security configuration in place, for example, authentication mechanisms and
Insecure Configuration Problem
No or an inadequate authentication mechanism has been implemented.
Insecure Configuration Recommendation
Applications should have proper authentication in place, for example, by using SAP HANA built-in authentication
mechanisms and, in addition, the SAP HANA XS cookie and session handling features. Application developers
must also consider and control which paths are exposed by HTTP to the outside world and which of these paths
require authentication.
Related Links
10.2.2.8 Server-Side JavaScript: Insecure Storage
for (and risk of) attacks against the insecure or lack of encryption of data assets.
Storage-Encryption Problem
Sensitive information such as logon credentials is exposed.
252
PUBLI C

Storage-Encryption Recommendation
To prevent unauthorized access, for example, in the event of a system break-in, data such as user logon
credentials must be stored in an encrypted state. Application developers can use the SAP HANA XS API, which
provides a secured key value store.
Related Links
10.2.2.9 Server-Side JavaScript: Missing URL Restrictions
for (and risk of) unauthorized access to URLs.
URL Access Problem
Unauthenticated users have access to URLs that expose confidential (unauthorized) data.
URL Access Recommendation
Make sure you have addressed the issues described in "Broken Authentication and Session Management" and
"Insecure Direct Object References". In addition, check if a user is allowed to access a specific URL before actually
executing the code behind that requested URL. Consider putting an authentication check in place for each
JavaScript file before continuing to send any data back to the client's Web browser.
Related Links
10.2.2.10 Server-Side JavaScript: Transport Layer Protection
for (and risk of) insufficient protection of the transport layer.
Transport Layer Protection Problem
Without transport-layer protection, the user's network traffic can be monitored, and attackers can steal sensitive
information such as logon credentials or credit-card data.
PUBLI C

Transport Layer Protection Recommendation
Turn on transport-layer protection in SAP HANA XS; the procedure is described in the SAP HANA security guide.
Related Links
10.2.2.11 Server-Side JavaScript: Invalid Redirection
for (and risk of) redirection and internal fowarding from the requested Web page.
Invalid Redirection Problem
Web applications frequently redirect users to other pages or use internal forwards in a similar manner.
Sometimes the target page is specified in an invalid (not permitted) parameter. This enables an attacker to
choose a destination page leading to the possibility of phishing attacks or the spamming of search engines.
Invalid Redirection Recommendation
To prevent invalidated redirects or forwards, application developers should validate the requested destination
before forwarding, for example, by checking if the destination is present in a whitelist.
Related Links
10.2.2.12 Server-Side JavaScript: XML Processing Issues
for (and risk of) attacks aimed at the process used to parse XML input and generate the XML output.
XML Processing Problem
There are several potential security issues related to processing XML as input or to generating XML as output. In
addition, problems with related technologies (for example, XSL Transformations or XSLT) can enable the
inclusion of other (unwanted) files.
254
PUBLI C

XML Processing Recommendation
Turn on transport-layer protection in SAP HANA XS; the procedure is described in the SAP HANA security guide.
Bear in mind the following rules and suggestions when processing or generating XML output:
When processing XML coming form an untrusted source, disable DTD processing and entity expansion unless
strictly required. This helps prevent Billion Laugh Attacks (Cross-Site Request Forgery), which can bring
down the processing code and, depending on the configuration of the machine, an entire server.
To prevent the inclusion (insertion) of unwanted and unauthorized files, restrict the ability to open files or
URLs even in requests included in XML input that comes from a trusted source. In this way, you prevent the
disclosure of internal file paths and internal machines.
Ensure proper limits are in place on the maximum amount of memory that the XML processing engine can
use, the amount of nested entities that the XML code can have, and the maximum length of entity names,
attribute names, and so on. This practice helps prevent the triggering of potential issues.
Related Links
10.2.3 Writing Server-Side JavaScript Application Code
SAP HANA Extended Application Services (SAP HANA XS) supports server-side application programming in
JavaScript. The server-side application you develop can use a collection of JavaScript APIs to expose authorized
data to client requests, for example, to be consumed by a client GUI such as a Web browser including SAPUI5
applications and mobile clients.
Since JavaScript programs are stored in the SAP HANA repository, the steps in this task description assume that
you have already created a workspace and a project (of type XS Project), and that you have shared the project
with other members of the development team. To write a server-side JavaScript application, you must perform
the following high-level steps.
1. Create a root package for your application, for example, helloxsjs.
2. Create an application descriptor for your application and place it in the root package you created in the
previous step.
The application descriptor is the core file that you use to describe an application's availability within SAP
HANA Extended Application Services. The application-descriptor file has no contents and no name; it only has
the file extension .xsapp.
Note
For backward compatibility, content is allowed in the .xsapp file but ignored.
3. Create an application-access file and place it in the package to which you want to grant access.
The application-access file does not have a name; it only has the file extension .xsaccess. The contents of
the .xsaccess file must be formatted according to JavaScript Object Notation (JSON) rules and associated
with the package the file belongs to. The rules defined in the .xsaccess file apply to the package it resides in
as well as any subpackages lower in the package hierarchy.
4. If you used the authorization keyword in the application-access file (.xsaccess) file for your application,
create an application-privileges file for the application and define the application privileges.
PUBLI C

The application-privileges file does not have a name; it only has the file extension .xsprivileges. The
contents of the .xsprivileges file must be formatted according to JavaScript Object Notation (JSON)
rules. Multiple .xsprivileges files are allowed, but only at different levels in the package hierarchy; you
cannot place two .xsprivileges files in the same application package. The privileges defined in
a .xsprivileges file are bound to the package to which the file belongs and can only be applied to this
package and its subpackages.
Note
The .xsprivileges file lists the authorization levels available for granting to an application package;
the .xsaccess file defines which authorization level is assigned to which application package.
5. Create the server-side JavaScript files that contain the application logic.
You can use the available JavaScript APIs (for example, the database or request-processing API) to expose
authorized data to client requests.
Server-side JavaScript files have the file suffix .xsjs, for example, hello.xsjs and contain the code that is
executed when SAP HANA XS handles a URL request.
$.response.setBody( "Hello, World!");
6. Check the layout workspace.
Your application package structure should have a structure that looks like the following example:
.
\
helloxsjs
\
.xsapp
.xsaccess
hello.xsjs
8. View the results.
The SAP HANA XS Web server enables you to view the results immediately after activation in the repository,
for example: http://dbhost:80<DB_Instance_Number>/helloxsjs/hello.xsjs
10.3 Server-Side JavaScript Libraries
The elements defined in normal server-side JavaScript programs cannot be accessed from other JavaScript
programs. To enable the reuse of program elements, SAP HANA Extended Application Services support server-
side JavaScript libraries.
Server-side JavaScript libraries are a special type of JavaScript program that can be imported and called in other
JavaScript programs. You can use JavaScript libraries to perform simple, repetitive tasks, for example, to handle
forms and form date, to manipulate date and time strings, to parse URLs, and so on.
Note
JavaScript libraries are internally developed extensions for SAP HANA.
256
PUBLI C

The following example shows how to import a JavaScript mathematics library using the import function:
// import math lib
$.import("sap.myapp.lib","math");
// use math lib
var max_res = $.sap.myapp.lib.math.max(3, 7);
The import function requires the following parameters:
Package name
Full name of the package containing the library object you want to import, for example, sap.myapp.lib
Library name
Name of the library object you want to import, for example, math
10.3.1 Writing Server-Side JavaScript Libraries
JavaScript programs. You can use JavaScript libraries to perform simple, repetitive tasks, for example, to handle
forms and form date, to manipulate date and time strings, to parse URLs, and so on.
JavaScript libraries are internally developed extensions for SAP HANA. However, you can write your own libraries,
too. JavaScript libraries exist in the context of a package, which is referenced when you import the library. To
write a JavaScript library to use in your server-side JavaScript application, perform the following steps:
1. Create the file that contains the JavaScript library you want to add to the package and make available for
import.
In SAP HANA XS, server-side JavaScript libraries have the file extension .xsjslib, for example
greetLib.xsjslib.
The following example creates a simple library that displays the word Hello along with a supplied name and
adds an exclamation point (!) as a suffix.
var greetingPrefix = "Hello, ";
var greetingSuffix = "!";
function greet (name) {
return greetingPrefix + name + greetingSuffix;
}
2. Save the new JavaScript library.
It is important to remember where the JavaScript library is located; you have to reference the package path
when you import the library.
3. Activate your new library in the repository so that it is available for import by other JavaScript applications.
10.3.2 Importing Server-Side JavaScript Libraries
JavaScript programs. You can use JavaScript libraries to perform simple, repetitive tasks, for example: handle
forms and form date, manipulate date and time strings, parse URLs, and so on.
PUBLI C

JavaScript libraries are internally developed extensions for SAP HANA. The libraries exist in the context of a
package, which is referenced when you import the library. The following example of a JavaScript library displays
the word "Hello" along with a name and an exclamation mark as a suffix.
var greetingPrefix = "Hello, ";
var greetingSuffix = "!";
function greet (name) {
return greetingPrefix + name + greetingSuffix;
}
Note
This procedure uses the illustrated example JavaScript library to explain what happens when you import a
JavaScript library, for example, which objects are created, when, and where. If you have your own library to
import, substitute the library names and paths shown in the steps below as required.
To import a JavaScript library for use in your server-side JavaScript application, perform the following tasks
1. Import the JavaScript library into a JavaScript application.
Open the server-side JavaScript file into which you want to import the JavaScript library.
Use the $.import function, as follows:
$.import("<package_your_library_was_deployed>","greetLib");
var greeting = $.path.to.your.library.filename.greet("World");
$.response.setBody(greeting);
2. Save and activate the changes to the JavaScript file.
Although the operation is simple, bear in mind the following points:
Additional objects in the package hierarchy
The import operation generates a hierarchy of objects below $ that resemble the library's location in the
repository, for example, for the library path/to/your/library/greetLib.xsjslib, you would see
the following additional object:
$.path.to.your.library.greetLib
Additional properties for the newly generated library object:
$.path.to.your.library.greetLib.greet()
$.path.to.your.library.greetLib.greetingSuffix
$.path.to.your.library.greetLib.greetingPrefix
Pre-import checks:
It is not possible to import the referenced library if the import operation would override any
predefined runtime objects.
Do not import the referenced library if it is already present in the package.
Library context
Imported libraries exist in the context defined by their repository location.
258
PUBLI C

10.4 Server-Side JavaScript APIs
SAP HANA Extended Application Services (SAP HANA XS) provides a set of server-side JavaScript application
programming interfaces (API) that enable you to configure your applications to interact with SAP HANA.
The SAP HANA XS API Reference lists all the functions that are available for use when programing interaction
between your application and SAP HANA. For example, you can use the database API to invoke SQL statements
from inside your application, or access details of the current HTTP request for SAP HANA data with the request-
processing API.SAP HANA XS includes the following set of server-side JavaScript APIs:
Database API
Enables access to the SAP HANA by means of SQL statements. For example, you can open a connection to
commit or rollback changes in SAP HANA, to prepare stored procedures (or SQL statements) for execution or
to return details of a result set or a result set's metadata.
Request-Processing API
Enables access to the context of the current HTTP request, for example, for read requests and write
responses. You can use the functions provided by this API to manipulate the content of the request and the
response.
Database API
The following example shows how to use the database API to prepare and execute an SQL statement. The
response to the SQL query is then prepared and displayed in a Web browser.
var pstmt = conn.prepareStatement( "select * from DUMMY" );
if (!rs.next()) {
$.response.setBody( "Failed to retreive data" );
} else {
$.response.setBody("Response: " + rs.getString(1));
}
rs.close();
pstmt.close();
conn.close();
The following example of usage of the SAP HANA XS database API shows how to establish a connection with SAP
HANA and return a result set that contains the information requested in the specified query (HOST, system ID,
instance number...) in the order specified. After returning the requested information, the connection is closed. If it
not possible to establish a connection with SAP HANA, then display the error message "Unable to retrieve
system." and close the connection.
var conn;
try {
conn = $.db.getConnection();
var query = 'SELECT SYSTEM_ID, HOST, SID, INSTANCE, PORT, USER, PASSWORD,
COMMENT FROM '
+ 'ACME_XSTC.com.acme.adm.xstc.db::SYSTEM WHERE SYSTEM_ID = ?';
PUBLI C

var pstmt = conn.prepareStatement(query);
pstmt.setInteger(1, id);
if (rs.next()) {
return {
id : rs.getInteger(1),
host : rs.getNString(2),
sid : rs.getString(3),
instance : rs.getDecimal(4),
port : rs.getDecimal(5),
user : rs.getNString(6),
password : rs.getNString(7),
comment : rs.getNString(8)
};
}
} catch (ex) {
$.trace.error("Unable to retrieve system. " + ex.toString());
ex.source = fileName;
throw ex;
} finally {
if (conn) {
conn.close();
}
}
Request-Processing API
The following example shows how to use the request-processing API to display the message Hello World in a
browser.
$.response.setBody( "Hello, World !");
In the following example, you can see how to use the request-processing API to get the value of parameters
describing the name and vendor ID of a delivery unit (DU) and return the result set in JSON-compliant form.
var duName = $.request.parameters.get("du_name");
var duVendor = $.request.parameters.get("du_vendor");
result = {
content_id : contentId.toString()
};
$.response.contentType = 'application/json';
$.response.setBody(JSON.stringify(result));
In the following example of use of the request-processing API, we show how to access to the request's meta data
(and body) and, in addition, how to set and send the response.
if($.request.method === $.net.http.GET) {
// get query parameter named id
var qpId = $.request.parameters.get("id");

// handle request for the given id parameter...
var result = handleRequest(qpId);

// send response
$.response.contentType = "plain/test";
$.response.setBody("result: " + result);
260
PUBLI C

} else {
// unsupported method
}
Related Links
SAP HANA XS JavaScript Reference
#unique_211
10.5 The SQL Connection Configuration File
In SAP HANA Extended Application Services (SAP HANA XS), you use the SQL-connection configuration file to
enable the execution of SQL statements from inside your server-side JavaScript application with credentials that
are different to the credentials of the requesting user.
In cases where it is necessary to execute SQL statements from inside your JavaScript application with credentials
that are different to the credentials of the requesting user, SAP HANA XS enables you to define the configuration
for individual SQL connections. Each connection configuration has a unique name, for example, Registration or
AdminConn, which is generated from the name of the corresponding connection-configuration file
(Registration.xssqlcc or AdminConn.xssqlcc) on activation in the repository. The administrator can
assign specific, individual database users to this configuration, and you can use the configuration name to
reference the unique SQL connection configuration from inside your JavaScript application code.
If you want to create an SQL connection configuration, you must create the configuration as a flat file and save the
file with the suffix .xssqlcc, for example, MYSQLconnection.xssqlcc. The new configuration file must be
located in the same package as the application that references it. You can activate repository files at any point in
time.
Note
An SQL connection configuration can only be accessed from a SAP HANA XS JavaScript application (.xsjs)
file that is in the same package as the SQL connection configuration itself. Neither subpackages nor sibling
packages are allowed to access an SQL connection configuration.
The following example shows the composition and structure of a configuration file AdminConn.xssqlcc for an
SAP HANA XS SQL connection called AdminConn. On activation of the SQL connection configuration file
AdminConn.xssqlcc (for example, in the package sap.hana.sqlcon), an SQL connection configuration with
the name sap.hana.sqlcon::AdminConn is created, which can be referenced in your JavaScript application.
sap.hana.sqlcon:AdminConn.xssqlcc
{
"description" : "Admin SQL connection"
}
To create a preconfigured SQL connection using the configuration object AdminConn, for example, from inside
your JavaScript application code, you reference the object using the object name and full package path, as
illustrated in the following code example.
Example
PUBLI C

Calling the SAP HANA XS SQL-Connection Configuration File
function test() {
var body;
var conn;
try {
conn = $.db.getConnection("sap.hana.sqlcon::AdminConn");
var pStmt = conn.prepareStatement("select CURRENT_USER from dummy");
var rs = pStmt.executeQuery();
if (rs.next()) {
body = rs.getNString(1);
}
rs.close();
pStmt.close();
} catch (e) {
body = "Error: exception caught";
$.response.status = $.net.http.BAD_REQUEST;
}
if (conn) {
conn.close();
}
$.response.setBody( body );
}
test();
To use the SQL connection from your application during runtime, you must bind the SQL connection configuration
to a database user. To maintain this user mapping, SAP HANA XS provides a dedicated application (the SQL
Connection Configuration Application), which is pre-installed and accessible under the URL http://
<host>:<port>/sap/hana/xs/sqlcc/. You can grant one of the following roles for users who want to use the
SQL Connection Configuration Application to view or maintain the available SQL connection configurations:
sap.hana.xs.sqlcc::Viewer
Required to display the available SQL Connections and the current user mapping
sap.hana.xs.sqlcc::Administrator
Required to change the user mapping
Troubleshooting Tips
If you are having problems implementing the SQL connection feature using an .xssqlcc configuration, check the
following:
Make sure that you grant the necessary user the activated role (for example,
sap.hana.xs.sqlcc::Administrator). You grant roles (or privileges) in the SAP HANA studio
Administration Console perspective. In the SAP HANA Systems view, choose Security Users The
granting user must have the object privilege EXECUTE on the procedure GRANT_ACTIVATED_ROLE.
Make sure that the SQL-role configuration file (.xssqlcc) you create is located in the same package as the
application that references it.
If your application is using form-based logon (configured in the application's .xsaccess file), make sure the
libxsauthenticator library is present and specified in the list of trusted libraries displayed in the SAP
HANA studio's Administration Console perspective ( Administration Configuration Tab xsengine.ini
application_container application_list . If the libxsauthenticator library is not in the list of authorized
libraries, an SAP HANA system administrator must add it.
262
PUBLI C

Note
If you have to authorize libxsauthenticator, you might also need to refresh the Web page in your
browser the next time you want to access .xssqlcc to display the logon dialog again.
10.6 Connection-language Settings in SAP HANA XS
HTTP requests can define the language used for communication in the HTTP header Accept-Language. This
header contains a prioritized list of languages (defined in the Browser) that a user is willing to accept. SAP HANA
XS uses the language with the highest priority to set the language for the requested connection. The language
setting is passed to the database as the language to be used for the database connection, too.
In server-side JavaScript, the session object's language property enables you to define the language an
application should use for a requested connection. For example, your client JavaScript code could include the
following string:
var application_language = $.session.language = 'de';
Note
Use the language-code format specified in BCP 47 to set the session language, for example: en-US (US
English), de-AT (Austrian German), fr-CA (Canadian French).
As a client-side framework running in the JavaScript sandbox, the SAP UI5 library is not aware of the Accept-
Language header in the HTTP request. Since the current language setting for SAPUI5 is almost never the same as
the language specified in the SAP HANA XS server-side framework, SAPUI5 clients could have problems relating
to text displayed in the wrong language or numbers and dates formatted incorrectly.
The application developer can inform the SAP UI5 client about the current server-side language setting, for
example, by adding an entry to the <script> tag in the SAPUI5 HTML page, as illustrated in the following
examples:
Script tag parameter:
<script id="sap-ui-bootstrap"
type="text/javascript"
data-sap-ui-libs="sap.ui.commons"
data-sap-ui-language="de">
</script>
Global sap-ui-config object:
<script>
window["sap-ui-config"] = {
"language" : "de"
}
</script>
[]
PUBLI C

[]
</script>
The sap-ui-config object must be created and filled before the sap-ui-bootstrap script.
It is important to understand that the session starts when a user logs on, and the specified language is associated
with the session. Although the user can start any number of applications in the session, for example, in multiple
Browser tabs, it is not possible to set a different language for individual applications called in the session,
Setting the Session Language on the Server side
The script tag for the SAPUI5 startup can be generated on the server side, for example, using the
$.session.language property to set the data-sap-ui-language parameter. Applications that have the SAPUI5
<script> tag in a static HTML page can use this approach, as illustrated in the following example:
type="text/javascript"
data-sap-ui-language="$UI5_LANGUAGE$">
</script>
The called XSJS page replaces the $UI5_LANGUAGE$ parameter with the value stored in $.session.language
when loading the static HTML page, as illustrated in the following example:
var objectId = $.repo.createObjectId("","sap.package_name","StaticPage","html");
var indexHtmlObject = $.repo.readObject(activeSession,objectId,activeVersion);

var respTxt = indexHtmlObject.cdata.replace("$UI5_LANGUAGE$", $.session.language);
$.response.setBody(respTxt);
Setting the Session Language with an AJAX Call
You can include an HTTP call in the static HTML page to fetch the correct language from the server using some
server-side JavaScript code, as illustrated in the following example:
<script>
var xmlHttp = new XMLHttpRequest();
xmlHttp.open( "GET", "getAcceptLanguage.xsjs", false );
xmlHttp.send( null );
window["sap-ui-config"] = {
"language" : xmlHttp.getResponseHeader("Content-Language")
}
</script>
</script>
264
PUBLI C

This approach requires an XSJS artifact (for example, getAcceptLanguage.xsjs) that responds to the AJAX
call with the requested language setting, as illustrated in the following example:
$.response.headers.set("Content-Language", $.session.language);
$.response.setBody("");
10.7 Server-Side JavaScript Tracing
The SAP HANA XS server-side JavaScript API provides tracing functions that enable your application to write
predefined messages in the form of application-specific trace output in the xsengine trace files (xsengine*.trc)
according to the trace level you specify, for example, info(information) or error.
If you use the server-side JavaScript API to enable your application to write trace output, you can choose from the
following trace levels:
debug
info
warning
error
fatal
For example, to enable debug-level tracing for your JavaScript application:
$.trace.debug("request path: " + $.request.path);
Note
You can view the xsengine*.trace files in the Diagnosis Files tab page in the Administration perspective of
the SAP HANA studio.
10.7.1 Tracing Server-Side JavaScript Applications
The server-side JavaScript API for SAP HANA XS enables you to activate the writing of trace messages into an
application-specific trace file; the following trace levels are available: debug, error, fatal, info, and warning.
By default, applications write messages of severity level error to the xsengine*.trc trace files; you can increase
the trace level manually, for example, to fatal. In SAP HANA XS, the following steps are required to enable trace
output for your server-side JavaScript application:
2. In the SAP HANA Systems view, double-click the SAP HANA instance to open the Administration view for the
repository where your server-side JavaScript source files are located.
3. Choose the Trace Configuration view.
4. In the Global Database Trace screen area, choose Edit Configuration.
PUBLI C

The Edit Configuration icon is only visible if you have the required privileges on the selected SAP HANA
system.
Note
If the Global Database Trace screen area is not displayed, check that you are using a version of SAP HANA
studio that is compatible (the same as) with the SAP HANA server where you want to set up tracing.
5. Select the Show All Components checkbox.
6. Enter the partial or full name of your application into the search box.
7. Expand the GLOBAL node, if necessary, to view registered traces for the application name returned by the
search operation.
8. Find the trace matching your application name and select the trace level you want to use to generate output.
The application name is the location (package) of the .xsapp file associated with the application you are
tracing. The trace topic is named xsa: <appName>.
9. Choose Finish to activate the trace level changes.
10.7.2 Viewing Server-Side JavaScript Application Trace Files
The server-side JavaScript API for SAP HANA XS enables you to instruct your JavaScript applications to write
application-specific trace messages in the xsengine*.trc trace files, which you can view in the Diagnosis Files
tab page of the Administration perspective in the SAP HANA studio. The following trace levels are available:
debug, error, fatal, info, and warning.
To view trace output for your server-side JavaScript application, perform the following steps:
2. In the Navigator view, double-click the SAP HANA instance to open the Administration view for the repository
where your server-side JavaScript source files are located.
3. Choose the Diagnosis Files tab page.
4. In the Filter box, enter a string to filter the list of search files displayed, for example, xsengine*.trc.
The timestamp displayed in the Modified column does not always reflect the precise time at which the trace
file was written or most recently modified.
5. Locate the trace file for your SAP HANA XS application and doubleclick the entry to display the contents of
the selected trace-file in a separate tab page.
10.8 Debugging Server-Side JavaScript
SAP HANA studio enables you to debug XS JavaScript files, including setting breakpoints and inspecting
variables.
266
PUBLI C

Caution
A port must be opened to enable debugging. For security reasons, the debug port is turned off by default.
1. In a browser, run the XS JavaScript source file that you want to debug.
2. Create a debug configuration for debug sessions for a specific SAP HANA installation.
a) Open the Debug perspective.
b) Choose and select Debug Configurations.
c) Create a new XS JavaScript configuration by double-click XS JavaScript.
d) Enter a name for the configuration.
e) Enter the host and debug port for your SAP HANA system.
f) Select Apply.
g) Select Close.
3. Set breakpoints in the JavaScript code by double-clicking on the left vertical ruler.
PUBLI C

4. Run your debug configuration for your server by choosing and selecting your debug configuration.
5. In the Select Session window, select the SAP HANA XS session ID you are debugging, and then click Select.
The session ID is the value of the xsSessionId cookie in your browser session.
6. Refresh the browser. The XS JavaScript is now running in debug mode. Your client is now attached to the
session, and execution is suspended at the first breakpoint.
You can now perform standard Eclipse debug tasks, such as resuming execution, stepping through execution, and
adding breakpoints. The following debug views are available:
Debug
Breakpoints
Variables
268
PUBLI C

Expressions, which you can use to inspect global variables, such as $.request and other SAP HANA XS
JavaScript API objects
Related Links
Debug Perspective [page 33]
The Debug perspective in the SAP HANA studio is the standard Eclipse Debug perspective, enabling you to start
debug sessions, create breakpoints and watch variables.
Opening a Port for Server-Side JavaScript Debugging [page 269]
10.8.1 Opening a Port for Server-Side JavaScript Debugging
To enable debugging of XS JavaScript code, you must open a debug port on SAP HANA.
The debug port is unsecured, as anyone can attach a debug client without needing credentials. A connected client
can attach to and debug any HTTP session, and can possibly run their own code.
Therefore, it is recommended that you enable this feature only in non-production or development environments
that are accessible by a limited number of developers and that do not contain sensitive and confidential data. In
production systems that contain sensitive or confidential data, it is recommended that you keep debugging
disabled or enable it only temporarily.
To increase security, customers should consider applying additional network-level security measures, for
example, limiting network access to the debug port.
1. In SAP HANA studio, open the Navigator view.
2. Right-click your system, and select Administration from the popup menu.
3. In the Configuration tab, add a section called xsengine.ini debugger (if it does not exist) and add the
following parameters:
enabled = true
listenport = <debug port>
10.8.2 Troubleshooting Server-Side JavaScript Debugging
If the execution of your XS JavaScript code is not stopping at a breakpoint, consider the following solutions.
Make sure the debug port is open on the server.
Make sure you have opened a session with the server by calling an XS JavaScript file from your browser
before starting to debug. Then, when starting to debug, make sure to select the correct session, whose ID is
found in the xsSessionId cookie in your browser session.
Restart your SAP HANA studio with the -clean option, for example:
hdbstudio.exe -clean
PUBLI C

You can tell if this solution is needed by checking whether the SAP HANA studio recognizes the breakpoints
as type SAP HANA XSE Script Breakpoint. To check, select the Breakpoints view menu, then select
Group By Breakpoint Types .
270
PUBLI C

11 Building UIs
11.1 Building UIs with SAPUI5
This section provides introductory information about UI development toolkit for HTML5.
UI development toolkit for HTML5 (SAPUI5) is a user interface technology that is used to build and adapt client
applications based on SAP HANA. You can install SAPUI5 in the SAP HANA studio to build user interfaces
delivered by SAP HANA's Web server.
SAPUI5 runtime is a client-side HTML5 rendering library with a rich set of standard and extension controls and a
lightweight programming model. To support you in developing applications, SAPUI5 application development
tools comes with a set of eclipse-based wizards and editors. SAPUI5 application development tools provides
wizards to create application projects and views according to the model-view-controller concept and other
features like JavaScript code completion, templates and snippets, and in-place application preview.
SAPUI5 provides many features to enable you to easily create and extend state-of-the-art user interfaces. SAPUI5
supports the following features:
RIA-like client-side features based on JavaScript.
CSS3, which allows you to adapt themes to your company's branding in an effective manner
Extensibility concept regarding custom controls, meaning that you can extend existing SAPUI5 controls as
well as develop your own controls
Open source jQuery library used as foundation
Full support of the SAP product standard
Compliance to Open Ajax; can be used together with standard JavaScript libraries
Produced in a release independent code line to enable short shipment cycles
SAPUI5 SDK
The SAPUI5 SDK (Demo Kit) provides the following sections:
Developer Guide with additional information about SAPUI5, the used programming languages, open source
technologies, development tools, and APIs
Controls containing running demo examples with descriptions and source codes
API reference with JavaScript documentation of Framework and Control API
Test Suite, which shows all controls running with different property settings where you can interactively adapt
the controls you use for test purposes
The Demo Kit is installed as a delivery unit. To directly access the Demo Kit, use the following link and replace the
placeholders accordingly: <xsengine protocol>://<xsengine server>:<xsengine port>/sap/ui5/1/sdk/index.html.
The following references to the sections of the Developer Guide in the Demo Kit base on this URL.
Building UIs
PUBLI C

11.1.1 Installing SAPUI5 Application Development Tools
This section provides a short description how to install SAPUI5 tools for SAP HANA.
Before you start the installation of SAPUI5 application development tools, make sure that you comply with the
following requirements:
SAP HANA studio is installed.
SAPUI5 application development tools are downloaded from the SAP Software Download Center on SAP
Service Marketplace (http://service.sap.com/swdc) and search for SAPUI5 TOOLS IDE PLUGIN 1.00.
For more information about the installation of SAPUI5 application development tools, see the SAPUI 5 Installation
Guide for SAP HANA on SAP Service Marketplace or SAP Note 1747308.
11.1.2 Creating an SAPUI5 Application
This section provides an example how to create an SAPUI5 application.
Prerequisites: You have installed the SAPUI5 application development tools in your SAP HANA studio.
Using the tools SAPUI5 provides, you create application projects and views according to the model-view-
controller concept with a clear separation between the user interface and the controller logic.
The following steps describe the generic procedure to create an SAPUI5 application for SAP HANA XS:
1. Create an SAPUI5 application project using SAPUI5 application development tools in Eclipse.
2. Develop the client-side JavaScript script code.
3. If the package that contains your SAPUI5 application project has not been exposed via HTTP already, you
need to create .xsapp and .xsaccess files, see Creating the Application Descriptors. This is not necessary, if
the project was created as a sub-package of an SAP HANA XS application. To make this project a valid SAP
HANA application, add the .xsapp and .xsaccess files.
4. Optional step (not part of the example below): Develop a server-side logic for the project by means of server-
side JavaScript or OData service.
5. Activate the project on SAP HANA's Web server.
6. Execute the application on SAP HANA'S Web server.
For SAPUI5 applications, the following view types are supported:
JavaScript: JSview (file extension: js)
XML: XMLview (file extension: xml)
JSON: JSONview (file extension: json)
The following procedure gives an example for the creation of an SAPUI5 application:
1. To open the wizard for creating an SAPUI5 application, choose File New Other SAPUI5 Application
Development Application Project .
2. Specify the name, for example MyApplication, and location for your project. Select Desktop and Create an
Initial View.
You can also add a view later on by creating a separate SAPUI5 application view. If a new view has been
created for an existing SAPUI5 application project, the view needs to be manually called either from the
index.html page, or from another view via view nesting. For more information, see the Demo Kit -
Developer Guide under Model View Controller Concept.
272
PUBLI C

Building UIs
3. Enter a name for the initial view in the Name field, for example, helloworld. Do not add a file extension. This
is done automatically based on the respective view type.
4. Choose JavaScript and finish the wizard.
The following application parts are created:
View file; in this example, a JSview file
Controller file
index.html file containing the references for the sap.ui.core and sap.ui.commons libraries, the
theme (sap_goldreflection), and information about the script type and the script ID. In a second script
block, the index file refers to the project name, view type, and name. A content default for placing the
controls on the UI later on and the ARIA application role.
5. Change the bootstrap tag (located in the index.html file) to enable the application to access the SAPUI5
libraries on the xsengine as follows: src="resources/sap-ui-core.js" to src="/sap/ui5/1/
resources/sap-ui-core.js".
6. Create a new folder i18n in the WebContent folder. Add a new file messagebundle.hdbtextbundle to the
i18n folder with the following content:
# TRANSLATE
# XBUT,30
MY_BUTTON_TEXT=Hello {0} button
Note
A specific suffix .hdbtextbundle is needed for the resource bundles on SAP HANA (so
called .properties file on other platforms).
7. To add a control to your view (in this example to the helloworld.view.js), insert the following coding:
createContent : function(oController) {
// require the jQuery.sap.resources module
jQuery.sap.require("jquery.sap.resources");
// load the resource bundle
var oBundle = jQuery.sap.resources({
// specify url of the .hdbtextbundle
url : "i18n/messagebundle.hdbtextbundle"
});
var aControls = [];
var oButton = new sap.ui.commons.Button({
id : this.createId("MyButton"),
// access the text using the welcome key and pass the value
// for the placeholder ( {0} ) via an array
text : oBundle.getText("MY_BUTTON_TEXT", [ "World" ])
});
aControls.push(oButton.attachPress(oController.doIt));
return aControls;
}
The coding is put into the createContent method, which creates the content of the view. The View wizard
creates the body of the method.
Note
Building UIs
PUBLI C

In your SAPUI5 coding, refer to the resource bundle with the URL containing the file extension, for
example:
var oBundle = jQuery.sap.resources({
url : "i18n/messagebundle.hdbtextbundle",
locale: sLocale})
var oResourceModel = new sap.ui.model.resource.ResourceModel({
bundleURL : "i18n/messagebundle.hdbtextbundle "})
8. To implement the doIt method for the button's press event, insert the following coding into the controller (in
this example: helloworld.controller.js):
doIt : function(oEvent) { alert(oEvent.getSource().getId() +
"does it!"); }
9. Create a .xsapp and a .xsaccess file.
10. Share, commit, and activate your project to transfer the SAPUI5 application to the SAP HANA repository.
11. To start the UI application on SAP HANA's Web server, use the following URL: <xsengine protocol>://
<xsengine server>:<xsengine port>/MyApplication/WebContent/index.html. Replace the
placeholders accordingly; MyApplication refers to the name of your project according to Step 2.
11.1.2.1 Local Testing of a SAPUI5 Application
This section provides a description how local testing can be performed as an optional step.
If you want to test the SAPUI5 application before you activate your project, you can test it within Eclipse on your
local PC. This local testing is optional. If you perform a local test, note the following:
The src attribute in the bootstrap tag of the index.html file needs to have a different URL than the src
attribute for execution on SAP HANA's Web server. This can be achieved by manually changing the URL
before submitting to SAP HANA's Web server, or by using a code snippet, that detects if the application runs
locally or not, and sets the URL accordingly.
If you call a data service in form of a server-side JavaScript or an OData service that resides in the SAP HANA
box, it is necessary to use a URL that points to SAP HANA's Web server. It is usually also necessary to use the
so-called proxy servlet to avoid cross-domain issues. In case of the bootstrap URL, for example, the URL for
local execution differs from the URL for execution on SAP HANA's Web server. For more information, see the
procedure below and the referenced section in the Demo Kit.
For more information, see the Demo Kit - Developer Guide under Testing a SAPUI5 Application.
To perform a local test, proceed as follows:
1. Change the URL set as "src" in the bootstrap tag of the index.html file (see explanation above). To avoid a
manual change before submitting the application, use the following code snippet that detects the context
automatically, in which the application runs, and sets the URL accordingly.
<script src="/sap/ui5/1/resources/sap-ui-core.js"
id="sap-ui-bootstrap"
data-sap-ui-theme="sap_goldreflection" >
</script>

<script>
274
PUBLI C

Building UIs
// use alternative bootstrap for testing locally in
// Eclipse during development time
// this is only a helper mechanism during development time
if (!window.sap) {
var oriTag = document.getElementById("sap-ui-bootstrap");
oriTag.id = "wrong";
oriTag.src = "wrong.js";
var script = "<script id='sap-ui-bootstrap' " +
"src='resources/sap-ui-core.js' " +
"data-sap-ui-theme='sap_goldreflection' " +
"data-sap-ui-libs='sap.ui.commons'> " +
"<\/script>";
document.write(script);
}
</script>
2. If you use requests to backend services, you need to use a suitable URL that points to the backend server and,
to avoid cross-domain issues, you need to use the so-called proxy servlet. Similar to the bootstrap tag, you
can change the respective code manually before submitting, or you can use a code snippet that determines
the URL automatically depending on the context, in which the application runs.
Note
As the server-side logic of an SAPUI5 application cannot be executed locally on the developer's PC, it can
only run on the server. Hence, this needs to be committed and activated before it can be tested.
For more information, see the Demo Kit - Developer Guide under Testing a SAPUI5 Application Enabling
Backend Access for Local Testing .
3. To test your application, choose one of the following options:
Open the context menu for the index.html file located in your project's WebContent folder and choose
Run As Web App Preview .
To run your application on a server, for example Tomcat, open the context menu of your project, choose
Run As Run on Server and select the respective server.
11.1.2.2 SAPUI5 Runtime Version Comparison
This section describes how to compare the SAPUI5 runtime versions installed locally and on SAP HANA' Web
server.
It is possible that the SAPUI5 runtime version installed locally in your Eclipse differs from the SAPUI5 runtime
version on SAP HANA's Web server.
To verify, which version is installed locally, you can check the SAP HANA studio Eclipse version via Help About
SAP HANA Studio Installation Details Features under "SAPUI5 Feature" (com.sap.ui5.uilib.feature).
To find out the SAPUI5 runtime version on SAP HANA's Web server, open the SAP HANA modeler perspective in
the SAP HANA studio Eclipse via Window Open Perspective Modeler . Choose Setup Delivery Units ...
. The version, support package, and patch version of the SAPUI5 runtime libs are displayed there. You can also
open the SAPUI5 runtime index page under <xsengine protocol>://<xsengine host>:<xsengine port>/sap/ui5/1/
index.html and press CTRL-ALT-SHIFT-P. The SAPUI5 runtime version is displayed in the <SAPUI5 Version>
field.
Building UIs
PUBLI C

If the local version is, for example, newer than the version on SAP HANA's Web server and you use code
completion or test the application locally in Eclipse, the results may differ or you may use features that are not
available on the server. We recommend the following:
To detect new features, check the @since tags provided in the JSDoc.
Before deployment, test the application locally in Eclipse with the runtime resources from the server by
configuring the resource servlet to fetch the runtime resources from the server. For more information, see the
Demo Kit - Developer Guide under Testing an SAPUI5 Application UI5 Library Location Used for Testing .
After deployment, always test the application on the server.
11.1.3 Supporting Translation
This section provides a description of what needs to be done to support the translation of resource bundles.
To enable the translation of the user interface, define a specific suffix for the resource bundles and use a specific
first line in the resource bundle file. This is described in step 6 in the Creating an SAPUI5 Application topic.
For more information, see the Demo Kit - Developer Guide under Localization.
Related Links
Creating an SAPUI5 Application [page 272]
This section provides an example how to create an SAPUI5 application.
11.2 Using UI Integration Services
SAP HANA UI Integration Services is a set of Eclipse-based and browser-based tools, as well as client-side APIs,
which enable you to integrate standalone SAP HANA client applications into Web user interfaces to support end-
to-end business scenarios.
These user interfaces are referred to as application sites. Pre-built standalone SAP HANA client applications that
are integrated into application sites are referred to as widgets.
Prerequisites
You are assigned to the sap.hana.uis.db::SITE_DESIGNER role, and end users are assigned to the
sap.hana.uis.db::SITE_USER role.
For more information, see Granting Privileges to Users [page 437]
The HANA_UI_INTEGRATION_SVC delivery unit is imported and activated.
To import the delivery unit, perform the following steps:
1. In the File menu, choose Import....
2. Select the import source SAP HANA Content Delivery Unit , and choose Next.
3. Select the target system for the delivery unit.
276
PUBLI C

Building UIs
4. Choose the Server option. In the dropdown list of available files, select <path>/
HANA_UI_INTEGRATION_SVC.tgz, and choose Finish.
For more information, see Importing Delivery Units [page 443].
Related Links
Creating an Application Site [page 277]
Before you can start designing an application site, you need to create it in the SAP HANA studio.
Designing an Application Site [page 278]
You can visually design and manage application sites in a browser-based design environment.
Developing Widgets [page 281]
SAP HANA UI Integration Services provides a number of tools and client-side APIs to use when developing
widgets for integration into application sites.
11.2.1 Creating an Application Site
Before you can start designing an application site, you need to create it in the SAP HANA studio.
To create a new application site in the SAP HANA studio:
1. In the project's context menu in Project Explorer, choose New Other .
2. In the New dialog box, choose SAP HANA Development Application Site , and then choose Next.
3. In the New Application Site dialog box, select a parent folder, enter the site properties, and choose Finish.
4. To open the site in the preferred browser once the wizard is completed, select the Open in Browser checkbox.
Note
Application sites are best viewed in Google Chrome or Mozilla Firefox. For convenience, select one of these
browsers as the default browser in SAP HANA studio.
Note
If the Open in Browser checkbox is disabled, it means that the site is not automatically committed to the
repository. To enable this, follow the instructions on screen.
5. If the site is not automatically committed to the repository, commit it by choosing Team Commit from
the site's .xsappsite file context menu.
6. To open the site in browser, double-click the site's .xsappsite file.
Note
If you open the site from its context menu, make sure you choose the default Application Site Editor.
Choosing another editor is not recommended.
7. To make the site available to end users, activate it by choosing Team Activate from the
site's .xsappsite file context menu.
Note
Building UIs
PUBLI C

The URL of the sites end-user version is displayed in the Site Properties pane in the design environment.
Related Links
Accessing Site Properties [page 278]
You can view and modify application site properties.
11.2.2 Designing an Application Site
You can visually design and manage application sites in a browser-based design environment.
In this environment, you can perform the following tasks:
Task Instructions
View and modify site properties Accessing Site Properties [page 278]
Add, organize, or remove pages Managing Pages [page 278]
Add widgets to pages Adding Widgets to Pages [page 279]
Manage widgets on pages Managing Widgets on a Page [page 279]
Choose site layout options Configuring Site Layout [page 280]
You perform most of these tasks in the design panel located to the right of the content area of the screen. By
default, the design panel is collapsed into a side bar with buttons. To expand the panel, click a side bar button
relevant for your current task.
11.2.2.1 Accessing Site Properties
You can view and modify application site properties.
To access the application site's properties, choose in the design panel. Most of the properties are read-only.
The URL property contains the URL of the sites runtime version that is available to end users after the site is
activated.
You can modify the Title or Description properties:
1. Double-click on the property field and type the new name.
2. The change is saved automatically. To discard the change while the field is in focus, press Esc.
11.2.2.2 Managing Pages
You can manage pages of an application site in a browser-based design environment.
To manage pages, choose in the design panel. In the Pages pane that opens, you can perform the following
tasks:
278
PUBLI C

Building UIs
Task Description
Add a page Click the Add box. A new page tab with the default name New Page is added
after the last page.
Rename a page Click the page name that you want to change, and edit it in the text box that
opens. The change is saved automatically. To discard the change while the text
box is in focus, press Esc.
Remove a page
Point the cursor at the page that you want to remove, and choose . If there
are widgets on the page, a confirmation box appears. If you choose OK in the
box, the page is removed along with all its widgets.
Move a page Drag and drop a page name to a required position in the list.
11.2.2.3 Adding Widgets to Pages
You can add available widgets to application site pages.
To add a widget to a page:
1. Open a page by choosing its tab.
2. In the design panel, choose to open the Widgets pane, which contains available widgets.
3. To locate the required widgets, type a widget name or a part of it in the search box. You can also filter the
widgets by folder: from the dropdown box, select a folder in which the widgets were created. The widgets
displayed in the pane match the selection criteria.
4. Double-click the widget of your choice, or drag and drop it onto the current page. Reposition the widgets on
the page as needed.
Related Links
Managing Widgets on a Page [page 279]
You can organize and manage widgets on application site pages.
11.2.2.4 Managing Widgets on a Page
You can organize and manage widgets on application site pages.
You can perform the following tasks with widgets on a page:
Task Description
Remove a widget
In the widget menu, choose Remove, and choose Yes in the
confirmation dialog.
Rename a widget
In the widget menu, choose Rename, and modify the text in the
title bar.
Building UIs
PUBLI C

Task Description
Toggle display of a widget's frame area
In the widget menu, toggle the Display Frame Area checkbox.
Toggle full screen view
Choose the or icon to open or close the full screen view.
Note
This setting is not persisted across sessions.
Resize a widget Drag the borders of the widget to resize it.
Rearrange widgets on a page Drag and drop widgets to the required positions.
11.2.2.5 Configuring Site Layout
You can configure certain site layout options.
You can adjust the appearance and behavior of the shell's navigation bar to the needs of specific applications. For
example, if your application site contains only one widget that has its own navigation controls, you might want to
disable the navigation capabilities of the shell.
1. In the design panel, choose to open the Layout pane.
2. In the Navigation Bar dropdown box, choose the required option:
Option Description
Full Fully functional navigation bar of the standard size
Narrow Fully functional narrow navigation bar
Header Only header with no navigation; suitable for sites with only one page
None No navigation bar at all; suitable for sites containing a single widget with its own header
The selected option takes effect for all end users of the application site once the site is activated.
11.2.3 Creating a Widget
To integrate a standalone SAP HANA XS client application or other Web application into application sites, you
need to create a widget based on this application.
To create an application-based widget in the SAP HANA studio, you need to write an XML specification file that
either references or embeds the client-side code of the application, and run the Create New Widget wizard as
follows:
1. In the project's context menu in Project Explorer, choose New Other... .
2. In the New dialog box, choose SAP HANA Development Widget , and choose Next.
280
PUBLI C

Building UIs
3. In the New Widget dialog box, choose the parent folder, enter the widget file name and specification file name,
and choose Next.
Note
The specification file should be located under the current project's folder.
4. In the next step of the wizard, set the widget properties, and choose Finish.
5. To make the widget available for application sites, activate the widget by choosing Team Activate from
the .xswidget file's context menu.
6. To edit the widget properties, open the .xswidget file, make the required changes, and then activate the
widget, as described previously.
Related Links
Sample Widget Specification Files [page 281]
A widget specification file should either reference the client-side code of an application, or directly embed the
application's HTML code in the <Content> section.
Developing Widgets [page 281]
11.2.4 Developing Widgets
Related Links
Sample Widget Specification Files [page 281]
Site Context API [page 283]
Provides a messaging mechanism that enables you to implement communication between widgets in the same
application site.
GadgetPrefs API [page 287]
Provides methods to persist widget preferences on the server side.
SetTitle [page 290]
A method that can be used to set a widget's title.
11.2.4.1 Sample Widget Specification Files
The following are sample widget specification files:
Building UIs
PUBLI C

References an application:

<?xml version="1.0" encoding="UTF-8" ?>
<Module>
<ModulePrefs title="Calendar">
</ModulePrefs>
<Content type="html" href="/content/applications/calendar/index.html">
</Content>
</Module>
Embeds HTML code of an application:

<?xml version="1.0" encoding="UTF-8" ?>
<Module>
<ModulePrefs title="Color Tester">
<Require feature="gadgetprefs"/>
</ModulePrefs>
<UserPref name="color" default_value="white"/>
<Content type="html">
<![CDATA[

<html>
<head>
<script language="javascript" type="text/javascript">
var gadgetPrefs = new gadgets.GadgetPrefs();
window.onload = initiateGadgetPrefs;
function initiateGadgetPrefs() {
setBGcolor();
}
function setColor() {
var selectedColor = document.getElementById('color').value;
gadgetPrefs.setPreference('color', selectedColor);
setBGcolor();
}
function setBGcolor() {
var currentColor = gadgetPrefs.getPreference("color");
document.getElementById('main').setAttribute("bgcolor",
currentColor);
}
</script>
</head>
<body id="main" bgcolor="#F290F0" style="font-family:Arial;">
<label style="font-weight: bold ">Select background color: </
label><br>
Color: <input type="text" name="color" id='color'/>
<input type="button" value="Change Color" onClick="setColor();"/
><br>
</body>
</html>
]]>
</Content>
</Module>

282
PUBLI C

Building UIs
11.2.4.2 Site Context API
Provides a messaging mechanism that enables you to implement communication between widgets in the same
application site.
A widget can publish a message to the dedicated context object so that other widgets can subscribe a callback
function to this message.
To use the API, declare this feature in the <ModulePrefs> section of the widget's specification file:
<ModulePrefs>
<Require feature="sap-context"/>
</ModulePrefs>
The Site Context API provides the following methods:
publish(key, value): Publish [page 283]
subscribe(callback): Subscribe [page 284]
unsubscribe(subscriptionID): Unsubscribe [page 285]
11.2.4.2.1 Publish
A method of the Site Context API that can be used by a widget to publish a message to other widgets.
publish(key, value)
Use to publish messages to the sapcontext object. A message is a key-value pair that is stored in the context.
The updated context is sent to all subscribed callbacks as a parameter.
A widget can publish multiple values with the same key. Each time a widget publishes a different value with the
same key, the new value replaces the previous one.
Parameters
Parameter Type Description
key String The key of the published pair
value String The value of the published pair
Building UIs
PUBLI C

Example
<script language ="JavaScript">
...
gadgets.sapcontext.publish(Country, France);
...
</script>
Related Links
Subscribe [page 284]
A method of the Site Context API that can be used by a widget for subscribing to messages published by other
widgets.
Unsubscribe [page 285]
A method of the Site Context API that can be used by a widget to remove an existing subscription.
11.2.4.2.2 Subscribe
widgets.
subscribe(callback)
Use for subscribing widgets to messages that are published to the sapcontext context object. The message is a
key-value pair that is stored in the context. The updated context is sent to all subscribed callbacks as a parameter.
Parameters
callback Function
The callback function that is called
in response to the published
message.
Receives two parameters: topic and
context.
The topic parameter always equals
sap-context.
284
PUBLI C

Building UIs
Returns
Subscription ID
Example
...

var callback = function (topic, context) {
var color = context.getPropertyByKey(widget-color);
paintBackground(color);
}
function subscribe() {
var subID = gadgets.sapcontext.subscribe(callback);
}

//To subscribe a widget immediately after it is loaded,
// register the subscribe function using gadgets.HubSettings.onConnect
gadgets.HubSettings.onConnect = subscribe;
...
</script>
You can also call the subcribe method at any point directly.
Related Links
Publish [page 283]
Unsubscribe [page 285]
11.2.4.2.3 Unsubscribe
unsubscribe(subscriptionID)
Use to remove an existing subscription.
Building UIs
PUBLI C

Parameters
subscriptionID Int ID of the subscription to remove
Example
...
...
gadgets.sapcontext.unsubscribe(subID);
...
</script>
Related Links
Subscribe [page 284]
widgets.
Publish [page 283]
11.2.4.2.4 Enabling Communication Between Widgets
You can enable widgets, running in the same application site, to pass data to each other so that their content and
behavior change dynamically according to this data.
The following code samples illustrate how to implement communication between two widgets by using the
publish-subscribe messaging mechanism provided by the Site Context API.
The first widget displays a list of sales orders. Once the user selects a sales order from the list, the second widget
displays the details of the order.
First (source) widget:
// Publish event when user selects a sales order from the list
function onSalesOrderSelected(var salesOrderID){
gadgets.sapcontext.publish("selectedSalesOrderID", salesOrderID);
}
</script>
Second (target) widget:
// Callback function to display order details
286
PUBLI C

Building UIs
function callback (topic, context) {
var salesOrderID =
context.getPropertyByKey(selectedSalesOrderID);
showSalesOrderDetails(salesOrderID);
}
function subscribe() {
}
// To subscribe a widget immediately after it is loaded,
// register the subscribe function using gadgets.HubSettings.onConnect
gadgets.HubSettings.onConnect = subscribe;
</script>
11.2.4.3 GadgetPrefs API
Provides methods to persist widget preferences on the server side.
The GadgetPrefs API enables you to persist modified widget preferences on the server side. Persisted
preferences of a widget override its default preferences. At design time, a single set of preferences is persisted for
a widget, whereas at runtime a separate set of widget preferences is persisted for each user.
To use the API, declare this feature under the <ModulePrefs>
tag of the widget's specification file:
<ModulePrefs>
<Require feature="gadgetprefs"/>
</ModulePrefs>
The GadgetPrefs API contains the following methods:
getPreference(key): GetPreference [page 287]
setPreference(key,value): SetPreference [page 289]
11.2.4.3.1 GetPreference
Methods of the GadgetPrefs API that can be used to retrieve a persisted preference of a widget.
getPreference
getPreference(key)
Retrieves the value of a preference specified by its key.
Building UIs
PUBLI C

Parameters
key String Key of a preference
Returns
The value assigned to the specified key, or undefined, if the key is not found.
getBool
getBool(key)
Retrieves the value of the specified preference as Boolean.
Returns
The Boolean value of the preference, or false, if no value is set or if the key is not found.
getString
getString(key)
Retrieves the value of the specified preference as string.
Returns
The string value of the preference, or an empty string, if no value is set or if the key is not found.
288
PUBLI C

Building UIs
Example
...
var val = gadgetPrefsApi.getPreference(key);
var boolVal = gadgetPrefsApi.getBool(key);
if (boolVal){
var strVal = gadgetPrefsApi.getString(key);
}
else
{
var strVal = Value not set;
}...
</script>
Related Links
SetPreference [page 289]
A method of the GadgetPrefs API that can be used to set a preference of a widget.
11.2.4.3.2 SetPreference
A method of the GadgetPrefs API that can be used to set a preference of a widget.
setPreference(key,value)
Sets a preference of a widget by defining a key-value pair.
Parameters
value String Value of the key
Example
...
var gadgetPrefsApi= new gadgets.GadgetPrefs();
gadgetPrefsApi.setPreference (key1,value1);
...
</script>
Building UIs
PUBLI C

Related Links
GetPreference [page 287]
Methods of the GadgetPrefs API that can be used to retrieve a persisted preference of a widget.
11.2.4.4 SetTitle
A method that can be used to set a widget's title.
setTitle(title)
Use to set the title of a widget in the current session. The title is not persisted across sessions.
To use the API, declare this feature under the <ModulePrefs> tag of the widget's specification file:
<ModulePrefs>
<Require feature="settitle"/>
</ModulePrefs>
Parameters
title String Title of a widget
Example
...
var newTitle="Hello World";
gadgets.window.setTitle(newTitle);
...
</script>
290
PUBLI C

Building UIs
12 Enabling Search
With a SAP HANA database, your users will want to search tables and views much like they would when searching
for information on the Internet. In SAP HANA, you can either directly query data using SQL queries or you can
build search apps using a UI toolkit.
Before enabling search, you must meet the following prerequisites:
Your SAP HANA database contains column-oriented tables.
A valid data type is assigned to each column. The data types of the columns determine how you can query
your data.
In SAP HANA, you can search on single or multiple columns of almost any visible data type. In addition to standard
string search, SAP HANA also supports full text search.
During a full text search, the SAP HANA search engine examines structured text, such as author and date
attributes, as well as unstructured text, such as body text. Unlike a string search, for a text search, the sequence
of words and characters is not critical for finding matches. A full text index enables this functionality by analyzing
and preprocessing the available text semantically. This includes normalization, tokenization, word stemming, and
parts of speech tagging.
To enable search, proceed as follows:
1. Create any required full text indexes for the columns in the table.
For columns defined with the data type TEXT or SHORTTEXT(n), full text indexes are automatically
generated. For columns of any other data type, you must manually create any required full text indexes.
When you create a full text index, you can also define synchronization and trigger text analysis.
2. Build SQL search queries.
Search queries use the SQL SELECT statement.
For searches on large object types or text, the queries must include the CONTAINS predicate.
3. Optionally, build search apps using the UI toolkit for SAP HANA Info Access .
The UI toolkit provides UI building blocks for developing browser-based search apps for end users.
The UI toolkit is based on HTML5 and JavaScript libraries.
The UI toolkit connects to the database using the SAP HANA Info Access service that wraps search and
analytic SQL queries and exposes them through an HTTP interface.
12.1 Creating Full Text Indexes
When you create a TEXT or SHORTTEXT column in a table, SAP HANA automatically creates a corresponding full
text index. However, for columns of other data types, you must manually create and define any necessary full text
indexes.
A full text index is an additional data structure that is created to enable text search features on a specific column
in a table. Conceptually, full text indexes support searching on columns in the same way that indexes support
searching through books.
To create a full text index, proceed as follows:
Enabling Search
PUBLI C

1. Determine the columns for which you require an index.
2. Use the CREATE FULLTEXT INDEX statement to create an index with the specified index name.
CREATE FULLTEXT INDEX <index_name> ON <tableref> '(' <column_name> ')'
[<fulltext_parameter_list>]
3. Specify any of the following additional parameters for the full text index:
LANGUAGE COLUMN <column_name>
LANGUAGE DETECTION '(' <string_literal_list> ')'
MIME TYPE COLUMN <column_name>
FUZZY SEARCH INDEX <on_off>
PHRASE INDEX RATIO <on_off>
CONFIGURATION <string_literal>
SEARCH ONLY <on_off>
FAST PREPROCESS <on_off>
TEXT ANALYSIS <on_off>
SYNC
|ASYNC
|ASYNC FLUSH [QUEUE] EVERY n MINUTES
|ASYNC FLUSH [QUEUE] AFTER n DOCUMENTS
|ASYNC FLUSH [QUEUE] EVERY n MINUTES OR AFTER m DOCUMENTS
If you do not specify any parameters, the default values are used.
The system creates a separate hidden full text index column for each source column that you have specified.
You can now create queries to search those columns.
You can check the parameters of an existing full text index by using the SYS.FULLTEXT_INDEXES monitoring
view.
Example
You want to create a full text index i1 for table A, column C, with the following characteristics:
Synchronous processing
Fuzzy search index disabled
Languages for language detection: English, German, and South Korean
To create the index, you use the following syntax:
CREATE FULLTEXT INDEX i1 ON A(C) FUZZY SEARCH INDEX OFF
SYNC
LANGUAGE DETECTION ('EN','DE','KR')
Related Links
Full Text Index Types [page 293]
SAP HANA automatically creates full text indexes for columns of type TEXT and SHORTTEXT(n). For other
column types, you must manually create any required full text indexes.
Synchronization [page 296]
Full text indexes in a SAP HANA database must be created and updated in synchronization with the
corresponding columns. This synchronization can be either synchronous or asynchronous.
Full Text Index Parameters [page 304]
The content and behavior of a full text index is configured by the use of both default and user-specified
parameters. To view the configuration of a full text index, you use the SYS.FULLTEXT_INDEXES view.
292
PUBLI C

Enabling Search
12.1.1 Full Text Index Types
SAP HANA automatically creates full text indexes for columns of type TEXT and SHORTTEXT(n). For other
column types, you must manually create any required full text indexes.
Characteristic TEXT SHORTTEXT (n) Manually Created
SQL data type exposed to
user
NCLOB NVARCHAR(n) Original data type
Data returned by SELECT Original data (returns
normalized data if the
SEARCH_ONLY parameter
is ON)
Original data Original data
SQL insertion mode SYNC SYNC SYNC or ASYNC
Drop index Yes (via drop column) Yes (via drop column) Yes
Text search via CONTAINS Yes Yes Yes
SQL string search Not possible Possible Depends on underlying
data type
Change parameters of full
text index
Partially Partially All (with rebuild)
Rebuild index No No Yes
Base type can be changed No No No (no dependency
between base-column and
index available)
12.1.1.1 TEXT Indexes
In a SAP HANA database, when you create a table that contains large text documents, you can define the columns
with the TEXT data type. This data type allows you to query large text documents and present content excerpts in
search hit lists. You can also reconstruct the document and display it in its original formatting.
When you create a TEXT column and insert content, SAP HANA extracts and processes the text from the original
document and then automatically generates a full text index for the column. To create this full text index, SAP
HANA replaces the original data in the column with the processed text. This text is then returned with the data
type NCLOB. The original data is no longer available.
If you insert new entries in the TEXT column, the full text index is automatically and synchronously updated.
Example
CREATE COLUMN TABLE <tablename>
(
k int primary key,
content TEXT
FAST PREPROCESS OFF
Enabling Search
PUBLI C

PHRASE INDEX RATIO 0.77
)
Limitations
The TEXT data type has the following search-relevant limitations:
SQL string searches are not supported.
The SQL functions CONCAT and JOIN are not supported.
TEXT columns cannot be converted to other data types.
Text analysis is not supported.
For TEXT columns, SAP HANA does not support the following SQL expressions:
HAVING
WHERE with strings or non-alphanumeric characters
ORDER BY
GROUP BY
Aggregate expressions (COUNT, MIN, MAX, etc.)
JOIN ON
Changes to TEXT Indexes
TEXT full text indexes are automatically generated and you do not specify names for them; therefore, you cannot
directly manipulate them. However, when you create, alter, or drop a table column, the same change is
automatically applied to the full text index for that column.
By using the ALTER TABLE statement to affect changes on the index, you can alter the following parameters:
PHRASE INDEX RATIO
FUZZY SEARCH INDEX
12.1.1.2 SHORTTEXT(n) Indexes
If the tables in your SAP HANA database contain columns with text strings that are relatively short in length, you
can define those columns with the SHORTTEXT(n) data type. The SHORTTEXT(n) data type enables both SQL
string search and full text search capabilities.
SAP HANA preprocesses the text in the column and stores that preprocessed text as a full text index in a hidden
column attached to the original column. When queried, the text in the full text index returns with the NVARCHAR
data type. The original text is still available; however, search queries are performed, by default, on the text in the
index.
294
PUBLI C

Enabling Search
When you create a column table and define a column with the data type SHORTTEXT(n), as in the following
example, a full text index is automatically generated. Whenever new entries are then inserted in the column, the
full text index is automatically and synchronously updated.
Example
CREATE COLUMN TABLE <tablename>
(
k int primary key,
content SHORTTEXT(100)
FAST PREPROCESS OFF
SEARCH ONLY ON
)
Changes to SHORTTEXT(n) Indexes
SHORTTEXT(n) full text indexes are automatically generated and you do not specify names for them; therefore,
you cannot directly manipulate them. However, when you create, alter, or drop a table column, the same change
is automatically applied to the index for that column.
When using the ALTER TABLE statement to affect changes on the index, you can only alter the following
parameters:
PHRASE INDEX RATIO
FUZZY SEARCH INDEX
Note
You cannot change the length of the original text and you cannot convert SHORTTEXT(n) to another data type.
12.1.1.3 Manually Created Indexes
If the tables in your SAP HANA database contain extensive columns that are frequently queried but do not have
automatically generated full text indexes, meaning they are not TEXT or SHORTTEXT(n) type columns, you can
improve search performance by manually creating full text indexes.
To manually create a full text index, the column must have one of the following SQL data types:
VARCHAR
NVARCHAR
ALPHANUM
CLOB
NCLOB
BLOB
When you manually create an index, the system attaches a hidden column to the specified column. This hidden
column contains textual data that SAP HANA Preprocessor has extracted from the text in the source column. The
original text in the source column remains unchanged. Search queries are then performed on the hidden column;
Enabling Search
PUBLI C

however, they always return the original text. Depending on the data type that is assigned to a source column,
string search may be possible.
You can manually create an index directly after creating a table or you can create the index later. Once you create
an index for a column, the system automatically processes any text that is inserted into this column and adds the
processed text to the index. Processing for manually created indexes can be performed synchronously or
asynchronously.
You can specify different parameters when you create a full text index. If parameter changes are required later,
you can change the values for the existing index directly or re-create the index with the parameters that you want
to change.
Related Links
Creating Full Text Indexes [page 291]
indexes.
Altering Full Text Index Parameters [page 303]
You can alter a full text index after it is created. Altering an index includes changing the values of the parameters
and altering the parameters by replacing the index.
12.1.2 Synchronization
Synchronous
In SAP HANA, indexes of the type TEXT and SHORTTEXT(n) are synchronous. Text preprocessing is
automatically performed when a column is created or new text is inserted and the full text index is then
automatically updated. The system cannot transport any data to the proper database tables until text
preprocessing is complete.
Asynchronous
If you manually create a full text index, you can specify whether the index is synchronous or asynchronous. By
default, manually created indexes are asynchronous. Text preprocessing is not initially performed when the table
296
PUBLI C

Enabling Search
or column is created or whenever new text is inserted. In this case, inserting the results of the text preprocessing
and writing the original data do not occur at the same time. Therefore, the full text information may not be
immediately available for searching.
To handle asynchronous processing of text, SAP HANA uses queues.
12.1.2.1 Queues
The queue is a mechanism used to enable a full text index to operate asynchronously. This means that when you
insert new entries into the column, the text is not made available in the column until it is preprocessed.
When you insert new entries, the queue sends the text to the preprocessor for analysis. It returns a serialized
instance of a DAF (document analysis format) object, which is then processed further by the HANA column store.
The result is stored in the full text index.
The SAP HANA queue manager automatically creates a queue when you create an asynchronous full text index or
when the index server is started and the queue manager finds the information that a specific queue is needed. The
queues are always created on the server on which the table is stored.
Every entry in the queue has one of the following processing states:
New
Preprocessing
Preprocessed
Indexing
Error
If the original column entry is modified or deleted during text processing, the queue is notified and, if necessary,
the entry is preprocessed again.
Note
The content of the queue is not made persistent at any stage. If the HANA index server process fails, the queue
data is lost and the queue manager automatically restarts the process for those entries that were not already
processed. Any incomplete text preprocessing is restarted from the beginning.
Flush Scheduling
When you create an asynchronous full text index, you can specify when documents are removed from the queue
after they are preprocessed and inserted into the full text index; this is called flushing. You can schedule flushing
based on either time or the number of documents. To do this, when you create the full text index, define one of the
following clauses with the ASYNC parameter:
FLUSH EVERY (n) MINUTES
FLUSH AFTER (n) DOCUMENTS
FLUSH EVERY (n) MINUTES OR AFTER (m) DOCUMENTS
Note
Enabling Search
PUBLI C

You cannot specify negatives values for minutes or documents.
To determine when the queue of an existing full text index is flushed, see the FLUSH_EVERY_MINUTES and
FLUSH_AFTER_ROWS attributes in the view FULLTEXT_INDEXES.
Related Links
Manipulating Queue Processing [page 298]
By default, queues are active initially and run automatically based on the parameters you specify when creating
the full text index. However, if necessary, you can manually manipulate the processing of an existing queue.
12.1.2.2 Manipulating Queue Processing
By default, queues are active initially and run automatically based on the parameters you specify when creating
the full text index. However, if necessary, you can manually manipulate the processing of an existing queue.
To manipulate the processing of a queue, the following commands are available:
FLUSH
Updates the full text index with the documents in the queue which have already been processed and removes
them from the queue.
SUSPEND
Suspends the full text index processing queue
ACTIVATE
Activates the full text index processing queue if it has been suspended
To manipulate the processing of a queue:
1. Identify which queue process you want to manipulate by using the monitoring view M_FULLTEXT_QUEUES. For
information about the specific content of the view, see SAP HANA System Tables and Monitoring Views.
2. Use the ALTER FULLTEXT INDEX statement to flush, suspend, or reactivate the queue.
Use the following syntax:
ALTER FULLTEXT INDEX <index name> FLUSH|SUSPEND|ACTIVATE QUEUE
Related Links
SAP HANA System Tables and Monitoring Views: M_FULLTEXT_QUEUES
12.1.3 Text Analysis
Text analysis provides a vast number of possible entity types and analysis rules for many industries in 20
languages. However, you do not have to deal with this complexity when analyzing your individual set of
documents. The language modules included with the software contain system dictionaries and provide an
extensive set of predefined entity types. The extraction process can extract entities using these lists of specific
entities. It can also discover new entities using linguistic models. Extraction classifies each extracted entity by
entity type and presents this metadata in a standardized format.
The following data types are enabled for text analysis: NVARCHAR, VARCHAR, and STRING.
298
PUBLI C

Enabling Search
The following text analysis options are delivered by SAP:
Table 12: Options
Name of Option Description
LINGANALYSIS_BASIC This option provides the following language processing
capabilities for linguistic analysis of unstructured data:
Segmentation - the separation of input text into its elements
LINGANALYSIS_STEMS This option provides the following language processing
Stemming - the identification of word stems or dictionary
forms
LINGANALYSIS_FULL This option provides the following language processing
Stemming - the identification of word stems or dictionary
forms
Tagging - the labeling of words' parts of speech
EXTRACTION_CORE This option extracts entities of interest from unstructured text,
such as people, organizations, places, and other parties described
in the document.
In most use cases, this option is sufficient.
EXTRACTION_CORE_VOICEOFCUSTOMER Voice of the customer content includes a set of entity types and
rules that address requirements for extracting customer
sentiments and requests. You can use this content to retrieve
specific information about your customers' needs and
perceptions when processing and analyzing text. The option
involves complex linguistic analysis and pattern matching that
includes processing parts of speech, syntactic patterns, negation,
and so on, to identify the patterns to be extracted.
Voice of the customer content is supported for these languages:
English
French
German
Spanish
To use the text analyzing function, create a full text index on the column that contains your texts with the following
parameters:
TEXT ANALYSIS ON
CONFIGURATION '<NAME OF OPTION>'
Note
Enabling Search
PUBLI C

Technical names of the options are case-sensitive.
If your tables contain a language indicator, enter the name of the column:
LANGUAGE COLUMN <NAME OF COLUMN CONTAINING THE LANGUAGE INDICATOR>
If no language is specified, EN will be used by default.
Once the indexing is started, the text analysis runs in the background. Depending on the number and size of the
texts, one analysis can take hours. To check the status of the text analysis, you can use the default monitoring
view FULLTEXT_INDEXES.
For each full text index, the system creates an additional table with the naming $TA_<index_name> within the
same schema that contains the source table.
This table stores the extracted entities and the analysis results. You can use this table to build joins with other
search enabled views, for example to use it for interactive navigation or auto-completion in search input fields.
For detailed information on this table, see Structure of the $TA Table.
To keep track of deletions in the source table, the keys in the $TA table need to be aligned to the keys of the
source table. To do this, use the following SQL statement:
ALTER TABLE "<schema>"."$TA_INDEX_NAME" ADD CONSTRAINT <constraint name> COMMAND
FOREIGN KEY("key_1", "key_2", "key_n") REFERENCES "<schema>"."<name of source
table>"("key_1","key_2","key_n") ON DELETE CASCADE
Example
Use the CREATE FULLTEXT INDEX statement to create an index named CUSTOMER_INDEX on your table
CUSTOMERS to index the customername column: CREATE FULLTEXT INDEX CUSTOMER_INDEX ON
"MY_SCHEMA"."CUSTOMERS" ('customername')[<fulltext_parameter_list>]
For triggering the text analysis using the option EXTRACTION_CORE, specify the following additional
parameters for the full text index:
TEXT ANALYSIS ON
CONFIGURATION 'EXTRACTION_CORE'
LANGUAGE COLUMN LANG
ALTER TABLE "MY_SCHEMA"."$TA_CUSTOMER_INDEX" ADD CONSTRAINT ALTER_COMMAND FOREIGN
KEY("KEY_1", "KEY_2") REFERENCES "MY_SCHEMA"."CUSTOMERS"("KEY_1","KEY_2") ON DELETE
CASCADE
Related Links
Text Data Processing Language Reference Guide
This guide describes in detail the standard extraction content and the linguistic analysis
300
PUBLI C

Enabling Search
12.1.3.1 Structure of the $TA Table
The table $TA_<index_name> is generated automatically after you have triggered the creation of the index. The
table is built from the key fields of the source table, additional key fields TA_RULE and TA_COUNTER, and eight
additional fields.
Key fields of the source table
Name of the analyzed column
Entity type, for example DATE or NOUN_GROUP
Entity value, for example October 29, 2010 or horizontal stabilizer
Normalized entity value, if available, for example 2010-10-29
Table 13: Structure of TA table
Column ID Key Description Data Type
<n key columns from
source table>
Yes To support a foreign key definition linking
from the $TA table to its source table, the
$TA table has to use exactly the same key
columns as its source table (in data type and
ID).
The $TA table includes all keys from the
source table.
Same as in source table
TA_RULE Yes Stores the rule package that yielded the
token. This is also required to distinguish
between LXP output and output from the
entity extraction.
NVARCHAR(200)
TA_COUNTER Yes The token counter counts all tokens across
the document. The order is only unique for a
given processing type (hence the previous
attribute as key).
BIGINT
TA_TOKEN - Term or entity - depending on processing
type.
NVARCHAR(250)
TA_LANGUAGE - Usually, the document's language is stated in
the source table. In rare cases where this is
not true, the language code is stored here.
Since there is no support for multi-language
documents, the language code is identical for
all result records of a document.
NVARCHAR(2)
TA_TYPE - The token type contains the linguistic or
semantic type of the token; for instance
"noun" (if option = LINGANALYSIS_*) or
"company" (if option = EXTRACTION_*).
NVARCHAR(100)
TA_NORMALIZED - Stores a normalized representation of the
token. This becomes relevant, for example,
for German with umlauts, or /ss.
NVARCHAR(250)
Enabling Search
PUBLI C

Column ID Key Description Data Type
Normalization with regard to capitalization
would not be as important as to justify this
column.
TA_STEM - Stores the linguistic stemming information,
for example, the singular nominative for
nouns, or the indicative for verbs. If text
analysis yields several stems, only the first
stem is stored, assuming this to be the best
match.
NVARCHAR(300)
TA_PARAGRAPH - Stores the total number of paragraphs in the
document.
INTEGER
TA_SENTENCE - Stores the total number of sentences in a
document.
INTEGER
TA_CREATED_AT - Stores the creation time. Used only for
administrative information; for example, for
reorganizing purposes.
TIMESTAMP
The $TA table can be modified like any other table, but it cannot be partitioned.
You can add additional columns to this table, for example, with a statement like ALTER TABLE
"TA"."$TA_SOURCE_TEXT_TA" ADD(C INTEGER GENERATED ALWAYS AS 1). In this example, the value 1 is
written only to existing rows in the new column C. The new column will not be automatically filled in delta updates.
Note
If the source table has a field name identical to one of the default fields from the $TA table, you receive an error
message after the CREATE FULTTEXT INDEX statement prompting you to rename the field of the source table.
After you rename the corresponding field, you can execute the CREATE FULLTEXT INDEX statement again.
12.1.4 Dropping Full Text Indexes
If you want to delete a full text index that you manually created, for example, because it is referenced only rarely
or preprocessing is too time-consuming, you can drop the full text index. For TEXT or SHORTTEXT full text
indexes, you cannot drop the full text index; instead, you must delete the related column in the table.
You also need to drop full text indexes when adding or removing index parameters. As parameters cannot be
added to or removed from an existing full text index, if you want to change parameters, you must first drop the full
text index and then create a new index with the new parameters.
To drop a full text index, you use the DROP FULLTEXT INDEX statement:
DROP FULLTEXT INDEX <index_name>
Note
302
PUBLI C

Enabling Search
Before you can drop a full text index, you must remove the relationship between the source table and any
existing $TA tables (for text analysis). To do so, use the following statement:
ALTER TABLE SCHEMA <$TA_table> DROP <name_constraint>
The name constraint must be the same as originally used when adding the constraint. For more information,
see Text Analysis.
Related Links
Altering Full Text Index Parameters [page 303]
Text Analysis [page 298]
12.1.5 Altering Full Text Index Parameters
To alter the parameters of a full text index, use the ALTER FULLTEXT INDEX statement.
You can only use this statement to alter the following parameters:
Fuzzy search index
Phrase index ratio
Example syntax:
ALTER FULLTEXT INDEX <index_name>
PHRASE INDEX RATIO <parameter value>
To alter any other parameter, you must replace the existing full text index as follows:
a) Delete the existing full text index by using the DROP FULLTEXT INDEX statement.
b) Create a new index using the new parameter values.
Related Links
indexes.
Queues [page 297]
The queue is a mechanism used to enable a full text index to operate asynchronously. This means that when you
insert new entries into the column, the text is not made available in the column until it is preprocessed.
Enabling Search
PUBLI C

12.1.6 Full Text Index Parameters
In SAP HANA, full text indexes are configured using the following parameters:
Parameter Data
Type
Default
(TEXT)
Default
(SHORTTEXT)
Default (Manually
Created)
Description
SCHEMA_NAME NVARCH
AR(256)
Specifies the schema name
TABLE_NAME NVARCH
AR(256)
Specifies the table name
TABLE_OID BIGINT
Specifies the object ID of the
table
INDEX_NAME NVARCH
AR(256)
Specifies the name of the full
text index
INDEX_OID BIGINT
Specifies the object ID of the
full text index
LANGUAGE_COL
UMN
NVARCH
AR(256)
None None None
Specifies the language used for
analyzing the document. If no
language is specified,
automatic language detection is
performed. The detected
language is stored with the
TEXT_AE attribute.
With this option, you can refer
to a column of the same
database table in which the
language for the document is
stored. The column is read by
the queue.
languageColumn in basetable
should be of type NVARCHAR/
VARCHAR(m) where m>=2
when not NULL:
table(tableoid).hasColumn(lang
uageColumn)
MIME_TYPE_CO
LUMN
NVARCH
AR(256)
None None None
This column holds a format/
mimetype indicator (for plain
text, the encoding can also be
specified), e.g. ('text/
plain:CP-1252','PDF','MS_WOR
D 3.0',..).
304
PUBLI C

Enabling Search
Parameter Data
Type
Default
(TEXT)
Default
(SHORTTEXT)
Default (Manually
Created)
Description
'FORMAT' is handled in the
same way as 'LANGUAGE'.
when not NULL:
table(tableoid).hasColumn(mi
meTypeColumn)
LANGUAGE_DET
ECTION
NVARCH
AR(5000
)
EN EN EN
Specifies the set of languages
to be considered for automatic
language detection. If reliable
language detection cannot be
performed, the first language in
the list will be used as the
default language.
This option is used to limit the
languages for text analysis. The
parameter is currently ignored
by text analysis. when not
NULL: for each language L in
languageDetection: L in (select
language from
SYS.M_TEXT_ANALYSIS_LANG
UAGES)
FAST_PREPROC
ESS
TINYINT ON ON ON
Specifies if fast preprocessing
should be performed. With fast
preprocessing the detected
language is always the default
language.
If this option is activated,
linguisitic analysis is skipped.
Only a simple tokenizer is used.
This does not work for
languages which do not use
spaces as word separators. The
fast analyzer cannot handle
binary documents.
FUZZY_SEARCH
_INDEX
TINYINT OFF OFF OFF
If this option is enabled, a
special index is created for the
fuzzy search. This index
accelerates the fuzzy search
but uses additional memory.
SEARCH_ONLY TINYINT OFF ON for LOB-types
OFF, otherwise ON
If set to ON, you cannot
reconstruct original content or
show the document in its
Enabling Search
PUBLI C

Parameter Data
Type
Default
(TEXT)
Default
(SHORTTEXT)
Default (Manually
Created)
Description
original formatting when using
the highlight function. With this
setting, the text attributes do
not store any DAFs, that
means, it is not possible to
retrieve the HTML-converted or
original data from the text
attribute. The document will,
however, use less memory.
If the text attribute is created
via a manually created full text
index, the source attribute that
contains the original data is not
affected by this setting.
IS_EXPLICIT TINYINT 0 0 1
Specifies whether the full text
index was manually created
(also known as explicit). By
default, manually created full
text indexes use ASYNC
synchronization.
FLUSH_AFTER_
DOCUMENTS
INTEGER N/A N/A
Specifies when asynchronous
full text indexes are flushed.
The specified value cannot be
negative.
FLUSH_EVERY_
MINUTES
INTEGER N/A N/A
Specifies when asynchronous
full text indexes are flushed.
The specified value cannot be
negative.
CONFIGURATIO
N
NVARCH
AR(5000
)
None None None
Specifies the type of text
analysis to be used. This
parameter requires the setting
TEXT ANALYSIS = ON.
You can use one of the following
types of text analysis delivered
by SAP:
LINGANALYSIS_BASIC
LINGANALYSIS_STEMS
LINGANALYSIS_FULL
EXTRACTION_CORE
EXTRACTION_CORE_VOICE
OFCUSTOMER
306
PUBLI C

Enabling Search
Parameter Data
Type
Default
(TEXT)
Default
(SHORTTEXT)
Default (Manually
Created)
Description
For more information about the
types of text analysis, see Text
Analysis [page 298].
INTERNAL_COL
UMN_NAME
NVARCH
AR(512)
Specifies the name of the
hidden column created for the
full text index (if the index is not
a TEXT index). Names of hidden
columns have the prefix
"$_SYS_SHADOW" .
PHRASE_INDEX
_RATIO
FLOAT 0.0 0.0 (0.2 if
length <=
1024)
0.0 (0.2 for non-
LOB type and
length <= 1024)
Stores information about the
occurrence of words and the
proximity of words to one
another.
The float value is between 0.0
and 1.0. 1.0 means that the
internal phrase index may use
100% of the memory size of the
fulltext
TEXT
ANALYSIS
TINYINT OFF OFF OFF
Enables text analysis
capabilities on the indexed
column. This parameter cannot
be enabled for TEXT or
SHORTTEXT columns. If set to
ON, the FAST_PREPROCESS
parameter is automatically set
to OFF.
Text analysis can extract
entities such as persons,
products, or places, from
documents and thus enrich the
set of structured information in
SAP HANA. You can use these
additional attributes, which are
stored in a new table, when
creating models (views) for
analytics and search scenarios
to enable improved analytics
and search in SAP HANA.
For more information, see Text
Analysis [page 298].
Enabling Search
PUBLI C

12.1.6.1 Memory Consumption of Full Text Index Parameters
In SAP HANA, certain full text index parameters can have a significant impact on memory consumption based on
how they are defined.
The following full text index parameters can have a significant impact on memory consumption:
Parameter Memory Impact Details
PHRASE_INDEX_RATIO
If the value is greater than 0.0, then there is additional
memory overhead. The maximum memory
consumption is the memory consumption of the full
text index multiplied by the parameter value.
FUZZY_SEARCH_INDEX
To increase response times for fuzzy search, when
enabled, this parameter creates additional in-memory
structures. For text-type columns, fuzzy search
indexes require approx. 10% of the memory size of the
column.
TEXT ANALYSIS
If set to ON, an additional table is created for storing
structured data extracted from the source text for text
analysis. The amount of extracted data depends on the
data in the source column, the text analysis rules, and
the structure of the results. In certain cases, the
memory consumption of the extracted data could
exceed the memory consumption of the source data.
12.2 Building SQL Search Queries
In column-oriented tables, you can perform searches using the SQL SELECT statement.
Before building SQL search queries, the following prerequisites must be met:
The tables you want to search are column-oriented.
You have created any required views for the tables you want to search.
You have created any required full text indexes for the columns you want to search.
SAP HANA supports the standard SQL search syntax and functions for search queries on nearly all visible data
types. You build SQL queries using the SELECT statement, as in the following example:
SELECT * FROM Contacts WHERE ID=1
However, in SAP HANA, columns of large object types and text have additional requirements. To enable search on
columns of these types, you must ensure the following requirements are met:
Full text indexes have been created for the search-relevant columns. For columns of type TEXT and
SHORTTEXT, this is done automatically.
Search queries on the columns use the CONTAINS predicate.
308
PUBLI C

Enabling Search
For large object types and text, you build SQL queries using the SELECT statement and CONTAINS predicate, as in
the following example:
SELECT * FROM Documents WHERE CONTAINS (*,'Comment')
To build a search query, proceed as follows:
1. Use the SQL SELECT statement and specify the table or view and column you want to search. If required,
include the CONTAINS predicate.
2. If required, specify scalar functions for the search.
3. Specify the search terms and, optionally, the search type (EXACT, LINGUISTIC, or FUZZY).
Note
If you do not specify a search type, by default, the search query is performed as an exact search.
Example
For further examples of the syntax used with the SELECT statement, see SAP HANA SQL Reference Manual.
Related Links
http://help.sap.com/hana/html/sql_select.html
12.2.1 Search Queries with CONTAINS
In SAP HANA, you can search one or multiple columns by creating a query that includes the CONTAINS predicate.
In SAP HANA, a search query with CONTAINS has a look and feel similar to common Internet search engines.
The CONTAINS predicate is optional for search queries on columns of most data types; however, for large object
types and text, this predicate is mandatory. You can build a search query with the CONTAINS predicate as follows:
SELECT * FROM <tablename>
WHERE CONTAINS ((<column1>, <column2>, <column3>), <search_string>)
When you specify the CONTAINS predicate, SAP HANA runs the following internal checks:
SAP HANA checks if the query contains one or more terms. If the query contains multiple terms, the terms
are tokenized and concatenated.
SAP HANA checks whether the query is to be run on one or more columns. If you only specify one column, to
optimize the search, additional processes are skipped and the query is run on the single column. If you specify
a wildcard, and therefore possibly numerous columns, SAP HANA automatically determines which columns
are relevant for the search query.
After the checks are performed, SAP HANA builds and runs an internal query on the relevant columns only.
Note
If a column has a full text index assigned, SAP HANA will automatically search on the index rather than on the
original column.
Enabling Search
PUBLI C

Determination of Search-Relevant Columns
You can specify the search-relevant columns either at the creation of the view or directly for the query. SAP HANA
determines which relevant columns to search based on the following hierarchy:
1. You specify a list of columns within the CONTAINS predicate. Even if a list of columns has been defined for the
view, it is overridden by the columns stated in the query.
2. If you enter an asterisk (*) instead of a column list but you specified a list of relevant columns when creating
the view, this list is used for the query.
3. If you enter an asterisk (*) and no list was provided when the view was created, all visible columns of the view
or table are considered as search-relevant.
For information about creating views, see Creating Views in the SAP HANA Administration Guide.
Search Operators and Syntax
With the CONTAINS predicate, SAP HANA supports the following search operators:
OR
Matches are returned that contain at least one of the terms joined by the OR operator.
- (minus)
With a minus sign, SAP HANA searches in columns for matches that do not contain the term immediately
following the minus sign.
" " (quotation marks)
Terms within the quotation marks are not tokenized and are handled as a string. Therefore, all search
matches must be exact.
Note
If you enter multiple search terms, the AND operator is automatically interpreted. Therefore, you do not need to
specify it.
For more information about the unique syntax requirements of the CONTAINS predicate, see the SAP HANA SQL
Reference.
Scalar Functions
For search queries using the CONTAINS predicate, you can use different scalar functions to either return
additional information about the results of your search queries or enhance how the results are displayed. These
functions include SNIPPETS, HIGHLIGHTED, and SCORE.
310
PUBLI C

Enabling Search
Limitations
The following limitations apply to search queries using the CONTAINS predicate:
You cannot search on more than one table or view at a time. If more than one table is joined in the SELECT
statement, then all columns mentioned in the CONTAINS predicate must come from only one of the tables.
You cannot enter a minus (-) search operator directly after OR.
Brackets are not supported as search operators.
Searches using the CONTAINS predicate do not consider non-physical columns, such as calculated columns,
as search-relevant because these columns are created during the search and, therefore, are not available
when SAP HANA internally checks the CONTAINS search query.
The CONTAINS predicate only works on column-oriented tables.
If you specify multiple CONTAINS predicates in the WHERE clause of the SELECT statement, only one of the
predicates is allowed to consist of more than one column in the list of <contains_columns>.
Related Links
http://help.sap.com/hana/html/_dsql_predicates.html#sql_predicates_contains_predicate
http://help.sap.com/hana/hana_admin_en.pdf
12.2.1.1 SNIPPETS Function
For search queries using the CONTAINS predicate, you can use the function SNIPPETS to return search results
with an excerpt of the text with your search term highlighted in bold. This short text excerpt provides some
context for you to see where and how the term is used in the document.
This function uses the following syntax:
SELECT *, SNIPPETS (<text_column>) FROM <tablename>
WHERE CONTAINS (<search_term>)
Limitations
The SNIPPETS function has the following limitations:
Only the first search term specified with the CONTAINS predicate is highlighted in the returned text.
The query result contains only the first hit of the first search term.
The text excerpt that is displayed with the search term is limited to a string of 12 tokens.
This function only works on columns of the TEXT data type or columns with a full text index.
12.2.1.2 HIGHLIGHTED Function
For search queries using the CONTAINS predicate, you can use the function HIGHLIGHTED to return the content of
the found document with your search term highlighted in bold.
Enabling Search
PUBLI C

Search queries using the HIGHLIGHTED function return the data type NCLOB.
SELECT *, HIGHLIGHTED (<text_column>) FROM <tablename>
Limitations
The HIGHLIGHTED function has the following limitations:
Only the first search term specified with the CONTAINS predicate is highlighted in the returned text.
The query result contains all hits of the first search term.
This function only works on columns of the TEXT data type or columns with a full text index.
12.2.1.3 SCORE Function
For search queries using the CONTAINS predicate, you can use the function SCORE to get the score, that means
the relevance, of a record found.
SAP HANA calculates a score based on the following information:
The relevance or weighting of attributes in a search using the CONTAINS predicate. The relevance of a hit
depends on the weight of the column that caused the hit. You can specify weights when you create the view or
in the CONTAINS predicate.
Fuzziness in fuzzy search. The more exact a hit is, the higher the score is.
Text ranking (TF-IDF).
SELECT SCORE (),* FROM <tablename>
12.2.2 EXACT Search
An exact search returns records only if the search term or search phrase is contained in the table column exactly
as specified. In the SELECT statement of the search query, you can specify the EXACT search type.
In an exact search, the search engine uses the word dictionary and the phrase index to detect the possible
matches. The search engine then checks whether the words appear and use exactly the same spelling.
For text columns, the search term must match at least one of the tokenized terms to return a column entry as
a match.
For string columns, the search term must match the entire string to return a column entry as a match.
Note
312
PUBLI C

Enabling Search
For more flexibility in a search query, you can use the supported wildcards % and *. Wildcards are supported for
both text and string columns.
You can perform an exact search by using the CONTAINS predicate with the EXACT option in the WHERE clause of a
SELECT statement. The exact search is the default search type. If you do not specify any search type in the search
query, an exact search will be executed automatically.
Example
WHERE CONTAINS (<column_name>, <search_string>, EXACT)
WHERE CONTAINS (<column_name>, <search_string>)
--- Exact search will be executed implicitly.
WHERE CONTAINS (<column_name>, '"cats and dogs"')
--- Phrase search.
12.2.3 LINGUISTIC Search
A linguistic search finds all words that have the same word stem as the search term. It also finds all words for
which the search term is the word stem. In the SELECT statement of the full text search query, you can specify the
LINGUISTIC search type.
When you execute a linguistic search, the system has to determine the stems of the searched terms. It will look up
the stems in the stem dictionary. The hits in the stem dictionary point to all words in the word dictionary that have
this stem
You can call the linguistic search by using the CONTAINS predicate with the LINGUISTIC option in the WHERE
clause of a SELECT statement.
A linguistic search for produced will also find producing and produce.
Example
WHERE CONTAINS (<column_name, 'produced' , LINGUISTIC)
Limitations
You can only perform linguistic searches on columns that meet the following conditions:
The columns contain text.
For the columns, the FAST_PREPROCESS parameter is specified as OFF.
Enabling Search
PUBLI C

12.2.4 FUZZY Search
Fuzzy Search is a fast and fault-tolerant search feature for SAP HANA. A fuzzy search returns records even if the
search term contains additional or missing characters or other types of spelling errors.
The term fault-tolerant search means that a database query returns records even if the search term (the user
input) contains additional or missing characters or other types of spelling error.
Fuzzy search can be used in various applications, for example:
Fault-tolerant search in text columns (for example, html or pdf): Search for documents on 'Driethanolamyn'
and find all documents that contain the term 'Triethanolamine'.
Fault-tolerant search in structured database content: Search for a product called 'coffe krisp biscuit' and find
'Toffee Crisp Biscuits'.
Fault-tolerant check for duplicate records: Before creating a new customer record in a CRM system, search for
similar customer records and verify that there are no duplicates already stored in the system. When, for
example, creating a new record 'SAB Aktiengesellschaft & Co KG Deutschl.' in 'Wahldorf', the system shall
bring up 'SAP Deutschland AG & Co. KG' in 'Walldorf' as a possible duplicate.
You can call the fuzzy search by using the CONTAINS predicate with the FUZZY option in the WHERE clause of a
SELECT statement.
Example
WHERE CONTAINS (<column_name>, <search_string>, FUZZY (0.8))
12.2.4.1 Supported Data Types
Fuzzy search is only available for column tables and attribute views and supports the following SQL types and
column store types.
SQL Type Column Store Type Supported Fuzzy Search Features
Character String Types
VARCHAR CS_STRING string features
NVARCHAR CS_STRING string features
SHORTTEXT CS_STRING text features
Large Object (LOB) Types
TEXT CS_TEXT text features
Datetime Types
DATE CS_DAYDATE date features
Other
FULLTEXT INDEX - text features
314
PUBLI C

Enabling Search
All other SQL types are not supported by fuzzy search.
String Types
String types support a basic fuzzy string search. The values of a column are compared with the user input, using
the fault-tolerant fuzzy string comparison.
When working with string types, the fuzzy string comparison always compares the full strings. If searching with
'SAP', for example, a record such as 'SAP Deutschland AG & Co. KG' gets a very low score, because only a very
small part of the string is equal (3 of 27 characters match).
For more information, see Fuzzy Search on String Columns [page 335].
Text Types
Text types support a more sophisticated kind of fuzzy search. Texts are tokenized (split into terms) and the fuzzy
comparison is done term by term.
For example, when searching with 'SAP', a record such as 'SAP Deutschland AG & Co. KG' gets a high score,
because the term 'SAP' exists in both texts. A record such as 'SAPPHIRE NOW Orlando' gets a lower score,
because 'SAP' is only a part of the longer term 'SAPPHIRE' (3 of 8 characters match).
For more information, see Fuzzy Search on Text Columns [page 346].
Date Types
Fuzzy search on date values checks for date-specific errors like dates that lie within a given range of days or dates
that have month and day exchanged (for example, American versus British date format).
For more information, see Fuzzy Search on DATE Columns [page 366].
12.2.4.2 Syntax
You can call the fuzzy search by using the CONTAINS() function with the FUZZY() option in the WHERE clause of a
SELECT statement.
For more information, see Using the CONTAINS() Predicate [page 316].
Basic example without additional search options
SELECT SCORE() AS score, *
FROM documents
WHERE CONTAINS(doc_content, 'Driethanolamyn', FUZZY(0.8))
ORDER BY score DESC;
Enabling Search
PUBLI C

Example with additional search options
Additional search options that change the default behavior of the fuzzy search can be specified as additional
string parameters in the FUZZY() function.
FROM documents
WHERE CONTAINS(doc_content, 'Driethanolamyn', FUZZY(0.8, 'option1=value1,
option2=value2'))
The search options are specified as a comma-separated list of key-value pairs.
For more information, see Available Fuzzy Search Options [page 327].
12.2.4.3 CONTAINS Predicate
12.2.4.3.1 Using the CONTAINS() Predicate
Supported Database Objects
Fuzzy search and the CONTAINS() predicate are available for SELECT statements on one column table or on one
attribute view.
Note
The following operations and database objects are not supported:
Joins between column tables or attribute views
Row tables
Database views created with the CREATE VIEW statement
Calculation views
Analytic views
Use Cases of CONTAINS()
The CONTAINS() predicate can be used in the WHERE clause of a SELECT statement. The type of search it
performs depends on its arguments:
1. A freestyle search on multiple columns
2. A full text search on one column containing large documents
3. A search on one database column containing structured data
All searches can be done either as an exact search or as a fuzzy search with additional tolerance for writing errors.
316
PUBLI C

Enabling Search
Freestyle Search on Multiple Columns
-- exact search
SELECT ... WHERE CONTAINS((col1, col2, col3), 'term1 term2 term3') ...;
-- or
SELECT ... WHERE CONTAINS((col1, col2, col3), 'term1 term2 term3', EXACT) ...;
-- fuzzy search
SELECT ... WHERE CONTAINS((col1, col2, col3), 'term1 term2 term3', FUZZY(0.7)) ...;
Full Text Search on One Column Containing Large Documents
To do a full text search, the column that is to be searched must be a text column.
-- exact search
SELECT ... WHERE CONTAINS(col1, 'term1 term2 term3') ...;
-- fuzzy search
SELECT ... WHERE CONTAINS(col1, 'term1 term2 term3', FUZZY(0.7)) ...;
Search on One Database Column Containing Structured Data
-- exact search
SELECT ... WHERE CONTAINS(col1, 'term1 term2 term3') ...;
-- fuzzy search
SELECT ... WHERE CONTAINS(col1, 'term1 term2 term3', FUZZY(0.7)) ...;
Multiple CONTAINS() Predicates in one SELECT
It is possible to use the CONTAINS() predicate more than once in a WHERE clause. In this case, only one
CONTAINS() can be used for a freestyle search on multiple columns. All other calls to CONTAINS() can access a
single column only.
SELECT ...
WHERE CONTAINS((col1, col2, col3), 'a b c', FUZZY(0.8))
AND CONTAINS(col4, 'x y z', FUZZY(0.7))
AND CONTAINS(col5, 'u v w', FUZZY(0.7))
AND ...
Reserved Words and Special Characters in the Search String
When searching with CONTAINS(), some terms and characters have a special meaning, as described below. For
more information, see the text search documentation.
Reserved Word/Special Character Description
OR A search such as CONTAINS(col, 'sap OR hana')
searches for all records that contain 'sap' or 'hana' in
column 'col'. The OR keyword is case sensitive, so
Enabling Search
PUBLI C

Reserved Word/Special Character Description
CONTAINS(col, 'sap or hana') searches for records
that contain 'sap' and 'or' and 'hana'.
Note
If your search input contains an 'or', make sure that
it is not in uppercase characters to avoid the OR
semantics.
Minus Sign (-)
A search such as CONTAINS(col, 'sap -hana') searches
for all records that contain 'sap' but not 'hana'. If the
second term is given as a phrase like in CONTAINS(col,
'sap "-hana"'), the database searches for records that
contain 'sap' and '-hana'. In text columns, the '-' is
removed from the second search term as it is a
delimiter symbol.
Note
If your search input contains terms starting with a
minus sign, make sure that these terms are
enclosed in double quotes to avoid the NOT
semantics.
Double Quotes (") A search that is enclosed in double quotes is searched
as a phrase. For example, CONTAINS(col, '"sap hana"')
searches for all records that contain 'sap hana' as a
phrase without any additional terms in between.
Asterisk (*) The asterisk activates a wildcard search. In this case,
no fuzzy search is done. An exact match with wildcards
is called instead.
Percent Sign (%) The percent sign is replaced with an asterisk (*) and a
wildcard search is called.
12.2.4.3.2 The SCORE() Function
When using CONTAINS() in the WHERE clause of a SELECT statement, the SCORE() function can be used to
retrieve the score that is a numeric value between 0.0 and 1.0.
The score defines the similarity between the user input and the records returned by the search. A score of 0.0
means there is no similarity. The higher the score, the more similar a record is to the search input.
When more than one CONTAINS() is given in the WHERE clause, the score is calculated as an average of the
scores of all columns.
SELECT SCORE(), col1, col2, ... FROM tab
WHERE CONTAINS(col1, 'x y z')
318
PUBLI C

Enabling Search
AND CONTAINS(col2, 'a b c')
AND ...
ORDER BY SCORE() DESC;
It is possible to assign a weight to each column. The weight is used to calculate the score as a weighted average of
the scores of all columns.
If a weight is not given, the default weight is 1.0.
LECT SCORE(), ... FROM tab
WHERE CONTAINS((col1, col2, col3), 'a b c', FUZZY(0.8), WEIGHT(1.0, 0.5, 0.5))
AND CONTAINS(col4, 'x y z', FUZZY(0.7), WEIGHT(0.7))
AND CONTAINS(col5, 'u v w', FUZZY(0.7))
AND ...
Note
Currently, the SCORE() function uses all conditions in the WHERE clause to calculate the score. It is not
restricted to the CONTAINS() predicates. However, this may change in future revisions, so an application must
not rely on the score calculation when combining CONTAINS() and other predicates in the WHERE clause.
For example, the condition col5 = 'u v w' is also used as part of the score calculation.
SELECT SCORE(), ... FROM tab
WHERE CONTAINS((col1, col2, col3), 'a b c', FUZZY(0.8))
AND CONTAINS(col4, 'x y z', FUZZY(0.7))
AND col5 = 'u v w'
AND ...
12.2.4.3.3 Examples
Fuzzy Search on One Column
1. Create the data.
CREATE COLUMN TABLE companies
(
id INTEGER PRIMARY KEY,
companyname SHORTTEXT(200) FUZZY SEARCH INDEX ON
);
INSERT INTO companies VALUES (1, 'SAP Corp');
INSERT INTO companies VALUES (2, 'SAP in Walldorf Corp');
INSERT INTO companies VALUES (3, 'ASAP');
INSERT INTO companies VALUES (4, 'ASAP Corp');
INSERT INTO companies VALUES (5, 'BSAP orp');
INSERT INTO companies VALUES (6, 'IBM Corp');
2. Perform the search on one column.
SELECT SCORE() AS score, * FROM companies
WHERE CONTAINS(companyname,'xSAP Corp Walldorf',
FUZZY(0.7,'textSearch=compare,bestMatchingTokenWeight=0.7'))
Enabling Search
PUBLI C

SCORE ID COMPANYNAME
0.94 2 SAP in Walldorf Corp
Fuzzy Search on Two Columns
1. Create the data.
CREATE COLUMN TABLE companies2
(
companyname SHORTTEXT(200) FUZZY SEARCH INDEX ON,
contact SHORTTEXT(100) FUZZY SEARCH INDEX ON
);
INSERT INTO companies2 VALUES (1, 'SAP Corp', 'Mister Master');
INSERT INTO companies2 VALUES (2, 'SAP in Walldorf Corp', 'Master Mister');
INSERT INTO companies2 VALUES (3, 'ASAP', 'Nister Naster');
INSERT INTO companies2 VALUES (4, 'ASAP Corp', 'Mixter Maxter');
INSERT INTO companies2 VALUES (5, 'BSAP orp', 'Imster Marter');
INSERT INTO companies2 VALUES (6, 'IBM Corp', 'M. Master');
2. Perform the search on two columns.
SELECT SCORE() AS score, * FROM companies2
WHERE CONTAINS(companyname, 'IBM',
AND CONTAINS(contact, 'Master',
SCORE ID COMPANYNAME CONTACT
0.91 6 IBM Corp M. Master
3. Perform a freestyle search.
WHERE CONTAINS((companyname,contact), 'IBM Master', FUZZY(0.7))
Note
Freestyle searches always use TF/IDF to calculate the score and do not support parameters like
'textSearch=compare' or 'bestMatchingTokenWeight=0.7' which influence score calculation. This
therefore results in a different score for the same record.
320
PUBLI C

Enabling Search
Perform a Freestyle Search on Two Columns
1. Create the data.
CREATE COLUMN TABLE companies2
(
companyname SHORTTEXT(200) FUZZY SEARCH INDEX ON,
contact SHORTTEXT(100) FUZZY SEARCH INDEX ON
);
INSERT INTO companies2 VALUES (1, 'SAP Corp', 'Mister Master');
INSERT INTO companies2 VALUES (2, 'SAP in Walldorf Corp', 'Master Mister');
INSERT INTO companies2 VALUES (3, 'ASAP', 'Nister Naster');
INSERT INTO companies2 VALUES (4, 'ASAP Corp', 'Mixter Maxter');
INSERT INTO companies2 VALUES (5, 'BSAP orp', 'Imster Marter');
INSERT INTO companies2 VALUES (6, 'IBM Corp', 'M. Master');
2. Perform a freestyle search on two columns.
WHERE CONTAINS((companyname,contact), 'IBM Master', FUZZY(0.7))
Note
Freestyle search always uses TF/IDF to calculate the score and does not support parameters that
influence score calculationl, such as 'textSearch=compare' or 'bestMatchingTokenWeight=0.7' . This
therefore results in a different score for the same record.
12.2.4.4 Fuzzy Score
The fuzzy search algorithm calculates a fuzzy score for each string comparison. The higher the score, the more
similar the strings are. A score of 1.0 means the strings are identical. A score of 0.0 means the strings have
nothing in common.
You can request the score in the SELECT statement by using the SCORE() function. You can sort the results of a
query by score in descending order to get the best records first (the best record is the record that is most similar
to the user input). If a fuzzy search of multiple columns is used in a SELECT statement, the score is returned as an
average of the scores of all columns used.
When searching text columns, a TF/IDF (term frequency/inverse document frequency) score is returned by
default instead of the fuzzy score. The fuzzy score influences the TF/IDF calculation, but it is important to keep in
mind that, with TF/IDF, the range of the score values returned is normed to the interval between 0.0 and 1.0, and
the best record always gets a score of 1.0, regardless of its fuzzy score.
The TF/IDF calculation can be disabled so that you get the fuzzy score instead. In particular, this makes sense for
short-text columns containing data such as product names or company names. On the other hand, you should
Enabling Search
PUBLI C

use TF/IDF for long-text columns containing data such as product descriptions, HTML data, or Word and PDF
documents.
12.2.4.4.1 Option similarCalculationMode
The option similarCalculationMode controls how the similarity of two strings (or, for TEXT attributes, terms) is
calculated.
Score Calculation Modes
Basically, the similarity of two strings is defined by the number of common characters, wrong characters,
additional characters in the search string, and additional characters in the reference string.
The following calculation modes exist:
Table 14: Modes
Mode Impact on wrong
characters
Impact on additional
characters in search
Impact on additional
characters in table
search high high low
compare (default) moderate high high
symmetricsearch high moderate moderate
Note that a high impact results in a lower score.
Table 15: Examples with score
Request Reference Compa
re
Search Symmetricsearch
search searching 0.76 0.96 0.86
search seerch 0.85 0.75 0.75
search searchingforextrater
restriallife
0.0 0.91 0.87
searchingforextrater
restriallife
searching 0.0 0.35 0.84
restriallife
seerch 0.0 0.24 0.79
restriallife
searchingforthemea
ningoflife
0.6 0.57 0.6
322
PUBLI C

Enabling Search
SQL Examples
Preparations
OP TABLE test_similar_calculation_mode;
CREATE COLUMN TABLE test_similar_calculation_mode
(
s NVARCHAR(255)
);
INSERT INTO test_similar_calculation_mode VALUES ('1','stringg');
INSERT INTO test_similar_calculation_mode VALUES ('2','string theory');
INSERT INTO test_similar_calculation_mode VALUES ('3','this is a very very very
long string');
INSERT INTO test_similar_calculation_mode VALUES ('4','this is another very long
string');
similarCalculationMode compare
SELECT TO_INT(SCORE()*100)/100 AS score, id, s
FROM test_similar_calculation_mode
WHERE CONTAINS(s, 'explain string theory', FUZZY(0.5,
'similarCalculationMode=compare'))
Score ID S
0.69 2 string theory
similarCalculationMode search
'similarCalculationMode=search'))
Score ID S
0.68 4 this is another very long string
0.66 3 this is a very very very long string
similarCalculationMode symmetricsearch
'similarCalculationMode=symmetricsearch'))
Score ID S
0.71 1 stringg
Enabling Search
PUBLI C

Score ID S
0.61 4 this is another very long string
0.58 3 this is a very very very long string
12.2.4.4.2 Option spellCheckFactor
The option spellCheckFactor defines the score for strings that are not identical but get a fuzzy score of 1.0.
There are two use cases for the option spellCheckFactor:
A) This option allows you to set the score for terms that are not fully equal but that would be a 100% match
because of the internal character standardization used by the fuzzy search.
For example, the terms 'Caf' and 'cafe' give a score of 1.0 although the terms are not equal. For some users it
may be necessary to distinguish between both terms.
The decision whether two terms are equal is based on the term representation stored in the column
dictionary. Therefore, the spellCheckFactor option works differently on string and text columns, as described
in the following sections.
B) The fuzzy search can return a 100% match for terms that are not identical but cannot be differentiated by
the fuzzy-string-compare algorithm.
For example, the fuzzy search cannot differentiate between the terms 'abaca' and 'acaba'. In this case, the
spellCheckFactor can be used to avoid a score of 1.0.
If A) and B) are not needed by an application, you can set the spellCheckFactor to 1.0 to disable the feature.
Standardization of Letters and Terms
All characters are replaced by lowercase characters without any diacritics before the fuzzy comparison takes
place. This is called standardization. It is therefore possible to get a 100% match when comparing two unequal
terms, because the standardization process returned two identical terms.
Standardization Examples
Original Letter Standardized Letter
E e
e e
e
e
e
e
324
PUBLI C

Enabling Search
The letter i is treated differently, since it is not standardized to an as would be the 'standard' rule.
I i
i
i i
i
German umlauts are replaced by two characters.
ae
ae
oe
oe
ue
ue
ss
Due to this standardization we get high fuzzy scores for common differences in the spelling of words.
Original term Standardized term
mller mueller
Mueller mueller
Cafe cafe
Caf cafe
Search on a String Column (VARCHAR, NVARCHAR)
The decision as to whether two strings are the same is based on the string representation stored in the dictionary
for the column in question. The contents of a string column are converted to lowercase characters before being
stored in the dictionary. No otther standardizations are carried out.
It is therefore not possible to use the spellCheckFactor option for example to distinguish between 'caf' and 'cafe'.
Example
CREATE COLUMN TABLE test_spell_check_factor
(
Enabling Search
PUBLI C

s NVARCHAR(255)
);
INSERT INTO test_spell_check_factor VALUES ('1','Muller');
INSERT INTO test_spell_check_factor VALUES ('2','Mueller');
INSERT INTO test_spell_check_factor VALUES ('3','Mller');
SELECT SCORE() AS score, id, s
FROM test_spell_check_factor
WHERE CONTAINS(s, 'Mller', FUZZY(0.5, 'spellCheckFactor=0.9'))
DROP TABLE test_spell_check_factor;
SCORE ID T Description
1.0 3 Mller
0.9 2 Mueller << spellCheckFactor got
used
0.88 1 Muller
0.88 4 Mller
Search on a Text Column (SHORTTEXT, TEXT or FULLTEXT INDEX)
Terms in text columns are standardized to lowercase characters without diacritics before being stored in the
dictionary. In text columns it is therefore not possible to distinguish for example between 'caf' and 'cafe' or
'Mller' and 'mueller'. In this case, the search always returns a score of 1.0.
The main use case of spellCheckFactor on text columns is therefore to avoid a score of 1.0 for terms like 'abaca'
and 'acaba'.
Example
CREATE COLUMN TABLE test_spell_check_factor
(
t SHORTTEXT(200) FUZZY SEARCH INDEX ON
);
INSERT INTO test_spell_check_factor VALUES ('1','Muller');
INSERT INTO test_spell_check_factor VALUES ('2','Mueller');
SELECT SCORE() AS score, id, t
FROM test_spell_check_factor
WHERE CONTAINS(t, 'Mller', FUZZY(0.5, 'spellCheckFactor=0.9,textSearch=compare'))
DROP TABLE test_spell_check_factor;
326
PUBLI C

Enabling Search
SCORE ID T
1.0 2 Mueller
1.0 3 Mller
0.88 1 Muller
0.88 4 Mller
12.2.4.5 Available Fuzzy Search Options
Note that some data types in the table below are data-type combinations.
String: SQL types VARCHAR and NVARCHAR
Text: SQL types TEXT and SHORTTEXT and any columns that have an additional FULLTEXT INDEX
Date: SQL type DATE
Table 16: Available Fuzzy Search Options
Name of Option Short
Name
Range Defa
ult
Applies to
Types
Description
emptyScore es 0.0..1.0 not
set
Text, String,
Date
Defines how an empty and a non-empty value
shall match.
For more information, see Option emptyScore
[page 332]
similarCalculatio
nMode
scm search,
compare,
symmetri
csearch
comp
are
Text, String
Defines how the score is calculated for a
comparison of strings (or terms in a Text
column).
For more information, see Option
similarCalculationMode [page 322] .
spellCheckFacto
r
scf 0.0..1.0 0.9 Text, String
Sets the score for strings that get a fuzzy score
of 1.0 but are not fully equal.
spellCheckFactor [page 324].
abbreviationSimi
larity
abs 0.0..1.0 0.0 Text
Activates abbreviation similarity and sets the
score.
abbreviationSimilarity [page 353].
andSymmetric as on,off,tru
e,false
off Text
Activates a symmetric AND content search.
For more information, see Partially Matching with
Parameter andThreshold [page 351].
Enabling Search
PUBLI C

Name
Range Defa
ult
Applies to
Types
Description
andThreshold at 0.0..1.0 1.0 Text
Activates a 'soft AND' and determines the
percentage of the tokens that need to match.
bestMatchingTo
kenWeight
bmtw 0.0..1.0 0 Text
Influences the score, shifts total score value
between best token score values and root mean
square of score values.
For more information, see Multi-Token Search on
Text Columns [page 348].
considerNonMat
chingTokens
cnmt max, min,
all, input,
table
max Text
Influences the score, defines the number of
terms used for score calculation.
For more information, see Multi-Token Search on
Text Columns [page 348].
minTextScore mts 0.0..1.0 0.0 Text
Minimum score of a TEXT field; if not reached,
the record will not be part of the result.
For more information, see Option minTextScore
[page 355].
stopwordListId sli Text
Activates the stopwords.
For more information, see Usage [page 357].
stopwordTable st not
set
Text
Activates the stopwords.
For more information, see Usage [page 357]
termMappingLis
tId
tmli Text
Activates the term mappings.
termMappingTa
ble
tmt not
set
Text
Activates the term mappings.
For more information, see Term Mappings [page
361].
textSearch ts fulltext,
compare
fullte
xt
Text, String,
Date
Switches between full text search with TF/IDF
score and duplicate search with fuzzy score.
For more information, see Option textSearch
[page 356].
maxDateDistanc
e
mdd 0..100 0 Date
Specifies the allowed date distance when using
fuzzy search on dates.
328
PUBLI C

Enabling Search
Name
Range Defa
ult
Applies to
Types
Description
For more information, see Fuzzy Search on DATE
Columns [page 366].
12.2.4.5.1 Allowed Combinations of Fuzzy Search Parameters
Possible combinations of search options are shown in the table below.
Depending on the following conditions, some of the searchOptions parameters are not allowed because they do
not make sense:
The data type of a column
The type of the search (freestyle or attribute search)
The selected method for score calculation on text columns (fuzzy score or TF/IDF, search option textSearch)
If the user sets an option that is not allowed, an SQL error is thrown and the SELECT aborts.
The following search types are used in the table below:
Freestyle search: the CONTAINS() predicate uses more than one column.
Attribute search: the CONTAINS() predicate uses a single column only.
Search on TEXT...: attribute search on a column of type TEXT, SHORTTEXT, or other type with an additional
FULLTEXT INDEX.
Search on types other than TEXT: attribute search that is not done on a TEXT column as defined above.
Search on TEXT
with Fuzzy Score
Search on TEXT
with TD/IDF Score
Freestyle Search Search on Types
Other than TEXT
attribute search and
textSearch=
compare and
datatype= TEXT
textSearch=fulltext
and datatype=TEXT
freestyle search (full
text search with
TD/IDF score)
datatype <>TEXT
textSearch=compar
e
YES n/a n/a NO
textSearch=fulltext n/a YES n/a (default mode) NO
similarCalculationM
ode
YES YES YES YES
spellCheckFactor YES YES YES YES
fuzzySimilarity > 0 YES YES YES YES
fuzzySimilarity = 0 YES NO NO YES
emptyScore YES NO NO YES (Only valid for
text, string, and date
types. Numeric
types are not
supported.)
Enabling Search
PUBLI C

Search on TEXT
with Fuzzy Score
Search on TEXT
with TD/IDF Score
Freestyle Search Search on Types
Other than TEXT
abbreviationSimilarit
y
YES NO NO NO
maxDateDistance NO NO YES YES (Only valid for
date types. Other
types are not
supported.)
termMappingTable/
ListId
YES YES YES NO
stopwordTable/
ListId
YES NO NO NO
andThreshold YES NO NO NO
andSymmetric YES NO NO NO
bestMatchingToken
Weight
YES NO NO NO
considerNonMatchin
gTokens
YES NO NO NO
minTextScore YES NO NO NO
Rank Calculation fuzzy score TF/IDF TF/IDF (text
columns) or fuzzy
score (other SQL
types)
fuzzy score
Legend
YES - The parameter is allowed.
NO - The parameter is not allowed and an error message is returned if the user sets this option.
12.2.4.5.2 Search Options and CONTAINS()
Here you can find several examples for using search options and the CONTAINS() statement.
Example without specifying additional search options
FROM documents
WHERE CONTAINS(doc_content, 'Driethanolamyn', FUZZY(0.8))
330
PUBLI C

Enabling Search
0.8 is the default value for the fuzziness parameter, so the above example can also be written as
FROM documents
WHERE CONTAINS(doc_content, 'Driethanolamyn', FUZZY)
Example with additional search options
Additional search options that change the default behaviour of the fuzzy search can be specified as an additional
string parameter to the FUZZY() function.
FROM documents
WHERE CONTAINS(doc_content, 'Driethanolamyn', FUZZY(0.8, 'option1=value1,
option2=value2'))
The search options are specified as a comma-separated list of key-value pairs.
Example with EXACT search and additional search options
It is also possible to use search options in combination with an exact search. So, for example, search options for
term mappings and stopwords can be used for fuzzy and exact search.
FROM documents
WHERE CONTAINS(doc_content, 'Driethanolamyn', EXACT('option1=value1,
option2=value2'))
12.2.4.6 Fuzzy Search - Options for All Column Types
dummy before conversion 4
There are search options that are not only valid for a specific column type such as text or string. These options are
described in this chapter.
Enabling Search
PUBLI C

12.2.4.6.1 Option emptyScore
Introduction
Many database tables often contain incomplete data. For example, the first name or the phone number of a
customer may be empty, either because the information was not known when the database record was created or
because of missing data maintenance.
Therefore, the search input may contain more information than the database record that the user is looking for. In
this case, the user still expects to get the result.
To get the expected behaviour using standard SQL, an application developer writes code as in the following
example:
SELECT score(), ... FROM ...
WHERE ...
AND (CONTAINS(firstname, 'Peter', FUZZY(0.8)) OR firstname IS NULL)
AND ...
It is not possible to specify the score for the 'firstname IS NULL' clause, so the overall score() for records with an
empty firstname may get an unexpected score() that probably does not fit into the sort order of other results. In
addition, SQL statements get longer and more complex with the additional WHERE clauses.
This is why the option 'emptyScore' has been introduced.
For symmetry reasons, the search works the other way around, too. So, when searching with an empty column,
records that contain a value in the column are also returned. This is important for batch processes, for example,
where the order of records processed is not known and results shall be the same, regardless of the order of
processing.
In the following sections an empty column value is a column value that is either an empty string ('') or a NULL
value. We do not distinguish between these two values.
Supported Data Types
The search option 'emptyScore' supports the following SQL data types:
VARCHAR
NVARCHAR
SHORTTEXT
TEXT
DATE
Columns with a FULLTEXT INDEX
Note
Numeric types like INTEGER, DECIMAL, FLOAT, and so on are currently not supported.
332
PUBLI C

Enabling Search
Example
DROP TABLE test_emptyscore;
CREATE COLUMN TABLE test_emptyscore
(
t TEXT FUZZY SEARCH INDEX ON
);
INSERT INTO test_emptyscore VALUES ('1', 'eins');
INSERT INTO test_emptyscore VALUES ('2', ''); -- empty string
INSERT INTO test_emptyscore VALUES ('3', ' '); -- n blanks
INSERT INTO test_emptyscore VALUES ('4', NULL); -- NULL value
Select 'eins' without emptyScore
SELECT SCORE() AS score, * FROM test_emptyscore WHERE CONTAINS(T, 'eins', FUZZY(0.5,
'textSearch=compare')) ORDER BY score DESC, id;
SCORE ID T
1.0 1 eins
Select 'eins' with emptyScore
SELECT SCORE() AS score, * FROM test_emptyscore WHERE CONTAINS(T, 'eins', FUZZY(0.5,
'textSearch=compare, emptyScore=0.5')) ORDER BY score DESC, id;
SCORE ID T
1.0 1 eins
0.5 2
0.5 4 <NULL>
Select empty string without emptyScore
SELECT SCORE() AS score, * FROM test_emptyscore WHERE CONTAINS(T, '', FUZZY(0.5,
'textSearch=compare')) ORDER BY score DESC, id;
SCORE ID T
1.0 2
Select empty string with emptyScore
'textSearch=compare, emptyScore=0.5')) ORDER BY score DESC, id;
SCORE ID T
1.0 1
1.0 4 <NULL>
0.5 1 eins
0.5 3
Enabling Search
PUBLI C

Select empty string with emptyScore and minTextScore
'textSearch=compare, emptyScore=0.5, mintextscore=0.8')) ORDER BY score DESC, id;
SCORE ID T
1.0 2
1.0 4 <NULL>
0.5 1 eins
0.5 3
Note
When searching with an empty value, for example, records that contain a value in the column are returned and
will get the score of the parameter emptyScore. The list will not be shortened, even though the parameter
minTextScore is set.
Getting Records with Empty Column Values
When the search is called with the emptyScore option as in the following example
WHERE ...
AND CONTAINS(firstname, 'Peter', FUZZY(0.8, 'textSearch=compare,
emptyScore=0.9, ...<otherOptions>...))
AND ...
the records returned by the search are the same as with the following SQL statement (scores differ of course
because of the emptyScore parameter)
WHERE ...
AND ( CONTAINS(firstname, 'Peter', FUZZY(0.8,
'textSearch=compare, ...<otherOptions>...))
OR firstname IS NULL
OR firstname = ''
)
AND ...
For records containing a NULL or an empty string in the firstname column, the value of emptyScore (0.9 in this
example) is used as firstname score when calculating the overall score. For records containing a non-empty
firstname, the fuzzy score is calculated and used.
For columns that do not allow empty strings as values (date types for example) the result of the search is the
same as the query
WHERE ...
AND ( CONTAINS(dateofbirth, '2000-01-02', FUZZY(0.8, '...<otherOptions>...))
OR dateofbirth IS NULL
334
PUBLI C

Enabling Search
)
AND ...
Note
Numeric types like INTEGER, DECIMAL and FLOAT are currently not supported.
Finding Non-Empty Column Values when Searching with an Empty Column Value
When the search is called with an empty string and with option emptyScore as in the following example:
WHERE ...
AND CONTAINS(firstname, '', FUZZY(0.8, 'textSearch=compare,
emptyScore=0.9, ...<otherOptions>...))
AND ...
the result of the search is the same as with the following SQL statement (scores differ, of course, because of the
emptyScore parameter):
WHERE ...
AND ( firstname IS NULL
OR firstname = ''
OR firstname LIKE '_%'
)
AND ...
For records that contain a NULL or an empty string in the firstname column, a score of 1.0 is used, because an
empty string is considered a 'perfect' match for a search with an empty string. Records with a non-empty
firstname get a firstname score of 0.9, which is the value of the emptyScore parameter in this example.
For columns that do not allow empty strings as a value (for example, date types), the result of the search is the
same as the query:
WHERE ...
AND ( dateofbirth IS NULL
OR dateofbirth IS NOT NULL
)
AND ...
12.2.4.7 Fuzzy Search on String Columns
String types support a basic fuzzy string search. The values of a column are compared with the user input using
the fault-tolerant fuzzy string comparison.
When working with string types, the fuzzy string compare always compares the full strings. When searching with
'SAP' for example, a record like 'SAP Deutschland AG & Co. KG' gets a very low score, because only a very small
part of the string is equal (3 of 27 characters match).
Enabling Search
PUBLI C

A fuzzy search on string types is a replacement for non-fault tolerant SQL statements like
SELECT ... FROM products WHERE product_name = 'coffe krisp biscuit' ...
which would not return any results because of the spelling errors.
The following SQL data types are supported:
VARCHAR
NVARCHAR
It is possible to speed up the fuzzy search by creating additional data structures called 'fuzzy search indexes'.
These are used for faster calculation of the fuzzy score. These indexes exist in the memory only, so no additional
disk space is needed.
To get the best response times possible, you should enable the fuzzy search indexes for all database columns with
a high load of fuzzy searches and for all database columns that are used in performance-critical queries.
The following search options influence the score calculation:
Option spellCheckFactor: Defines the score of terms that get a fuzzy score of 1.0 but are not equal.
Option similarCalculationMode: Defines how the score is calculated when comparing terms. Defines options
to search with substrings of terms.
Option emptyScore: Defines the score of empty column values when searching with non-empty user input
(and the other way round).
12.2.4.7.1 Speeding Up the Fuzzy Search with the Fuzzy
Search Index
You can speed up the fuzzy search on string types by creating a special data structure called a fuzzy search index.
The additional index will increase the total memory footprint of the loaded table. In unfavourable cases the
memory footprint of the column can be doubled.
CREATE COLUMN TABLE mytable
(
col1 VARCHAR(100) FUZZY SEARCH INDEX ON,
col2 NVARCHAR(100) FUZZY SEARCH INDEX ON
);
Additional performance improvements are possible when creating database indexes on the columns.
CREATE INDEX myindex1 ON mytable(col1);
The state of the fuzzy search index can be changed at a later point in time by using the ALTER TABLE statement.
ALTER TABLE mytable ALTER
(
col1 VARCHAR(100) FUZZY SEARCH INDEX OFF,
col2 NVARCHAR(100)
);
336
PUBLI C

Enabling Search
The view SYS.TABLE_COLUMNS shows the current state of the fuzzy search index. When working with attribute
views, this information is also visible in SYS.VIEW_COLUMNS.
SELECT column_name, data_type_name, fuzzy_search_index
FROM table_columns
WHERE table_name = 'MYTABLE';
12.2.4.7.2 Content-Specific Fuzzy Search on String Columns
Use Case: Fuzzy Search - House Numbers
Score Calculation
The house number comparison aims for a 'simple' solution that is easy to understand, gives good results, and
works for most countries. The limitations of the algorithm are:
The algorithm focuses on numeric values - either a single number ('8') or a range of numbers ('8 - 12').
House number additions (for example, the 'a' in '8a') are either equal or not equal.
When comparing two strings containing house numbers with each other, the score is calculated in accordance
with the rules described below.
House number addition. A house number addition in terms of this backlog item is any additional text that is
written before or after the numeric value of a house number.
House number ranges. When a string contains at least two numbers and there is a dash between the first and
second number, this is treated as a house number range. The first number is the lower bound of the range, the last
number is the upper bound.
Multiple numbers. When multiple numbers are part of a house number string that does not define a house
number range, the first number is the house number used for the comparison. All remaining information is used
as a house number addition.
Whitespace characters. For all rules, whitespace characters are ignored when comparing the house numbers.
For the score calculation it does not matter if a house number is given as '8a' or '8 a' or if it is '8-10' or '8 - 10'.
Symmetry. In all examples, the score calculation is symmetric. This means that either string 1 or string 2 can be
the user input and the other string is stored in the database table.
Rule 1 - House Numbers or House Number Ranges Are Identical
For identical house numbers, a score of 1.0 is returned. Identical house numbers are house number strings that
are equal when whitespace characters are ignored.
Examples:
Enabling Search
PUBLI C

String 1 String 2 Score
5 5 1.0
5a 5 a 1.0
8-12 8-12 1.0
9 in the backyard 9 in the backyard 1.0
Rule 2 - House Numbers or House Number Ranges Are Very Similar (House
Number Additions Are Different)
House numbers or house number ranges are considered very similar when the numerical values are identical but
the additional information differs.
Examples:
5 5 a 0.9
5a 5 b 0.9
5 Nr. 5 0.9
8-12 8 - 12a 0.9
8-12 Nr. 8-12 0,9
8-12 8 - 12/5 0.9
8 this is a long text -12 8 - 12a 0.9
7 below 7 0.9
9 9 in the backyard 0.9
in the backyard 9 9 in the backyard 0.9
Rule 3 - House Numbers or House Number Ranges Are Less Similar
House numbers and house number ranges are considered less similar in the following cases:
1. A house number is compared to a house number range and the numerical value of the house number equals
the lower or upper bound of the range.
2. Two house number ranges are compared and the numerical values of either the lower or upper bounds are
equal.
8 8-12 0.8
12a 8-12 0.8
338
PUBLI C

Enabling Search
8-10 8-12 0.8
8-10 8-12 0.8
10-12a 8-12 0.8
8 in the backyard 8-12 0.8
Rule 4 - Overlapping House Number Ranges
House numbers and house number ranges overlap in the following cases:
1. A house number is compared to a house number range and the numerical value of the house number lies
within the range.
2. Two house number ranges are compared and the ranges overlap.
Examples:
10 8-12 0.7
10a 8-12 0.7
9 8-12 0.7
8-12 10-14 0.7
8-12a 10b-14 0.7
Last Rule - House Numbers Are Not Equal
Examples:
5 6 0.0
8a 9a 0.0
6 8-12 0.0
8-10 12-14 0.0
House Number Columns and Other String Search Options
The following search options available for string column types are not valid for house number columns:
Enabling Search
PUBLI C

SpellCheckFactor: When comparing house numbers, the search option 'spellCheckFactor' is ignored. So for
house numbers, the results are always the same as with 'spellCheckFactor=1.0'.
SimilarCalculationMode: When comparing house numbers, the search option 'similarCalculationMode' is
ignored and has no effect on the search result.
Both options are ignored. No error is returned when any of the options is given.
SQL Syntax
To enable the search for house numbers on an (N)VARCHAR column, the FUZZY SEARCH MODE clause is used in
a CREATE TABLE statement.
CREATE COLUMN TABLE tab
(
col1 NVARCHAR(20) FUZZY SEARCH MODE 'housenumber'
);
To enable or disable the house number search mode at a later point in time, use the ALTER TABLE statement. The
fuzzy search mode is not changed if the FUZZY SEARCH MODE clause is omitted.
-- enable housenumber search
ALTER TABLE tab ALTER
(
col1 VARCHAR(20) FUZZY SEARCH MODE 'housenumber'
);
-- disable housenumber search
(
col1 VARCHAR(20) FUZZY SEARCH MODE NULL
);
-- do not change the status of the search mode
(
col1 VARCHAR(20)
);
You can query the state of the fuzzy search mode using the system view TABLE_COLUMNS.
SELECT column_name, data_type_name, fuzzy_search_mode
FROM table_columns
WHERE table_name = 'TAB';
Note
You cannot use a fuzzy search index in combination with the house number search mode.
340
PUBLI C

Enabling Search
Example
The following example creates a table that contains only a single house number column and executes some
searches on this column.
CREATE COLUMN TABLE housenumbers
(
housenumber NVARCHAR(50) FUZZY SEARCH MODE 'housenumber'
);
INSERT INTO housenumbers VALUES ('5');
INSERT INTO housenumbers VALUES ('5a');
INSERT INTO housenumbers VALUES ('5 a');
INSERT INTO housenumbers VALUES ('Nr. 5');
INSERT INTO housenumbers VALUES ('8-12');
INSERT INTO housenumbers VALUES ('8 - 12');
INSERT INTO housenumbers VALUES ('8 - 12a');
INSERT INTO housenumbers VALUES ('Nr. 8-12');
INSERT INTO housenumbers VALUES ('8 - 12/5');
INSERT INTO housenumbers VALUES ('10-12a');
SELECT TO_DECIMAL(SCORE(),3,2), *
FROM housenumbers
WHERE CONTAINS(housenumber, '5', FUZZY(0.8))
ORDER BY TO_DECIMAL(SCORE(),3,2) DESC;
FROM housenumbers
WHERE CONTAINS(housenumber, 'Nr. 5', FUZZY(0.8))
FROM housenumbers
WHERE CONTAINS(housenumber, '8a-12', FUZZY(0.8))
FROM housenumbers
WHERE CONTAINS(housenumber, '10-12', FUZZY(0.8))
FROM housenumbers
WHERE CONTAINS(housenumber, '9 in the BACKYARD', FUZZY(0.8))
Use Case: Fuzzy Search - Postcodes
Postcodes in almost all countries are ordered by region. This means that if the leading characters of the
postcodes of two different addresses are the same, the addresses are near to each other. In Germany, for
example, addresses within large cities share the first or even the first two digits of their postcode.
Enabling Search
PUBLI C

The only exception known to the development team is Cambodia, where postcodes are not ordered by region.
When doing a fuzzy search on addresses, it makes sense to return a higher score for postcodes that are 'near' to a
given user input than for postcodes that are 'far away' from the user input. It makes sense to give a higher weight
to the leading characters and a lower weight to the trailing characters of the postcode.
Valid addresses may contain a country code in front of the postcode (for example, 'D-12345' or 'DE-12345' for a
German address). This is also supported by the fuzzy postcode search.
Score Calculation
Before the fuzzy score is calculated, the postcode strings are standardized.
1. Country codes are separated from the postcode strings. Country codes in this case consist of one to three
letters (a-z only, no numbers) at the beginning of the postcode, followed by a minus sign. Longer words are
not considered a country code because postal standards do not allow country names in front of the postcode.
2. Country codes are standardized to enable a comparison of different codes for the same country, for example,
'D-', 'DE-' and 'DEU-' for German postcodes. All unknown/invalid country codes are standardized to one
special 'dummy' country code.
3. Spaces and dashes are removed from the remaining postcode.
4. All letters are standardized to uppercase.
User Input Country Code Remaining Postcode
71691 71691
716 91 71691
D-71691 D 71691
DE-71 691 DE 71691
D 71691 D71691
Germany-71691 GERMANY71691
GB-A1H 2ZU GB A1H2ZU
A1H-2ZU A1H2ZU
gb-A1h 2zu GB A1H2ZU
XY-12345 XX 12345
zz-12345 XX 12345
AI-2640 AI 2640
The last example is the only known example where the country code is part of the postcode (AI = Anguilla). The
algorithm works here as well, since the country code is also compared. The two examples directly above the AI
example show invalid country codes. Both are standardized to the same non-existent 'dummy' country code.
342
PUBLI C

Enabling Search
Postcode Comparison
The standardized postcodes are compared using a variation of the fuzzy string comparison. This variation gives a
higher weight to the first two characters of the postcode.
Country codes are given the same weight as a single character at the end of the postcode.
Only postcodes with the same country code can get a score of 1.0.
If one country code is given and the second country code is empty, the score of the postcode comparison is
less than 1.0.
If both country codes are given and are different, the score of the postcode comparison is also less than 1.0.
Parameter similarCalculationMode
The search option 'similarCalculationMode' with options 'search' and 'symmetricsearch' is available for postcode
columns.
When using the search option 'similarCalculationMode', a postcode search with a postcode prefix will find all
addresses in a given area.
A search with '71' returns all postcodes beginning with '71'.
A search with '1234' returns all postcodes starting with a sequence similar to '1234' and, with a lower score, all
postcodes that contain a '1234'.
Parameter spellCheckFactor
Two postcodes may be considered identical by the fuzzy string comparison, but may still be different. In this case,
the value of the parameter 'spellCheckFactor' is applied and the score is multiplied by the spellCheckFactor.
Examples of non-equal postcodes that get a score of 1.0 are:
'123456' and '12 34 56'
'7070717' and '7071707'
The default value of the search option spellCheckFactor is 0.9. To disable this feature, set 'spellCheckFactor=1.0'.
Example
The following example uses a spellCheckFactor of 1.0, which is not the default value.
Postcode 1 Postcode 2 Score Remarks
71691 71691 1.0
71691 71 691 1.0
Enabling Search
PUBLI C

Postcode 1 Postcode 2 Score Remarks
71691 81691 0.51 Highest weight on the
first digit
71691 72691 0.7 High weight on the
second digit
71691 71692 0.96 Lower weight on all
other digits
71691 D-71691 0.96 Country code missing
in one column
D-71691 A-71691 0.96 Country codes are
different
71691 D-71692 0.92
D-71691 A-71692 0.92
GB-A1H 2ZU Gb-a1h2zu 1.0
XX-12345 YY-12345 1.0 Invalid country codes
are 'equal'
D-12345 YY-12345 0.96 Valid and invalid
country code
SQL Syntax
(N)VARCHAR columns have to be defined as postcode columns to enable the fuzzy postcode search. You do this
using the FUZZY SEARCH MODE clause.
You can also improve the performance of the postcode search by activating a fuzzy search index and by creating a
database index on the postcode column.
CREATE COLUMN TABLE tab
(
postcode NVARCHAR(20) FUZZY SEARCH INDEX ON FUZZY SEARCH MODE 'postcode'
);
CREATE INDEX myindex1 ON tab(postcode);
You can enable or disable the postcode search at a later point in time with the ALTER TABLE statement. To
disable the postcode search, do not specify the FUZZY SEARCH MODE for the postcode column.
-- enable postcode search
(
postcode NVARCHAR(100) FUZZY SEARCH MODE 'postcode'
);
-- disable postcode search
(
344
PUBLI C

Enabling Search
postcode NVARCHAR(100) FUZZY SEARCH MODE NULL
);
-- do not change the status of the search mode
(
postcode NVARCHAR(100)
);
You can query the status of the fuzzy search index and the fuzzy search mode from the system view
TABLE_COLUMNS.
SELECT column_name, data_type_name, fuzzy_search_index, fuzzy_search_mode
FROM table_columns
WHERE table_name = 'TAB';
Example
CREATE COLUMN TABLE postcodes
(
postcode NVARCHAR(50) FUZZY SEARCH INDEX ON FUZZY SEARCH MODE 'postcode'
);
INSERT INTO postcodes VALUES ('71691');
INSERT INTO postcodes VALUES ('716 91');
INSERT INTO postcodes VALUES ('A1H 2ZU');
INSERT INTO postcodes VALUES ('A1H2ZU');
INSERT INTO postcodes VALUES ('D-71691');
INSERT INTO postcodes VALUES ('A-71691');
INSERT INTO postcodes VALUES ('A-71692');
INSERT INTO postcodes VALUES ('DE-71 691');
INSERT INTO postcodes VALUES ('D 71691');
INSERT INTO postcodes VALUES ('GB-A1H 2ZU');
INSERT INTO postcodes VALUES ('XX-12345');
FROM postcodes
WHERE CONTAINS(postcode, '71691', FUZZY(0.5, 'spellCheckFactor=1.0'))
FROM postcodes
WHERE CONTAINS(postcode, 'D-71691', FUZZY(0.5, 'spellCheckFactor=1.0'))
FROM postcodes
WHERE CONTAINS(postcode, 'Gb-a1h2zu', FUZZY(0.5, 'spellCheckFactor=1.0'))
FROM postcodes
WHERE CONTAINS(postcode, 'YY-12345', FUZZY(0.5, 'spellCheckFactor=1.0'))
Enabling Search
PUBLI C

SELECT TO_DECIMAL(SCORE(),3,2), *FROM postcodes
FROM postcodes
12.2.4.7.3 Speeding Up the Fuzzy Search on String Columns
Fuzzy search on string types can be sped up by creating a special data structure called a fuzzy search index. The
additional index will increase the total memory footprint of the loaded table. In unfavourable cases the memory
footprint of the column can be doubled.
(
col1 VARCHAR(100) FUZZY SEARCH INDEX ON,
col2 NVARCHAR(100) FUZZY SEARCH INDEX ON
);
Additional performance improvements are possible when creating database indexes on the columns.
The state of the fuzzy search index can be changed at a later point in time by using the ALTER TABLE statement.
(
col1 VARCHAR(100) FUZZY SEARCH INDEX OFF,
col2 NVARCHAR(100)
);
The view SYS.TABLE_COLUMNS shows the current state of the fuzzy search index. When working with attribute
views, this information is also visible in SYS.VIEW_COLUMNS.
SELECT column_name, data_type_name, fuzzy_search_index
FROM table_columns
12.2.4.8 Fuzzy Search on Text Columns
Text types support a more sophisticated kind of fuzzy search. Texts are tokenized (split into terms), and the fuzzy
comparison is performed term by term. When searching with 'SAP' for example, a record like 'SAP Deutschland
AG & Co. KG' gets a high score, because the term 'SAP' exists in both texts. A record like 'SAPPHIRE NOW
Orlando' gets a lower score, because 'SAP' is just a part of the longer term 'SAPPHIRE' (3 of 8 characters match).
Fuzzy search on text columns replaces non-fault tolerant statements like
346
PUBLI C

Enabling Search
SELECT ... FROM documents WHERE doc_content LIKE '% Driethanolamyn %' ...
The following SQL data types are supported:
TEXT
SHORTTEXT
fulltext index
A fulltext index is an additional index structure that can be defined for non-text columns to add text search
features. Supported column types include NCLOB and NVARCHAR.
It is possible to speed up the fuzzy search by creating data structures called 'fuzzy search indexes', which are
used for faster calculation of the fuzzy score. These indexes exist in the memory only, so no additional disk space
is needed.
To get the best response times possible, you should enable the fuzzy search indexes for all database columns with
a high load of fuzzy searches and for all database columns that are used in performance-critical queries.
Fuzzy Search on SQL Type TEXT
A call to contains that references a TEXT column is automatically processed as a text search. In this case, the
mode textsearch=compare and all fuzzy search options are allowed:
(
col1 TEXT
);
SELECT score() AS score, * FROM mytable WHERE contains(col1, 'a b', fuzzy(0.8,
'textsearch=compare'));
Fuzzy Search on SQL Type SHORTTEXT
When a SHORTTEXT column is created, a column of column store type cs_string and a second hidden text
column are created. A call to contains that references the SHORTTEXT column is automatically redirected by the
freestyler to the additional hidden TEXT column. In this case, the mode textsearch=compare and all fuzzy search
options are allowed:
(
col1 SHORTTEXT(200)
);
Enabling Search
PUBLI C

Fuzzy Search on a FULLTEXT INDEX
When a full text index is created on a column that is not of type TEXT (e.g. NVARCHAR, NCLOB, ...) a hidden text
column is added to the table. A call to contains that references the non-TEXT column is automatically redirected
by the freestyler to the additional text column. In this case, the mode textsearch=compare and all fuzzy search
options are allowed:
(
col1 NVARCHAR(2000)
);
CREATE FULLTEXT INDEX myindex ON mytable(col1);
Merge Delta for Better Performance
When inserting or loading a large number of rows into a table that has a TEXT or SHORTTEXT column or uses a
FULLTEXT INDEX, it is important to merge the delta part of the table in order to ensure satisfactory search
performance.
A delta merge can be started manually using the following SQL statement:
MERGE DELTA OF mytable;
Alternatively, a delta merge can be triggered automatically by the mergedog process.
12.2.4.8.1 Multi-Token Search on Text Columns
When using more than one token in a query, the default content type is AND (... WHERE CONTAINS (mycolumn,
'software firm', FUZZY(0.5)) ... for example will return entries that contain a token similar to 'software' and a token
similar to 'firm').
Alternatively, you can use OR by adding the key word between the tokens (... WHERE CONTAINS (mycolumn,
'apple OR 'banana', FUZZY(0.5)) ... for example will return entries that contain a token similar to 'apple' and
entries that contain a token similar to 'banana').
PHRASE is similar to AND but restricts hits to those containing the tokens as a phrase, in other words in the same
order and with nothing between them. A PHRASE is indicated by adding double quotes around the tokens, within
the single quotes (... WHERE CONTAINS (mycolumn, '"day dream"', FUZZY(0.5)) ... for example will not return an
entry containing 'I dream of a day').
The content type AND that is used for a full-text search (default behavior: textSearch=fulltext) is implemented as
a logical AND to achieve better performance. A search for 'Miller & Miller AG' for example with content type AND
matches 'Miller AG'.
For duplicate detection and to compare company names, product names and so on, textSearch=compare should
be used. The earch results are better here because of the strict AND comparison that is used. In other words,
348
PUBLI C

Enabling Search
when searching for 'Miller & Miller' with content type AND, only records that contain the term 'Miller' at least twice
are returned.
A strict AND assigns terms from the user input to terms in the database entry only once (and vice versa). For
more information, see Partially Matching with Parameter andThreshold [page 351].
Parameters Influencing the Score
Name of Option Range Default Applies to Types
bestMatchingTokenWeigh
t
0.0..1.0 0 TEXT
considerNonMatchingTok
ens
max, min, all, input, table max TEXT
Formula for score calculation:
score = hingTokenWeight x max(tokenScores) + (1-bestMatchingTokenWeight) x ((tokenScore)/tokenCount)
tokenCount is determined according to the parameter considerNonMatchingTokens as follows:
input: Use search term token count
table: Use column value token count
all: Use sum of search and column value token count (where the numerator is multiplied by 2)
min: Use smaller value of token counts
max: Use larger value of token counts (default)
Recommendations for specific search content types
If you are using an "OR" search (searching for "this or that"), you should set considerNonMatchingTokens to get a
meaningful score assessment.
Example
DROP TABLE test_table;
CREATE COLUMN TABLE test_table
(
);
INSERT INTO test_table VALUES ('1','eins');
INSERT INTO test_table VALUES ('2','eins zwei');
INSERT INTO test_table VALUES ('3','eins zwei drei');
INSERT INTO test_table VALUES ('4','eins zwei drei vier');
INSERT INTO test_table VALUES ('5','eins zwei drei vier funf');
SELECT TO_DECIMAL(SCORE(),3,2) AS score, * FROM test_table
WHERE CONTAINS(t, 'eins zwoi drei',
ORDER BY score DESC, id;
Enabling Search
PUBLI C

SCORE ID T
1 3 eins zwei drei
1 4 eins zwei drei vier
1 5 eins zwei drei vier funf
CT TO_DECIMAL(SCORE(),3,2) AS score, * FROM test_table
FUZZY(0.5,'textSearch=compare,bestMatchingTokenWeight=0'))
SCORE ID T
0.92 3 eins zwei drei
0.8 4 eins zwei drei vier
0.71 5 eins zwei drei vier funf
FUZZY(0.5,'textSearch=compare,bestMatchingTokenWeight=0,considerNonMatchingTokens=al
l'))
SCORE ID T
FUZZY(0.5,'textSearch=compare,bestMatchingTokenWeight=0,considerNonMatchingTokens=in
put'))
SCORE ID T
Parameters Influencing the Result Set
Option Range Default Applies to Types
andSymmetric on,off,true,false off TEXT
andThreshold 0.0..1.0 1.0 TEXT
350
PUBLI C

Enabling Search
For examples, see Symmetric Search with Parameter andSymmetric [page 352] and Partially Matching with
Parameter andThreshold [page 351] .
12.2.4.8.2 Multi-Token Search with Soft AND
Partially Matching with Parameter andThreshold
It is possible to specify a 'partial AND' that requires a subset of request tokens only to match the reference tokens.
You will then get better results when comparing data like company names.
The 'andThreshold' parameter defines the percentage of tokens that have to match when comparing the user
input with a row stored in a TEXT column. In other words, the ratio between the number of matching tokens and
the number of input tokens has to be greater than or equal to the given andThreshold.
Option Name Range Default Applies to Types Description
andThreshold 0.0..1.0 1.0 TEXT Determines the
percentage of
tokens that need to
match
andThreshold = 1.0 -> all tokens have to match, 'strict AND'
0.0 < andThreshold < 1.0 -> some of the tokens have to match, 'soft AND'
andThreshold = 0.0 -> at least one token has to match, 'OR'
Summary
The andThreshold parameter is available as a searchOption.
Allowed values are between 0.0 and 1.0. An SQL error is returned for values outside this range.
The parameter works on TEXT columns only and is ignored for all other SQL types.
The parameter is used with content type AND only and is ignored for all other content types.
The parameter influences performance.
Example
DROP TABLE test_soft_and;
CREATE COLUMN TABLE test_soft_and
(
);
INSERT INTO test_soft_and VALUES ('1','eins');
INSERT INTO test_soft_and VALUES ('2','eins zwei');
INSERT INTO test_soft_and VALUES ('3','eins zwei drei');
INSERT INTO test_soft_and VALUES ('4','eins zwei drei vier');
INSERT INTO test_soft_and VALUES ('5','eins zwei drei vier funf');
Enabling Search
PUBLI C

Search with andThreshold
SELECT SCORE() AS score, * FROM test_soft_and
WHERE CONTAINS(T, 'eins XXX drei vier',
FUZZY(0.5,'andThreshold=0.75,bestMatchingTokenWeight=0.5,textSearch=compare'))
SCORE ID T
Symmetric Search with Parameter andSymmetric
In addition to the parameter andThreshold, it is possible to specify a 'symmetric AND' that also returns a record
when all tokens of a database entry are part of the user input. The parameter 'andSymmetric' was added to the
fuzzy search and activates the symmetric AND comparison when comparing the user input with a row stored in a
TEXT column.
Option Name Range Default Applies to Types Short Description
andSymmetric on,off,true,false off TEXT Activates a
symmetric AND
content search
Example
When using content type AND, a search with 'SAP Deutschland AG' returns 'SAP Deutschland AG & Co KG' but
not 'SAP Deutschland' or 'SAP Walldorf'. When using the symmetric and, the search with 'SAP Deutschland AG'
returns 'SAP Deutschland AG & Co KG' and 'SAP Deutschland', but not 'SAP Walldorf'.
Summary
The parameter andSymmetric is available as a searchOption.
Allowed values are [on,true] and [off, false]. An SQL error is returned for other values.
The parameter works on TEXT columns only and is ignored for all other SQL types.
The parameter is used with content type AND only and is ignored for all other content types.
The parameter works in combination with andThreshold to activate a symmetric 'soft AND'.
The parameter influence the performance.
Example
DROP TABLE test_soft_and;
CREATE COLUMN TABLE test_soft_and
(
);
352
PUBLI C

Enabling Search
INSERT INTO test_soft_and VALUES ('1','one');
INSERT INTO test_soft_and VALUES ('2','one two');
INSERT INTO test_soft_and VALUES ('3','one two three');
INSERT INTO test_soft_and VALUES ('4','one two three four');
INSERT INTO test_soft_and VALUES ('5','one two three four five');
Search with andSymmetric=off
WHERE CONTAINS(T, 'one two three', FUZZY(0.5,
'andSymmetric=off,bestMatchingTokenWeight=0.5,textSearch=compare'))
SCORE ID T
1 3 one two three
0.933012723922729 4 one two three four
0.887298345565796 5 one two three four five
Search with andSymmetric=on
WHERE CONTAINS(T, 'one two three',
FUZZY(0.5,'andSymmetric=on,bestMatchingTokenWeight=0.5,textSearch=compare'))
SCORE ID T
1 3 one two three
0.933012723922729 4 one two three four
0.90824830532074 2 one two
0.887298345565796 5 one two three four five
0.788675129413605 1 one
12.2.4.8.3 Option abbreviationSimilarity
The abbreviation similarity option allows you to search for and with initial characters. You can thus find
abbreviations when searching with long strings or find long strings when searching with initial characters.
abbreviationSimilarit
y
0.0..1.0 0.0 TEXT Defines the similarity
that is returned for a
matching initial
character.
Note
abbreviationSimilarity = 0.0 disables the abbreviation search.
Enabling Search
PUBLI C

The abbreviationSimilarity option works for TEXT fields only. It is available for all term actions available in SQL, like
EXACT, SIMILAR, LINGUISTIC. It accepts values between 0.0 and 1.0 . If the given value for abbreviationSimilarity
is out of this range, the system returns a SQL error.
Search for inital characters:
"Peter" finds "P." and "P"
"P." always finds "P" with similarity 1.0
"Hans-Peter" finds "Hans P", "H. Peter" and "H.-P."
Search with initial characters:
"P." and "P" find "Peter"
"P." always finds "P" with similarity 1.
"Hans P" and "H P" find "Hans-Peter"
Example
The abbreviationSimilarity option is used to search for and find a word using its first character and vice versa with
a given score. With abbreviationSimilarity = 0.9, a SELECT retrieves word with SCORE = 0.9 if you search with
w (and vice versa).
CREATE COLUMN TABLE abbrev
(
name SHORTTEXT(200) FUZZY SEARCH INDEX ON
);
INSERT INTO abbrev VALUES ('1','Peter');
INSERT INTO abbrev VALUES ('2','Hans');
INSERT INTO abbrev VALUES ('3','H.');
INSERT INTO abbrev VALUES ('4','P.');
INSERT INTO abbrev VALUES ('5','Hans-Peter');
INSERT INTO abbrev VALUES ('6','H.-P.');
INSERT INTO abbrev VALUES ('7','HP');
INSERT INTO abbrev VALUES ('8','G Gerd');
INSERT INTO abbrev VALUES ('9','G');
INSERT INTO abbrev VALUES ('10','Gerd');
Search one token with abbreviationSimilarity
SELECT SCORE() AS score, id, name FROM abbrev
WHERE CONTAINS(name, 'HP', FUZZY(0.5,
'abbreviationSimilarity=0.80,textSearch=compare'))
SCORE ID NAME
1 7 HP
0.800000011920929 3 H.
0.565685451030731 6 H.-P.
354
PUBLI C

Enabling Search
Search two tokens with abbreviationSimilarity
WHERE CONTAINS(name, 'Hans Peter',FUZZY(0.5,
SCORE ID NAME
1 5 Hans-Peter
0.800000011920929 6 H.-P.
Search two tokens with abbreviationSimilarity
WHERE CONTAINS(name, 'Hans Peter',FUZZY(0.5,
SCORE ID NAME
0.905538558959961 8 G Gerd
12.2.4.8.4 Option minTextScore
The minTextScore option allows you to set the score a text field has to reach to be a match.
A text field can contain more than one token. Each token is evaluated against the fuzzySimilarity parameter which
defines the minimum score a single token has to reach. The overall score of a text field differs from the
fuzzySimilarity because the overall Text score is computed from the fuzzySimilarity of all tokens and parameters
found, like bestMatchingTokenWeight, considerNonMatchingTokens and term mapping weights.
minTextScore defines the minimum score the whole content of a text field has to reach.
minTextScore 0.0..1.0 0.0 TEXT Minimum score that
all records in the
result have to reach.
A minTextScore value of 0.0 means that all records that match the conditions defined by fuzzySimilarity and the
search options are returned. The result list will not be cut.
Note
If you use a fuzzySimilarity of 0.0, the minTextScore parameter becomes redundant.
minTextScore does not work with a freestyle or fulltext search.
When using minTextScore in combination with the emptyScore parameter, rows found because of the
emptyScore parameter are returned even if emptyScore is less than minTextScore. See the code example
for emptyScore.
DROP TABLE tab_mintextscore;
Enabling Search
PUBLI C

CREATE COLUMN TABLE tab_mintextscore
(
t TEXT FAST PREPROCESS ON FUZZY SEARCH INDEX ON
);
INSERT INTO tab_mintextscore VALUES ('1','Bert');
INSERT INTO tab_mintextscore VALUES ('2','Berta');
INSERT INTO tab_mintextscore VALUES ('3','Bart');
Search on a text column
SELECT SCORE() AS score, id, t FROM tab_mintextscore
WHERE CONTAINS(t, 'Ernie OR Bert', FUZZY(0.100, 'textSearch=compare,
bestMatchingTokenWeight=0,minTextScore=0.70'))
SCORE ID T Description
0.7 1 Bert << rank value of 0.7 is
reached, this is a match
0.65 1 Berta << minTextScore not
reached - cut off
0.52 1 Bart << minTextScore not
reached - cut off
12.2.4.8.5 Option textSearch
The textSearch option is used to select the search algorithm for TEXT columns:
textSearch=fulltext (default value): A full-text search is performed on a TEXT column. IDF calculation or
specialOrRanking (depending on search flags) is used. This is the 'old' NewDB behavior.
textSearch=compare: A search similar to a Fuzzy Double search is performed. Additional search options are
enabled.
The value of the textSearch search option defines which other search options are allowed for a search.
Rank calculation
When 'fulltext' is specified, the search is performed using IDF or specialOrRanking (depending on the search
flags). The fuzzy score is not calculated.
When 'compare' is specified, the fuzzy score is calculated using bestMatchingTokenWeight and
considerNonMatchingTokens. The user does not have to set these options. In this case, the default values are
used.
356
PUBLI C

Enabling Search
Example
WHERE CONTAINS(T, 'eins', FUZZY(0.5, 'textSearch=compare'))
Select with 'Fuzzy Score' (textSearch=compare)
WHERE CONTAINS(T, 'eins', FUZZY(0.5, 'textSearch=compare'))
SCORE ID T
1.0 1 eins
0.7071067690849304 2 eins zwei
0.5773502588272095 3 eins zwei drei
Select with 'IDF Score' (textSearch=fulltext)
WHERE CONTAINS(T, 'eins', FUZZY(0.5, 'textSearch=fulltext'))
SCORE ID T
0.888888955116272 1 eins
0.6666666865348816 2 eins zwei
0.5333333611488342 3 eins zwei drei
12.2.4.8.6 Stopwords
Usage
Stopwords are terms that are less significant for a search and are therefore not used to generate the result set. In
other words, the search is carried out as if the stopwords are not present (either in the user input or in the
database column).
However, stopwords influence the score that is calculated. A record with stopwords identical to the user input
gets a higher score than a record with differing or missing stopwords.
Stopwords can be defined either as single terms or as stopword phrases consisting of multiple terms. Stopword
phrases are only applied when all terms of the stopword appear in exactly the given order.
Use case example: When searching for company names, the legal form (Ltd, SA, AG, and so on) is less significant
and less selective than the other parts of the name.
Enabling Search
PUBLI C

Stopwords are stored in a column-store table with the following format:
CREATE COLUMN TABLE mystopwords
(
stopword_id VARCHAR(32) PRIMARY KEY,
list_id VARCHAR(32) NOT NULL,
language_code CHAR(2),
term NVARCHAR(200) NOT NULL
);
Stopwords are language dependent. It is possible to define the language that a stopword is valid for. You can also
define stopwords for all languages by not setting a language.
As with term mappings, stopwords can be grouped together in multiple groups. Groups of stopwords are
identified by the value of the list_id column that is part of the stopword table.
INSERT INTO mystopwords VALUES (1, 'legalform', '', 'Ltd');
INSERT INTO mystopwords VALUES (2, 'legalform', 'de', 'GmbH');
INSERT INTO mystopwords VALUES (3, 'legalform', 'de', 'Gesellschaft mit
beschrnkter Haftung');
INSERT INTO mystopwords VALUES (4, 'legalform', 'de', 'AG');
INSERT INTO mystopwords VALUES (5, 'legalform', 'de', 'Aktiengesellschaft');
To activate stopwords for a search on a TEXT column, you need to provide two search options (similar to the
options used for term mappings):
stopwordListId=mylist1,mylist2,mylist3
stopwordTable=[<schemaname>.]<tablename>
SELECT TO_DECIMAL(SCORE(),3,2) as score, company FROM mydata
WHERE CONTAINS(company, 'xy gmbh', FUZZY(0.7, 'textsearch=compare,
stopwordTable=MYSTOPWORDS, stopwordListId=legalform'))
To activate language-specific stopwords, you must provide the language parameter:
SELECT TO_DECIMAL(SCORE(),3,2) as score, company FROM mydata
WHERE CONTAINS(company, 'xy gmbh', FUZZY(0.7, 'textsearch=compare,
stopwordTable=MYSTOPWORDS, stopwordListId=legalform'), language('de'))
Note
In this case, all stopwords where language_code is set to 'de' or empty will be used. Any stopwords with other
language codes will be ignored.
Stopwords are removed from the search term first. In this example, the result set of the search is the same as for
the search '... WHERE CONTAINS(company, 'xy', ...'.
When calculating the score, the fuzzy scores of the non-stopword terms have the biggest influence on the
resulting score. Stopwords in the user input and in the database records are also given less weight than the non-
stopword terms, so records with matching stopwords get a higher score than records with differing or missing
stopwords.
The result of the above example is as follows:
358
PUBLI C

Enabling Search
Score Company Comment
1.00 XY GmbH
0.95 XY Missing stopword
0.95 XY Aktiengesellschaft Differing stopword
0.92 XY Gesellschaft mit beschrnkter
Haftung
Many differing stopwords
0.78 XY Company Additional non-matching term, no
stopword
The value given for stopwordTable can be any valid SQL identifier as defined in the SQL reference manual. If
schema is omitted, the current schema is used. The following examples all reference the same stopword table.
SET SCHEMA schema1;
SELECT ... WHERE CONTAINS(c, 'xy', FUZZY(0.7,
'stopwordTable=schema1.mystopwords, ...'))...;
SELECT ... WHERE CONTAINS(c, 'xy', FUZZY(0.7,
'stopwordTable="SCHEMA1"."MYSTOPWORDS", ...'))...;
SELECT ... WHERE CONTAINS(c, 'xy', FUZZY(0.7, 'stopwordTable=mystopwords, ...'))...;
Stopword Example
CREATE COLUMN TABLE stopwords
(
language_code CHAR(2) NOT NULL,
);
(
);
INSERT INTO companies VALUES ('1', 'SAP AG');
INSERT INTO companies VALUES ('2', 'SAP Aktiengesellschaft');
INSERT INTO stopwords VALUES ('1', '01', 'de', 'AG');
INSERT INTO stopwords VALUES ('2', '01', 'de', 'Aktiengesellschaft');
INSERT INTO stopwords VALUES ('3', '01', 'de', 'blub');
Query 1: User input without stopwords. Stopwords in the database table only.
SELECT TO_DECIMAL(SCORE(), 3, 2) AS score, *
FROM companies
WHERE CONTAINS(companyname, 'sap', FUZZY(0.8, 'stopwordTable=stopwords,
stopwordListId=01, textSearch=compare'))
ORDER BY score DESC, ID;
0.95 1 SAP AG
0.95 2 SAP Aktiengesellschaft
Enabling Search
PUBLI C

Query 2: User input with stopword. Other stopwords in the database table.
FROM companies
WHERE CONTAINS (companyname, 'sap blub', FUZZY(0.8, 'stopwordTable=stopwords,
stopwordListId=01, textSearch=compare'))
0.95 1 SAP AG
Query 3: User input with stopword. One record with matching stopword, one record with differing stopword.
FROM companies
WHERE CONTAINS (companyname, 'sap aktiengesellschaft', FUZZY(0.8,
'stopwordTable=stopwords, stopwordListId=01, textSearch=compare'))
1 2 SAP Aktiengesellschaft
0.95 1 SAP AG
Query 4: User input with two stopwords. Database records with one matching stopword.
FROM companies
WHERE CONTAINS (companyname, 'sap ag aktiengesellschaft', FUZZY(0.8,
'stopwordTable=stopwords, stopwordListId=01, textSearch=compare'))
0.97 1 SAP AG
Stopwords Combined With Term Mappings
When stopwords and term mappings are combined in a single query, term mappings are applied first. Stopwords
are then applied to all variations of the search term created by the term mappings.
Let us assume that you have defined the following term mapping:
Term 1 Term 2
Incredible Busy Machines IBM
Ltd Limited
Now you search for "Incredible Busy Machines Ltd".
360
PUBLI C

Enabling Search
The search would be carried out for all possible search terms:
Search Terms
Incredible Busy Machines Ltd
Incredible Busy Machines Limited
IBM Ltd
IBM Limited
Let us assume that you have defined the following stopwords:
Stopword
busy machines
ltd
The stopwords will not be searched, so the resulting search terms would be:
Search Terms
Incredible Busy Machines Ltd
Incredible Busy Machines Limited
IBM Ltd
IBM Limited
12.2.4.8.7 Term Mappings
Facts About Term Mappings
Term mappings have the following characteristics:
Term mappings can be used to extend the search by adding additional search terms to the user input.
When the user enters a search term, the search term is expanded and synonyms, hypernyms, hyponyms, and
so on are added. The result that is returned to the user contains additional records or documents related to
the search term that may be useful to the user.
Term mappings are defined in a column table and can be changed at any time.
The current term mapping definition is applied when a search is started. The definition of term mappings does
not change the data that is stored in the database tables (in contrast to the definition of synonyms in Text
Analysis, where a change of synonyms requires the text data to be reloaded or reindexed).
Term mappings can be grouped.
Each group of term mappings is identified by a list_id that is stored in the term mapping table. By grouping
term mappings, it is possible to apply different sets of term mappings to different columns of a table. For
example, you may want to use some term mappings when searching company names and other term
mappings when searching documents. When starting a search, it is possible to specify the term mapping
list_ids that are applied to each column.
Term mappings can be assigned a weight.
Enabling Search
PUBLI C

In this case, records that are found because of a term mapping will get a lower score than records found with
the original user input. From the user's view, the result list is sorted in a more useful way.
Term mappings are defined as a unidirectional replacement.
For a term mapping definition of 'term1' -> 'term2', 'term1' is replaced with 'term2', but 'term2' is not replaced
with 'term1'. This is helpful if you want a search with a hypernym to find all hyponyms, but not the other way
round. If a bidirectional replacement is needed (as for synonyms), both directions have to be added to the
term mapping table.
Term mappings are language dependent.
It is possible to define the language that a term mapping is valid for. You can also define term mappings for all
languages by not setting a language.
Use Cases
Synonyms
If you have a large database of company names, you might want to map the companies' legal forms.
For example:
Searching for You would also like to find With a weight of
AG Aktiengesellschaft 1.0
Ltd Limited 1.0
Since these are synonyms, the term mappings have to be added to the term mapping table in both directions, as
shown in the example below.
Usually, synonym definitions get a weight of 1.0, because records found when the term mapping is applied are as
good as records found with the original user input.
Hypernyms, Hyponyms
If you search with a hypernym, you might also find other documents related to this topic.
For example:
Searching for You would also like to find With a weight of
car VW Golf 0.8
Since these are not synonyms, and a search with 'VW Golf' should not return all documents about cars, the term
mapping is added to the term mapping table in this direction only.
Format of the Term Mapping Table
Column Name Type Primary Key Description Comment
MAPPING_ID VARCHAR(32) x Primary key For example, a GUID
362
PUBLI C

Enabling Search
Column Name Type Primary Key Description Comment
LIST_ID VARCHAR(32) Term mapping list ID Used to group term
mappings
LANGUAGE_CODE CHAR(2) Language code
(ISO2)
NULL: term mapping
is valid for all
languages
TERM_1 NVARCHAR(200) Term 1, the term to
be replaced
TERM_2 NVARCHAR(200) Term 2, the term
that replaces term 1
WEIGHT DECIMAL Weight, 0.0 <=
Weight <= 1.0
The definition of the term mapping table is checked, so tables with other column names or data types cannot be
used for a fuzzy search. Nevertheless, the table may contain additional columns that are ignored by the fuzzy
search engine.
Example code for creating a term mapping table:
CREATE COLUMN TABLE termmappings
(
mapping_id VARCHAR(32) PRIMARY KEY,
list_id VARCHAR(32),
term_1 NVARCHAR(255),
weight DECIMAL
);
Basic Example
The value given for the termMappingTable parameter can be any valid SQL identifier as defined in the SQL
reference manual. If no schema is specified, the current schema is used.
(
list_id VARCHAR(32),
weight DECIMAL
);
(
);
INSERT INTO companies VALUES ('1','SAP AG');
INSERT INTO companies VALUES ('2','SAP Aktiengesellschaft');
INSERT INTO termmappings VALUES ('1','01','de','AG','Aktiengesellschaft','0.9');
INSERT INTO termmappings VALUES ('2','01','de','Aktiengesellschaft','AG','0.9');
SELECT TO_DECIMAL(SCORE(),3,2) AS score, *
Enabling Search
PUBLI C

FROM companies
WHERE CONTAINS(companyname, 'sap aktiengesellschaft',
FUZZY(0.8, 'termMappingTable=termmappings,termMappingListId=01,textSearch=compare'))
0.9 1 SAP AG
SELECT TO_DECIMAL(SCORE(),3,2) AS score, * FROM companies
WHERE CONTAINS(companyname, 'sap ag',
FUZZY(0.8,'termMappingTable=TERMMAPPINGS,termMappingListId=01,textSearch=compare'))
1.0 1 SAP AG
To activate language-specific term mappings, you must provide the language parameter:
SELECT TO_DECIMAL(SCORE(),3,2) AS score, * FROM companies
WHERE CONTAINS(companyname, 'sap ag',
FUZZY(0.8,'termMappingTable=TERMMAPPINGS,termMappingListId=01,textSearch=compare'),
language('de'))
Note
In this case, all term mappings where language_code is set to 'de' or empty are used. Any term mappings with
other language codes are ignored.
12.2.4.8.8 Speeding Up the Fuzzy Search on Text Columns
You can speed up the fuzzy search for all SQL types except DATE by creating a special data structure called a
fuzzy search index. The additional data structures increase the total memory footprint of the loaded table. In
unfavorable cases, the memory footprint of the column can double.
TEXT Columns
TEXT columns offer the option 'FUZZY SEARCH INDEX' to enable and disable the indexes when creating a table:
(
col1 TEXT FUZZY SEARCH INDEX ON
);
You can change this later using the ALTER TABLE command.
(
364
PUBLI C

Enabling Search
col1 TEXT FUZZY SEARCH INDEX OFF
);
The view SYS.TABLE_COLUMNS shows the current state of the fuzzy search indexes. When working with
attribute views, this information is also visible in SYS.VIEW_COLUMNS.
SHORTTEXT Columns
SHORTTEXT columns offer the option 'FUZZY SEARCH INDEX' to enable and disable the indexes when creating a
table:
(
col2 SHORTTEXT(100) FUZZY SEARCH INDEX ON
);
You can change this later using the ALTER TABLE command.
(
col2 SHORTTEXT(100) FUZZY SEARCH INDEX OFF
);
Note
The following limitation applies: since SHORTTEXT creates a hidden column of type text, the state of the fuzzy
search index is not visible in SYS.TABLE_COLUMNS and SYS.VIEW_COLUMNS.
SYS.FULLTEXT_INDEXES can be used to query the state of the fuzzy search indexes.
SELECT index_name, internal_column_name, fuzzy_search_index
FROM fulltext_indexes
Full-Text Indexes
A full-text index offers the option 'FUZZY SEARCH INDEX' to enable and disable the indexes when creating a full-
text index:
(
col3 NVARCHAR(2000)
);
CREATE FULLTEXT INDEX myindex ON mytable(col3)
FUZZY SEARCH INDEX ON;
You can change this later using the ALTER FULLTEXT INDEX command:
ALTER FULLTEXT INDEX myindex FUZZY SEARCH INDEX OFF;
The view SYS.FULLTEXT_INDEXES shows the current state of the fuzzy search indexes.
SELECT index_name, internal_column_name, fuzzy_search_index
FROM fulltext_indexes
Enabling Search
PUBLI C

12.2.4.9 Fuzzy Search on DATE Columns
A fuzzy search on DATE columns supports two types of error:
Date-specific typos
Dates lying within a user-defined maximum distance
It is not possible to create additional data structures for date types to speed up the search. The queries run with
optimal performance without any database tuning.
Note
Fuzzy search for dates works with valid dates only. A search with an invalid date does not return any results.
The following example returns no results, because '31' is not a valid month:
...WHERE CONTAINS(mydate, '2012-31-01', FUZZY(0.8))...
Score Calculation for Typos
Instead of using Levenshtein distance or other string-compare algorithms, the following date-specific typos and
errors are defined as similar:
1. One wrong digit at any position (for example, 2011-08-15 instead of 2011-08-25). This type of error gets a
score of 0.90.
2. Two digits interchanged within one component (day, month, or year) (for example, 2001-01-12, 2010-10-12, or
2010-01-21 instead of 2010-01-12). This type of error gets a score of 0.85.
3. Month and day interchanged (US versus European date format) (for example, 2010-10-12 instead of
2010-12-10). This type of error gets a score of 0.80.
Only one of these errors is allowed. Dates with more than one error are not considered similar, so the score is 0.0.
Dates with a score less than the fuzzySimilarity parameter are not returned.
Example:
SELECT TO_DECIMAL(SCORE(),3,2) AS score, * FROM dates
WHERE CONTAINS(dateofbirth, '2000-01-10', FUZZY(0.8))
For this example we get:
2000-01-09 -> 0.00 (not returned to the user)
2000-01-10 -> 1.00
2000-01-11 -> 0.90
2000-01-12 -> 0.90
...
2000-01-21 -> 0.00 (not returned to the user)
...
366
PUBLI C

Enabling Search
2000-10-01 -> 0.80
Score Calculation for Date Distance
The maximum allowed distance between dates can be defined using the search option 'maxDateDistance', which
defines a number of days.
The default for this option is 0, that is, the feature is disabled. This is shown in the following example:
WHERE CONTAINS(dateofbirth, '2000-01-10', FUZZY(0.95, 'maxDateDistance=5'))
This query returns all dates between 2000-01-05 and 2000-01-15.
The fuzzy score for dates is calculated as follows:
1. The identical date gets a score of 1.0.
2. The date that is maxDateDistance days away from the search input gets a score that equals the
fuzzySimilarity parameter (0.95 in the example above).
3. The score of dates between the identical date and maxDateDistance is calculated as a linear function between
the two dates defined above. In other words, for each day the score is reduced by ((1-fuzzySimilarity) /
maxDateDistance).
4. For dates outside the range of maxDateDistance, the score is 0.0.
Therefore, for the example above we get:
2000-01-04 -> 0.00
2000-01-05 -> 0.95
2000-01-06 -> 0.96
...
2000-01-09 -> 0.99
2000-01-10 -> 1.0
2000-01-11 -> 0.99
2000-01-12 -> 0.98
...
2000-01-15 -> 0.95
2000-01-16 -> 0.00
The distance between dates is calculated following the rules of the Gregorian calendar.
The special case 'fuzzySimilarity = 1.0' and maxDateDistance=n is allowed and returns all dates within a range of n
days with a rank of 1.0.
Enabling Search
PUBLI C

Dates That Meet Both Conditions
If a date meets the conditions of a typo and also the conditions of the maxDateDistance parameter, two scores
are calculated for the same date. In this case, the score() function returns the highest of both scores. This is
shown in the following example:
WHERE CONTAINS(dateofbirth, '2000-01-10', FUZZY(0.8, 'maxDateDistance=5'))
This query returns the following:
2000-01-04 -> 0.00
2000-01-05 -> 0.80
2000-01-06 -> 0.84
2000-01-07 -> 0.88
2000-01-08 -> 0.92
2000-01-09 -> 0.96
2000-01-10 -> 1.0
2000-01-11 -> 0.96
2000-01-12 -> 0.92
2000-01-13 -> 0.90
2000-01-14 -> 0.90
2000-01-15 -> 0.90
2000-01-16 -> 0.90
12.2.4.10 Search Rules
Overview
The fuzzy search in SAP HANA allows you to search for structured database content that is similar to the user
input. In this case, the user input and the records in the database are nearly the same but differ in their spelling
(for example, typing errors) or contain additional information (for example, additional or missing terms).
One of the use cases for the fuzzy search on structured data is the prevention of duplicate records. New database
records can be checked for similar and already existing records in real time, just before the new record is saved.
For example, before saving a new customer to the database, the application checks for similar customers that
may be duplicates of the new customer that has been entered. The application does some searches and then
presents to the user any existing customers that are similar to the user input. The user then decides whether to
create a new customer (because the records presented are similar, but not really duplicates) or to accept one of
the existing customers and continue with this customer record.
368
PUBLI C

Enabling Search
The searches performed by the application are defined by business rules that define when two customers are
similar. For example, the application may consider two customers to be similar as soon as one of the following
conditions is true:
1. The customers' names and addresses are similar.
2. The customers' last names and addresses are identical but the first names are different (may be persons
living in the same household).
3. The customers' names are similar and the dates of birth are identical.
These rules can be hardcoded in the application by writing three SELECT statements that perform the three
searches defined above. Whenever the requirements for the search rules change, the application code has to be
changed, tested, and deployed to the productive system. This may be costly in terms of time and the development
resources needed.
Alternatively, the application can use search rules to store the rules as a configuration object in the SAP HANA
database. Instead of embedding the rules in SELECT statements that are part of the application code, the
application has to call a database procedure to process all rules defined in the configuration object.
If the business requirements change, only the search rule definition in the configuration object has to be changed,
tested, and deployed. The interface to the database procedure that is called by the application remains
unchanged. So, without any change to the application code, the definition of the search rules is changed and the
user immediately gets search results according to the new requirements. This results in much less effort and
more flexibility when changing search rules.
Supported Database Objects
You can search on attribute views. The views have to be modeled using the SAP HANA studio and have to be
stored as objects in the SAP HANA repository.
Other database objects, such as row tables, column tables, calculation views, or analytic views, are not supported.
Important Terms
A search rule set is the configuration object that is stored in the SAP HANA repository and that contains the
definition of the search rules. When the database procedure is called to do a search, a search rule set is executed.
This means that all rules that are defined in the search rule set are executed.
A search rule defines a condition when two records the user input and a record in the database are
considered similar. Each rule in a search rule set is converted to a SELECT statement and is executed when the
search rule set is processed.
12.2.4.10.1 Creating Search Rule Sets
Before you can create a search rule set in the SAP HANA Studio, you have to create a workspace and a project in
the SAP HANA studio. In this workspace, you create a package that will contain your rule set.
Enabling Search
PUBLI C

To set up the project in the SAP HANA Studio, use the connection of the user who modeled the attribute view (for
example the user from the MODELOWNER tutorial) and add the package that you want to contain the search rule
sets to the project (for example the package from the apps.customer tutorial).
1. In the SAP HANA modeler, open the Project Explorer view and navigate to your package.
2. From the context menu of your package, choose New Other Search Rule Set .
3. Enter a file name for your rule set. The file has to have the extension .searchruleset.
4. Open and edit the search rule set in the search rule set editor. See Working with the Search Rule Set Editor
[page 370]
5. Define the attribute view, key columns, score selection parameter, stopwords, and term mappings. See
Configuring Search Rule Sets [page 371]
6. From the context menu of your package or search rule, choose Team Activate .
You can now execute a search with the rule set. See Executing a Search With a Rule Set [page 376]
Related Links
Search Rules - A Tutorial [page 381]
This tutorial documentation describes the development of an application that uses the search rule sets to search
on a table containing a company's customers.
12.2.4.10.2 XML Structure of a Search Rule Set
Search rule sets are stored in XML files in the local file system. The search rule set editor reflects the structure of
these files and displays the search rule set in a tree-like structure. To edit search rule sets, you need some basic
knowledge of the structure of the XML files. The XML structure of a search rule set configuration is as follows:
+Document Root
+Rule Set
+Attribute View
+Key Column 1
+...
+Key Column n
+Rule 1
+Column 1
+optional: Column Options (text, string or date options)
+...
+Column n
+optional: Column Options
+...
+Rule n
+Column 1
+...
+Column n
12.2.4.10.3 Working with the Search Rule Set Editor
To open a rule set in the editor, double-click a rule set file or, choose Open With Search Rule Set Editor in
the context menu.
370
PUBLI C

Enabling Search
Action Description
Adding new nodes, children or siblings
To add new nodes, children or siblings, you can use the
context menu of each node.
You can a new key column node in the following ways
for example:
Select the attribute view node and, choose New
Key Column in the context menu.
Select a key column node and choose New Key
Column in the context menu.
Deleting a node
To delete a node from the search rule set, select the
node and choose Delete in the context menu.
This deletes the node together with all its child nodes.
Changing the order of nodes
To change the order of nodes or to move nodes to
other parent nodes, you can drag and drop them.
With the SAP HANA studio running on Microsoft
Windows, you can copy nodes below the same parent
node or even to another parent node by pressing CTRL
while dragging.
Changing the node properties
To change properties, click on the value of the property
in the Properties view and enter the new value.
Each node contains a set of properties that define the
behavior of the search rules. Node properties are
displayed in the Properties view in the SAP HANA
studio when a node is selected in the tree. If the
Properties view is not displayed, you can open it by
choosing Window Show View Properties .
Note
Some node properties refer to database objects or
to column names. These properties are case-
sensitive, so be sure to enter all names in the correct
notation.
12.2.4.10.4 Configuring Search Rule Sets
1. Define the attribute view.
The search rule set searches on an attribute view that is modeled using the SAP HANA studio. The attribute
view to be used is defined in the attribute view node of the search rule set. The name property contains the
name of the attribute view. The FQN notation (fully qualified name) is used to specify the view.
Enabling Search
PUBLI C

You can type in the name of the view in the "Properties" view using the correct syntax for catalog and
repository objects.
After you defined the view and your project is shared with a HANA repository workspace, you don't need to
enter all the column names and the names of the key columns manually. In this case the "Property View" of
the corresponding node (column or key column) provides you a combobox with all availble fields.
Example attribute view name: apps.customer::CUSTOMER
2. Define the key columns and the score selection parameter.
A search may return the same record more than once because it matches more than one rule.
To enable the search to return each record only once, key columns must be defined in a way that makes
records in the result set unique. For an attribute view, there is no definition of key columns available, so the
key columns have to be defined for each search rule set.
The key columns are a set of one or more columns of the attribute view that clearly identify a record in the
view. So the key columns are similar to the primary key columns of a database table.
As for primary keys, LOB types (BLOB, CLOB, NCLOB, TEXT) are not allowed as key columns.
Note
It is possible to create an invalid key column definition that does not make the result set unique. In this
case, when running a search, an error is raised when records returned by a single rule are not unique.
By default, each search rule set contains one key column node below the attribute view node. If more columns
are needed to make records unique, more key column nodes can be added below the attribute view node.
In each key column node enter the name of the attribute view column in the properties panel.
In addition to the key columns, you have to define how the result set shall be made unique. Records returned
by more than one rule usually have different scores assigned. Only one combination of score and rule name
can be returned in the result set.
The score selection parameter defines whether the values with the highest score or the values found with the
first matching rule are returned.
You define the score selection parameter in the properties panel of the rule set node.
3. Optional: Define stopwords and term mappings.
To use stopwords and term mappings in a search rule set, the configuration options have to be added to the
rule set.
First, open the context menu of the Search Rule Set node and select New Stopwords (table-based) or New
Term Mappings (table based).
In the properties of the new nodes, you can define the stopword table and term mapping table that is used.
On the Stopwords (table-based) or Term Mappings (table-based) node, select New Column to enable the
stopwords or term mappings on a column. In the properties panel, you can define the name of the column
where stopwords and term mappings shall be applied.
Below the Column nodes, create new List ID nodes (open the context menu and select New List ID). In each
list ID node, you can configure a stopword or term mapping list that is applied on the column.
The stopword and term mapping settings are valid for a given column in all rules.
372
PUBLI C

Enabling Search
4. Add rules.
To add a new rule to the search rule set, open the context menu on the Search Rule Set node and select New
Rule.
You can change the order of the rules by dragging a rule with the mouse to a new position.
12.2.4.10.5 Configuration Settings and Properties
Search Rule Set Properties
The only property of a search rule set is the Score Selection. Score selection defines which rule number is
returned for a record if it was found by multiple rules. highestScore will choose the rule that gives the highest
score for the record (used for non-hierarchical views or together with Rule Cut) and firstRule will pick the
first rule (top down) that found this record (used for hierarchical rules).
Attribute View Properties
A search rule set needs an attribute view on which the rules are executed. To connect the search rule set with the
attribute view, the node Attribute View with the property Name is used. Name takes the full name consisting of
package and attribute view, separated by a double colon. _SYS_RT needs SELECT granted on this view.
Key Columns
When searching with multiple search rules, it is likely that the same row is returned by more than one rule. In this
case, the search rules runtime has to make a decision on what score, rule ID, and rule number are returned for this
record because EXECUTE_SEARCH_RULE_SET returns every row only once.
To make the results unique, the EXECUTE_SEARCH_RULE_SET function needs a definition of the 'key columns'
that identify each row. This is done by adding a Key Column node for each column the key consists of. Its
property Name takes the column's name. If the key consists of multiple columns, the same number of Key
Column nodes have to be defined. LOB types (CLOB, NCLOB, TEXT) are not allowed for key columns.
Stopwords and Term Mapping Properties
To use stopwords, the node Stopwords (table-based) has to be added. It takes Schema and Table names of
the stopword table as properties. _SYS_RT needs SELECT granted on this table. For each column that uses
stopwords, a Column node has to be added. Its property Name takes the column name for which stopwords are to
be used. Multiple List ID nodes can be defined under the Column node. Their property Id indicates a list ID of
the stopword table that is used for the given column.
The same is valid for term mappings with the node Term Mappings (table-based).
Enabling Search
PUBLI C

Rule Properties
For each rule that shall be defined, a node Rule has to be added. The order of the rules (top down) in the resource
set also defines the order of the execution. You name the rule by setting the Name property.
Min Fuzziness defines the minimum score that the rule's columns have to result in to trigger this rule. All
records with a lower score will not show in the result. The property Weight is a factor that lowers the rule's score
(for example, a score of 0.9 with a weight of 0.9 will result in 0.81). This is used to show that the rule's hits have a
lower value. Rule Cut is used to stop the processing before a rule is used. Whenever the total number of hits of
earlier rules exceeds the value of Rule Cut, no more rules are processed. A value of 0 (zero) deactivates Rule
Cut for this rule.
Rule Cut is used for the following reasons:
To separate rules (or groups of rules) that have a different hierarchy by putting a Rule Cut = 1 into the next
rule (or first rule of the next group). If you do this, the next rule (or next group) will not be executed if the
previous rule (or group of rules) results in hits.
To speed up processing. This is done by putting a Rule Cut with the desired number of hits into every rule.
As soon as this number is reached, no more rules are processed.
Column Properties
A rule needs at least one column to search on. The column's Name property points to a column of the Attribute
View defined above. Min Fuzziness corresponds to the fuzziness of the fuzzy() function within the contains()
statement. The property Weight is a factor that lowers the column's score (for example, a score of 0.9 with a
weight of 0.9 will result in 0.81). This is used to show that the column's hits have a lower value.
The column of the attribute view has a certain type in the table it is based on. The fuzzy search allows certain
column options for each type. These translate directly into column properties:
Column Type Column Option Node Available Properties
TEXT, SHORTTEXT, any STRING
column with fulltext index
Text Column Options
abbreviationSimilarity
andSymmetric
andThreshold
bestMatchingTokenWeight
considerNonMatchingToken
s
emptyScore
enableStopwords
enableTermMappings
minTextScore
similarCalculationMode
spellCheckFactor
Other STRING columns String Column Options
emptyScore
374
PUBLI C

Enabling Search
Column Type Column Option Node Available Properties
similarCalculationMode
spellCheckFactor
DATE, TIMESTAMP Date Column Options
emptyScore
maxDateDistance
The column option nodes are optional; if not specified, default values for their properties will be used.
TEXT columns (as well as SHORTTEXT and any STRING columns with fulltext index) always get the fuzzy search
option textsearch=compare set. This is to ensure that the fuzzy score will be retrieved instead of the TF/IDF
score.
The two properties enableStopwords and enableTermMappings are not standard fuzzy() parameters.
However, since stopwords and term mappings are defined for a column and are independent of the rules, these
parameters are needed to individually switch them off (false) for single rules. The default value of
enableStopwords and enableTermMappings is true.
Example Search Rule Set
Let us look at this example of a search rule set:
This rule set includes the following properties:
Enabling Search
PUBLI C

12.2.4.10.6 Executing a Search With a Rule Set
HANA exports a new built-in function: SYS.EXECUTE_SEARCH_RULE_SET, with which you can execute a
previously defined search rule set.
The function has one parameter, which is an XML string to run the search. Within this XML, you have to specify:
The ruleset that should be executed
The way the result will be returned
The limitation of the result
The input values
The EXECUTE_SEARCH_RULE_SET method returns by default a ResultSet object. This object contains all the
columns from the referenced attribute view plus additional columns _SCORE and _RULE_ID. Alternatively, the
EXECUTE_SEARCH_RULE_SET can write the result into a results table that needs to be created by the user.
376
PUBLI C

Enabling Search
For code samples of SYS.EXECUTE_SEARCH_RULE_SET calls, see Searching with a Rule Set - Code Examples
[page 379].
Transaction Isolation Level
The EXECUTE_SEARCH_RULE_SET function creates one SELECT statement for each of the rules and runs the
statements independent of each other. The statements are executed in the transaction context of the calling
application and use the same isolation level as the application. The isolation level has an influence on the results of
the EXECUTE_SEARCH_RULE_SET function when there are other transactions running in parallel that change the
contents of the database tables.
When the isolation level 'READ COMMITTED' is used, each of the SELECT statements of the search rule set sees
all changes that have been committed at the time the execution of the SELECT statement begins. So, for example,
the second rule of a rule set may see a new record that was not committed when the first rule has been executed.
In this case, the new record may be returned by the 'wrong' rule and the user gets an incorrect result.
When the isolation levels 'REPEATABLE READ' or 'SERIALIZABLE' are used, all SELECT statements see the same
state of the database. So the results returned by EXECUTE_SEARCH_RULE_SET are always correct.
Available XML Tags and Parameters
XML Tag Occurence Parameter Description
query 1 limit Defines the maximum number of rows that are
returned
1 offset Defines the number of rows skipped
Note
limit and offset work as the SQL options
limit and offset
ruleset 0..1 name Name of the ruleset that will be executed
resulttableschema 0..1 name Schema name of the resulttable
resulttablename 1..n name Table name of the resulttable
column 0..n name Input column name
resultsetcolumn name Defines the column that shall be returned. Columns
are returned in the order that is defined by these
tags. If no resultsetcolumn is defined, all columns
are returned that are defined in the attribute view
plus _SCORE and _RULE_ID.
Note
This tag is not valid if you use a resulttable for the
search result. In this case the resulttable defines
the structure of the result.
Predefined Columns
Enabling Search
PUBLI C

Column SQL Type Description
_SCORE The Fuzzy Score for each hit in the result. It can
be used in combination with a result table and the
resultset.
_RULE_ID The name of the rule that provided this hit. It can
be used in combination with a result table and the
result set.
Examples for Executing Rule Sets
Execute a search - Return the result as a ResultSet
CALL EXECUTE_SEARCH_RULE_SET('
<query>
<ruleset name="documentation.customersearch:Search.searchruleset" /> --
specifies the SearchRuleSet
<column name="FIRSTNAME">Herbert</column> -- specifies the input value for
column FIRSTNAME
<column name="LASTNAME">Hofmann</column> -- specifies the input value for
column LASTNAME
</query>
');
Execute a search - Write the result to a column table provided by the user
--First create the result table
set schema MY_SCHEMA;
CREATE COLUMN TABLE MY_RESULT_TABLE (
_SCORE FLOAT,
_RULE_ID VARCHAR(255),
"FIRSTNAME" TEXT FUZZY SEARCH INDEX ON FAST PREPROCESS ON,
"LASTNAME" TEXT FUZZY SEARCH INDEX ON FAST PREPROCESS ON
};
-- Afterwards you can execute the search using the created result table.
<query>
<resulttableschema name="MY_SCHEMA"/> -- specifies the schema of the result
table
<resulttablename name="MY_RESULT_TABLE"/> -- specifies the name of the result
table
column FIRSTNAME
378
PUBLI C

Enabling Search
column LASTNAME
</query>
');
-- get the result
select * from MY_RESULT_TABLE;
Limiting the number of rows returned by a search
When calling the system procedure EXECUTE_SEARCH_RULE_SET, the application can define the maximum
number of rows that are returned by setting a limit parameter.
By default, this parameter is undefined, which means that an unlimited number of rows is returned. The limitation
takes place after each rule and at the end when all rules are performed. In the following example, a maximum
number of 100 rows will be returned.
-- run the search
<query limit="100">
<ruleset name="documentation.customersearch:Search.searchruleset" />
<column name="FIRSTNAME">billy</column>
<column name="LASTNAME">smith</column>
</query>
');
12.2.4.10.7 Searching with a Rule Set - Code Examples
This topic contains code examples for search rule sets.
Executing a search and returning the result as a ResultSet
CALL SYS.EXECUTE_SEARCH_RULE_SET('
<query>
column FIRSTNAME
column LASTNAME
</query>
');
Enabling Search
PUBLI C

Executing a search and writing the result to a column table provided by the
user
--First create the result table
set schema MY_SCHEMA;
CREATE COLUMN TABLE MY_RESULT_TABLE (
_SCORE FLOAT,
_RULE_ID VARCHAR(255),
"FIRSTNAME" TEXT FUZZY SEARCH INDEX ON FAST PREPROCESS ON,
"LASTNAME" TEXT FUZZY SEARCH INDEX ON FAST PREPROCESS ON
};
-- Afterwards you can execute the search using the created result table.
<query>
<resulttableschema name="MY_SCHEMA"/> -- specifies the schema of the result
table
<resulttablename name="MY_RESULT_TABLE"/> -- specifies the name of the result
table
column FIRSTNAME
column LASTNAME
</query>
');
-- get the result
select * from MY_RESULT_TABLE;
Limiting the number of rows returned by a search
Note
When calling the system procedure EXECUTE_SEARCH_RULE_SET, the application can define the maximum
number of rows that are returned by setting a limit parameter.
In the default setting, this parameter is undefined. This means that an unlimited number of rows is returned.
The limitation takes place after each rule and at the end when all rules are performed. In the following example,
a maximum number of 100 rows will be returned.
-- run the search
<query limit="10" offset="100">
<ruleset name="documentation.customersearch:Search.searchruleset" />
<column name="FIRSTNAME">billy</column>
<column name="LASTNAME">smith</column>
</query>
');
You can use this parameter with the ResultSet object and with the custom result table.
380
PUBLI C

Enabling Search
12.2.4.10.8 Search Rules - A Tutorial
on a table containing a company's customers.
Prerequisites
As a developer who wants to create search rule sets, you are familar with the SAP HANA development
environment.
The development tools for search rule sets are part of the development environment for native SAP HANA
applications. Nevertheless, search rule sets can be used in all types of SAP HANA applications.
The Scenario
on a table containing a company's customers. The data model used for this tutorial is very simple. Let us assume
that all customers are persons, so no company information can be stored in the table.
Users and Roles
When building an application based on SAP HANA, different database users are created for modeling content and
for running the application. Privileges have to be granted to give each user a minimum set of access rights needed
to perform tasks corresponding to his/her role.
The scenario in this documentation uses the following database users:
The data stored in the column tables is owned by the first user called TABLEOWNER.
Attribute views and search rules are modeled by a second user called MODELOWNER.
Finally, the application runs with a third user called APPOWNER.
Instead of granting privileges to users directly, you can use roles.
Creating the Application Package
In our example, the database content (the attribute view and the search rule set) will be stored in a package called
"apps.customer".
In the SAP HANA studio, create a new package named customer.
You can create the package in the SAP HANA Systems view of the SAP HANA studio (either in the Modeler
perspective or in the SAP HANA Development perspective).
Enabling Search
PUBLI C

Use the SYSTEM user or any other user with sufficient privileges.
Creating the Users
You have to create three users and and assign a set of privileges to them. The users are created by the
administration user SYSTEM or by any other user with sufficient privileges.
1. Creating the TABLEOWNER
CREATE USER tableowner PASSWORD "********";
-- privilege needed to create the application schema
GRANT CREATE SCHEMA TO tableowner;
2. Creating the MODELOWNER
CREATE USER modelowner PASSWORD "********";
-- Allow repository access from SAP HANA studio
GRANT EXECUTE ON _sys_repo.repository_rest TO modelowner;
-- Privileges needed to access packages in the HANA repository
-- read access to packages and designtime objects (native and imported objects)
GRANT REPO.READ ON "apps.customer" TO modelowner;
-- all kinds of inactive changes to design-time objects in native packages
GRANT REPO.EDIT_NATIVE_OBJECTS ON "apps.customer" TO modelowner;
-- activate / reactivate design-time objects in native packages
GRANT REPO.ACTIVATE_NATIVE_OBJECTS ON "apps.customer" TO modelowner;
-- create, update or delete native packages, or create subpackages of native
packages
GRANT REPO.MAINTAIN_NATIVE_PACKAGES ON "apps.customer" TO modelowner;
-- permission go grant and revoke privileges on activated content (like, for
example, attribute views)
382
PUBLI C

Enabling Search
GRANT EXECUTE ON GRANT_PRIVILEGE_ON_ACTIVATED_CONTENT TO modelowner;
GRANT EXECUTE ON REVOKE_PRIVILEGE_ON_ACTIVATED_CONTENT TO modelowner;
-- A user needs an analytical privilege to access activated attribute views.
--
-- The predefined privilege _SYS_BI_CP_ALL gives access to all activated content.
-- Create your own analytical privilege if a limited access to activated content
is needed.
CALL _SYS_REPO.GRANT_ACTIVATED_ANALYTICAL_PRIVILEGE('_SYS_BI_CP_ALL',
'MODELOWNER');
-- privileges needed to run the search rules
GRANT EXECUTE ON sys.execute_search_rule_set TO modelowner;
Note
Additional privileges for SAP HANA 1.0 SPS5:
For SAP HANA 1.0 SPS5, an additional privilege is needed to execute search rule sets.
GRANT SELECT ON _sys_rt.search_rule_sets TO modelowner;
Note
_SYS_RT.SEARCH_RULE_SETS table
The _SYS_RT.SEARCH_RULE_SETS must not be queried in applications using search rule sets.
If you do not want to assign a large number of privileges to each modeling user, you can create a new role
containing all necessary privileges and assign this role to the users. This is the preferred option, but
alternatively you can use the predefined roles MODELING or CONTENT_ADMIN to grant the privileges.
3. Creating the APPOWNER
CREATE USER appowner PASSWORD "********";
-- A user needs an analytical privilege to access activated attribute views.
CALL _SYS_REPO.GRANT_ACTIVATED_ANALYTICAL_PRIVILEGE('_SYS_BI_CP_ALL',
'APPOWNER');
-- privileges needed to run the search rules
GRANT EXECUTE ON sys.execute_search_rule_set TO appowner;
Note
Additional privileges for SAP HANA 1.0 SPS5:
For SAP HANA 1.0 SPS5, an additional privilege is needed to execute search rule sets.
GRANT SELECT ON _sys_rt.search_rule_sets TO appowner;
Note
_SYS_RT.SEARCH_RULE_SETS table
The _SYS_RT.SEARCH_RULE_SETS must not be queried in applications using search rule sets.
Enabling Search
PUBLI C

Creating the Database Schema and Tables
You have to create a new database schema called 'CUSTOMERS' to store the database tables.
Then, you create the table 'CUSTOMER' in the new schema to store information about customers such as name,
address, phone number, and date of birth. The example uses a simple table definition only, but more complex
table definitions including 1:n relationships between names and addresses are possible.
The schema and table are created by the user TABLEOWNER.
Creating the Database Schema and Tables
CREATE SCHEMA customers;
SET SCHEMA customers;
-- create sample table
CREATE COLUMN TABLE customer
(
firstname SHORTTEXT(100) FUZZY SEARCH INDEX ON,
lastname SHORTTEXT(100) FUZZY SEARCH INDEX ON,
streetname NVARCHAR(100) FUZZY SEARCH INDEX ON,
housenumber NVARCHAR(20) FUZZY SEARCH MODE
'housenumber',
postcode NVARCHAR(20) FUZZY SEARCH INDEX ON FUZZY SEARCH MODE
'postcode',
cityname NVARCHAR(100) FUZZY SEARCH INDEX ON,
countrycode NVARCHAR(2),
phone NVARCHAR(20),
dateofbirth DATE
);
-- needed to model an attribute view on top of the table
GRANT SELECT ON customer TO modelowner;
-- needed to allow activation of attribute views that use this table
GRANT SELECT ON customer TO _sys_repo WITH GRANT OPTION;
-- for better performance, database indexes should be created
-- on all NVARCHAR columns that are used in the search rules
-- example:
CREATE INDEX customer_cityname ON customer(cityname);
-- insert a sample record:
INSERT INTO customer VALUES(1, 'Billy', 'Smith', 'Summerset Drv', '1001',
'123456789', 'Littleton', 'US', '555-1234', '1950-12-01');
-- to be able to use stopwords a stopword table is needed:
CREATE COLUMN TABLE stopwords
(
language_code VARCHAR(2),
);
GRANT SELECT ON stopwords TO _sys_repo WITH GRANT OPTION;
INSERT INTO stopwords VALUES('1', 'firstname', '', 'Dr');
-- and for term mappings another table:
(
384
PUBLI C

Enabling Search
language_code VARCHAR(2),
term_1 NVARCHAR(255) NOT NULL,
term_2 NVARCHAR(255) NOT NULL,
weight DECIMAL NOT NULL
);
GRANT SELECT ON termmappings TO _sys_repo WITH GRANT OPTION;
INSERT INTO termmappings VALUES('1', 'firstname', '', 'William', 'Will',
'1.0');
INSERT INTO termmappings VALUES('2', 'firstname', '', 'William', 'Bill',
'0.9');
INSERT INTO termmappings VALUES('3', 'firstname', '', 'William', 'Billy',
'0.9');
INSERT INTO termmappings VALUES('7', 'firstname', '', 'Will', 'William',
'1.0');
INSERT INTO termmappings VALUES('8', 'firstname', '', 'Will', 'Bill',
'0.9');
INSERT INTO termmappings VALUES('9', 'firstname', '', 'Will', 'Billy',
'0.9');
INSERT INTO termmappings VALUES('4', 'firstname', '', 'Bill', 'William',
'0.9');
INSERT INTO termmappings VALUES('5', 'firstname', '', 'Bill', 'Will',
'0.9');
INSERT INTO termmappings VALUES('6', 'firstname', '', 'Bill', 'Billy',
'1.0');
Defining the Attribute View
As the user MODELOWNER, you create the attribute view that is used for the search.
The attribute view is created in the SAP HANA Systems view that is part of the SAP HANA studio (either in the
Modeler perspective or in the SAP HANA Development perspective).
1. In the package 'apps.customer', create a new attribute view called CUSTOMER (containing all columns from
table CUSTOMERS.CUSTOMER).
2. Save and activate the attribute view.
Enabling Search
PUBLI C

A SELECT privilege is required for a user to see the contents of the activated attribute view.
-- allow users MODELOWNER and APPOWNER to see the contents of the attribute view
CALL GRANT_PRIVILEGE_ON_ACTIVATED_CONTENT('SELECT', '"apps.customer::CUSTOMER"',
'MODELOWNER');
CALL GRANT_PRIVILEGE_ON_ACTIVATED_CONTENT('SELECT', '"apps.customer::CUSTOMER"',
'APPOWNER');
The contents of the view are now visible to users MODELOWNER and APPOWNER.
-- test the view
SELECT * FROM "apps.customer::CUSTOMER";
Creating the Search Rule Set
Read about the steps needed to define a search rule set and to run the final application.
1. Preparing the Development Environment
The SAP HANA studio is used to create, modify, and activate search rule sets.
Before creating your first search rule set, you have to set up a native SAP HANA Development project in the
SAP HANA studio:
2. Create a new repository workspace in the SAP HANA Repositories view.
Use the database connection of user MODELOWNER for the repository workspace.
3. Create a new project called 'project' in package 'apps.customer'.
Select 'General/Project' for a project that contains search rule sets only.
Select 'SAP HANA Development/XS Project' for native SAP HANA projects.
386
PUBLI C

Enabling Search
Use 'Team/Share Project' to move the project to the repository package 'apps.customer'.
2. Defining the Search Rule Set
The next step for user MODELOWNER is to define the search rule set that will be used by the application. This
is done in the Project Explorer view in the SAP HANA studio.
The steps to define the search rule set are as follows:
1. In your project, create a new search rule set 'Search.searchruleset'.
The search rule set is created in the package 'apps.customer.project'.
2. Edit the search rule set.
Define view "apps.customer::CUSTOMER" as the view that will be searched.
Define at least one rule (for example, use columns FIRSTNAME and LASTNAME in the rule).
3. Save the search rule set.
4. Commit and activate your changes.
Now you can call the search in the SAP HANA studio:
-- run the search
<query>
<ruleset name="apps.customer:Search.searchruleset" />
<column name="FIRSTNAME">Dr. bill</column>
<column name="LASTNAME">smiths</column>
<column name="CITYNAME">littleton</column>
<column name="DATEOFBIRTH">1950-12-02</column></query>
');
3. Running the Application
The application user (APPOWNER) can now build an application that uses the attribute view and the search
rules defined previously.
In the application, you can call the following statements:
-- show contents of attribute view
SELECT * FROM "apps.customer::CUSTOMER";
-- run the search
<query>
<ruleset name="apps.customer:Search.searchruleset" />
<column name="FIRSTNAME">Dr. bill</column>
<column name="LASTNAME">smiths</column>
<column name="CITYNAME">littleton</column>
<column name="DATEOFBIRTH">1950-12-02</column></query>
');
12.2.4.11 Support Information
12.2.4.11.1 Monitoring Views
This topic lists the monitoring views related to fuzzy search and fuzzy search indexes.
Enabling Search
PUBLI C

Checking the memory usage with M_HEAP_MEMORY
All data structures for fuzzy search share a common 'Pool/FuzzySearch' allocator. Statistics can be obtained
through the system view M_HEAP_MEMORY.
SELECT * FROM m_heap_memory WHERE category LIKE '%FuzzySearch%';
HOST "lu3412587"
PORT 30003
VOLUME_ID 2
STATISTICS_ID 36723
CATEGORY "Pool/FuzzySearch"
DEPTH 2
INCLUSIVE_SIZE_IN_USE 4989216
INCLUSIVE_COUNT_IN_USE 14399
INCLUSIVE_ALLOCATED_SIZE 15428600
INCLUSIVE_DEALLOCATED_SIZE 10439384
INCLUSIVE_ALLOCATED_COUNT 49177
INCLUSIVE_DEALLOCATED_COUNT 34778
INCLUSIVE_MAX_SINGLE_ALLOCATION_SIZE 524288
INCLUSIVE_PEAK_ALLOCATION_SIZE 4989216
EXCLUSIVE_SIZE_IN_USE 4989216
EXCLUSIVE_COUNT_IN_USE 14399
EXCLUSIVE_ALLOCATED_SIZE 15428600
EXCLUSIVE_DEALLOCATED_SIZE 10439384
EXCLUSIVE_ALLOCATED_COUNT 49177
EXCLUSIVE_DEALLOCATED_COUNT 34778
EXCLUSIVE_MAX_SINGLE_ALLOCATION_SIZE 524288
EXCLUSIVE_PEAK_ALLOCATION_SIZE 4989216
EXCLUSIVE_ALLOC_ERRORS 0
MALLOC_PROXY_CACHE_MISSES 0
FLAGS "(none)"
This information is also available in the Management Console.
hdbadm@lu3412587:/usr/sap/HDB/HDB00 $ hdbcons
SAP HANA DB Management Client Console (type '\?' to get help for client commands)
Try to open connection to server process 'hdbindexserver' on system 'HDB', instance
'00'
388
PUBLI C

Enabling Search
SAP HANA DB Management Server Console (type 'help' to get help for server commands)
Executable: hdbindexserver (PID: 2644)
[OK]
--
> mm list Pool/FuzzySearch -s
+----------------+-----+-----------------+-----+-----------------+------+-------+
|Name |Used |Local size |TUsed|Total size |ACount|TACount|
+----------------+-----+--------+--------+-----+--------+--------+------+-------+
|Pool/FuzzySearch|14399|4989216B|4872.2KB|14399|4989216B|4872.2KB| 49177| 49177|
+----------------+-----+--------+--------+-----+--------+--------+------+-------+
[OK]
--
>
12.2.4.11.2 Sizing Information
This topic contains information about the sizing of fuzzy search indexes.
When setting FUZZY SEARCH INDEX ON, index structures are created in memory to make the fuzzy search
faster. It is important to be aware of the additional memory usage when sizing an SAP HANA server. It is not
possible to give exact numbers here, since the size of a fuzzy search index depends on the contents of the column
and on the compression mode used for this column. A fuzzy search index on a VARCHAR column for example is
large if each row contains a distinct value. The index is much smaller if the number of distinct values in this column
is small. The size of the fuzzy search index can be determined with the M_FUZZY_SEARCH_INDEXES monitoring
view.
More information: Monitoring Views [page 387].
Fuzzy Search Indexes for String Types
When creating a fuzzy search index on a string-type column (VARCHAR, NVARCHAR), the size of the fuzzy search
index can be up to twice the memory size of the column itself.
Fuzzy Search Indexes for Text Types
The fuzzy search index on a text-type column (SHORTTEXT, TEXT or a hidden column created with the CREATE
FULLTEXT INDEX statement) needs about 10 percent of the memory size of the column. Labels parameters
12.2.4.11.3 Activating the Trace in the SAP HANA Studio
Useful tracing and logging information for a fuzzy search is stored in the database traces. The traces are written
for several services of the system (for example, INDEXSERVER and NAMESERVER). If the database trace is
configured, trace information for the specified database will be written to files named as follows:
<servicename>_<hostname>.<portnumber>.000.trc.
Enabling Search
PUBLI C

Tracing the output of a fuzzy search is available for the components fuzzysearch, searchrulesexecution,
and searchruleruntime. The trace of the component fuzzysearch shows information about selected search
options and the scoring and number of search results. You should use this trace for single fuzzy search requests.
If the trace for searchrulesexecution and/or searchruleruntime is configured, information about the
search rule configuration and search rule set results are tracked. The components searchrulesexecution
and/or searchruleruntime should be used if a search rule set that combines multiple fuzzy search requests is
used.
1. In the SAP HANA studio, open the Administration perspective by double-clicking your system name in the
navigation pane.
2. Select the tab Trace Configuration and choose the icon Edit Configuration at the right corner of the tracing
section you want to configure.
3. Set the trace levels in the window Database Trace.
390
PUBLI C

Enabling Search
Trace Level Description
NONE Tracing is switched off. However, information about
error situations is still recorded.
ERROR and FATAL Information about errors is recorded.
WARNING Information about potentially problematic situations
is recorded.
INFO Information about activity in the database is
recorded.
DEBUG Detailed information intended for debugging is
recorded.
4. To configure the trace targets, choose Show All Components and filter for fuzzysearch or searchrules, as
shown below.
To get information about a fuzzy search, you only have to configure the INDEXSERVER. All other servers do
not output relevant information regarding a fuzzy search.
Enabling Search
PUBLI C

To get basic tracing data with information for every search, set the trace level to INFO.
To get detailed information for every search, set the trace level to DEBUG.
Note
DEBUG information cannot be read by end users and should be used for support issues only.
The default name for the trace file is indexserver_$host_$port.000.trc.
After executing a fuzzy search statement, you can find all trace files in the section Diagnosis Files of your
administration perspective. To get only the information of the last request, you should delete older trace files
before executing a fuzzy search. By clicking the column header of the traces table, you can reorder the trace
files view, for example, by the date of the last modification.
392
PUBLI C

Enabling Search
To open a selected trace file, double-click it.
If you have problems with a fuzzy search, you can download trace files of your search and send them to SAP
customer support. You can download an opened trace file by using the function Download File it the left
corner of the trace file window.
12.2.4.12 Frequently Asked Questions
Why are there results with a score lower than the requested fuzzySimilarity?
In text fields, the parameter fuzzySimilarity sets the minimum similarity that a token has to match to be included
in the search result. All other fuzzy search operations (for example, applying term mappings, considering
stopwords, abbreviationSimilarity) can influence the score that you will see.
Enabling Search
PUBLI C

How many misspellings are allowed with a particular fuzzySimilarity?
This question is not easy to answer. The scoring algorithm is not linear to the number of misspellings; the position
of the misspelling is also important. You can use the following example to familiarize yourself with it:
DROP TABLE test;
CREATE COLUMN TABLE test
(
);
INSERT INTO test VALUES ('1','abc');
INSERT INTO test VALUES ('2','abx');
INSERT INTO test VALUES ('3','xbc');
INSERT INTO test VALUES ('4','axc');
INSERT INTO test VALUES ('5','abcx');
INSERT INTO test VALUES ('6','xabc');
SELECT TO_DECIMAL(SCORE(),3,2) AS score, * FROM test
WHERE CONTAINS(companyname, 'abc',
1 1 abc
0.89 5 abcx
0.82 2 abx
0.75 6 xabc
0.61 3 xbc
0.61 4 axc
SELECT TO_DECIMAL(SCORE(),3,2) AS score, * FROM test
WHERE CONTAINS(companyname, 'abcx',
1 5 abcx
0.89 1 abc
0.88 6 xabc
0.75 3 abx
0.59 3 xbc
0.59 4 axc
How do I find out if the fuzzy search index is enabled for column x?
See Using the CONTAINS() Predicate [page 316].
394
PUBLI C

Enabling Search
How do I enable the fuzzy search index for a particular column?
See Using the CONTAINS() Predicate [page 316].
The additional data structures will increase the total memory footprint of the loaded table. In unfavorable cases
the memory footprint of the column can double.
How can I see how much memory is used for a fuzzy search index?
See Monitoring Views [page 387].
Is the score between request and result always stable for TEXT columns?
It depends on how you look at the topic. The algorithm is indeed deterministic, but you need to take all parameters
into account. Cases can be constructed where a small change in the fuzzySimilarity will change the rank between
the same strings.
Why is this? The fuzzySimilarity is the minimum score that tokens need to reach to be considered for the result. If
you use andThreshold or the keyword "OR" in your search, not all tokens have to reach the fuzzySimilarity to be
part of the result. This can lead to a change in the total score if you change the fuzzySimilarity. Let us look at an
example:
DROP TABLE companies;
(
companyname SHORTTEXT(200)
);
INSERT INTO companies VALUES(1, 'aktien gesellschaft');
INSERT INTO companies VALUES(2, 'aktiv gesellschaft');
Important: The similarity between "aktien" and "aktiv" is 0.77.
If the fuzzySimilarity is lower than 0.77, the token scoring will be part of the result score. If the fuzzySimilarity is
higher than 0.77, the token scoring will not be considered, so the total scoring will be lower.
SELECT TO_DECIMAL(SCORE(),3,2) AS score, id, companyname
FROM companies
WHERE CONTAINS(companyname, 'aktiv OR gesellschaft', FUZZY(0.75,
'textSearch=compare'))
1 2 aktiv gesellschaft
0.89 1 aktien gesellschaft
SELECT TO_DECIMAL(SCORE(),3,2) AS score, id, companyname
FROM companies
WHERE CONTAINS(companyname, 'aktiv OR gesellschaft', FUZZY(0.80,
Enabling Search
PUBLI C

'textSearch=compare'))
1 2 aktiv gesellschaft
0.71 1 aktien gesellschaft
12.3 Building Search Apps
12.3.1 Introduction to the UI Toolkit for Info Access
The UI toolkit for SAP HANA Info Access provides UI building blocks for developing browser-based search apps on
SAP HANA. Such applications provide real-time information access and faceted search features for huge volumes
of structured and unstructured text data.
The toolkit enables a freestyle search of a SAP HANA attribute view, displaying and analyzing the result set. The
toolkit provides UI elements (widgets) such as a search box, a result list with a detailed view, and charts for basic
analytics on the result set. The widgets are interconnected and adapt in real-time to user entries and mouse-over
(hover) selections.
The toolkit is based on HTML5 and JavaScript libraries such as JQuery/JQueryUI, d3 (Data Driven Documents),
and Tempo. The widgets use the SAP HANA Info Access HTTP service. You do not need an additional layer to run
the UI; SAP HANA and a Web browser are sufficient. The toolkit is tested on Mozilla Firefox 17. It also runs on
Microsoft Internet Explorer 9 in standard mode and Google Chrome.
Along with the actual toolkit (the widgets), we deliver a fully operational demo Web app built using this toolkit. The
demo app enables you to explore a search application dealing with a set of sample data for electronic products.
You can use the demo app as a template or for reference when you start developing your own app.
You can, for example, use the demo site to easily connect your own data and compose your Web site. To create a
search application this way, it is sufficient to have experience in HTML and to have read this documentation. For
advanced customizing and layout of the site, more in-depth knowledge of CSS and the above-mentioned
JavaScript libraries is required.
12.3.1.1 What is the SAP HANA Info Access HTTP Service?
In the first place, SAP HANA is a database. It stores any type of raw data in its tables: master data, transactional
data, numbers, texts, and so on. But SAP HANA can do more than a classic database. SAP HANA also enables you
to turn raw data into meaningful information.
For the world of numbers, there are the analytics features of SAP HANA. By joining, aggregating, and calculating,
they can turn raw facts and numbers into meaningful measures and key figures.
For the world of texts, there are the full text search and text analysis features of SAP HANA. By tokenizing,
stemming, normalizing, and analyzing semantically, they can turn a set of strings into a structured and searchable
text corpus.
396
PUBLI C

Enabling Search
So there is meaningful information in SAP HANA. But it is still not instantly available to business end users. SQL is
a very powerful tool for accessing and processing SAP HANA information. However, it operates on a very
technical level deep down in the stack and far from an end user UI.
This is where the info access service of SAP HANA steps in. As long as you only need read-access, meaning
search or basic analytics, the service provides shortcuts to Web and mobile UIs.
12.3.1.2 SAP HANA Info Access Architecture
The SAP HANA info access HTTP service wraps search and analytic SQL queries and exposes them through an
HTTP interface. On the UI layer, Info Access offers the HTML5 development kit including UI widgets and Web site
templates. No intermediate layer is required. To provide Info Access apps to your users, you only need SAP HANA
and a Web browser.
12.3.2 Installing the Service and the Toolkit
The UI toolkit and the service are part of the default SAP HANA shipment, but they are not installed automatically.
They are shipped as separate delivery units that you can import and activate quickly as described below.
12.3.2.1 Importing the Delivery Units
SAP HANA Appliance Software, support package stack (SPS) 05 or higher, running on a server.
SAP HANA studio and SAP HANA client installed on your local machine.
1. To import the UI toolkit and the service, in the SAP HANA studio, choose the Modeler perspective and select
your database instance.
2. From the menu, choose File Import .
Enabling Search
PUBLI C

3. Select HANA Content Delivery Unit and choose Next.
4. Select Client and browse for the HCO_INA_UITOOLKIT.tgz delivery unit.
The delivery units can be found on the SAP HANA host under SYS/global/hdb/content.
5. Select both actions and choose Finish.
6. Perform the same import procedure for the HCO_INA_SERVICE.tgz delivery unit.
7. In the Navigator view, under Content, check that the following packages are available:
sap\bc\ina\uitoolkit
sap\bc\ina\demos\epm
sap\bc\ina\service\v2
8. To get started developing apps using the toolkit, set up an application project.
Related Links
12.3.2.2 Registering and Activating the Info Access Service
1. To register the service, switch to the Lifecycle Management perspective of the SAP HANA studio and select
your database instance.
2. Choose the Configuration tab page and expand xsengine.ini\application_container.
3. Double-click the application_list line and, under the System column, add the entry
InformationAccess.
Caution
Once InformationAccess is running, SAP HANA data is exposed by an HTTP service (port 80<instance>).
4. To restart SAP HANA XS, choose the Landscape tab page.
5. From the context menu of the xsengine service, choose Kill...
The service stops completely and restarts automatically.
12.3.2.3 Importing the Demo Data
1. Go to the SAP Code Exchange Web page of the UI toolkit.
2. From the Documents tab page, download the epm_data.zip archive.
3. Follow the instructions in the readme.txt file contained in the archive.
Use the system user for importing and running the demo.
Related Links
398
PUBLI C

Enabling Search
SAP Code Exchange Web page of the UI toolkit
12.3.3 Getting to Know the Demo App
You can now explore what an app built using the UI toolkit looks and feels like. Start the sample UI in a Web
browser by opening the following URL:
http://<HANAhost>:80<instance>/sap/bc/ina/demos/epm/search.html
Example: http://hana1.acme.corp:8001/sap/bc/ina/demos/epm/search.html
Start exploring the UI.
The demo app consists of a header row, a column on the left, and a large content area on the right. The column on
the left contains some widgets; the search box and some charts. The charts are used for displaying the count of
result items with distinct values. In the EPM example, it is the count of sold products by category and by currency.
The content area on the right is for displaying an enlarged view of one of the widgets, the result list is enlarged at
page load.
The app starts with a search for all. By typing in the search box or filtering in the charts, you can refine the result
set. The results are displayed in the list ranked by relevancy.
From the right margin of the page, you can pull in the facet repository containing all configured facets that are not
displayed in the column and the content pane. Facets are the widgets, such as the results list and charts, that
display the dimensions of the result set.
Enabling Search
PUBLI C

You can drag and drop all facets to any position on the UI, be it the column, the content area or the facet
repository. When a chart is enlarged in the content pane, you can change the chart type with a click.
12.3.4 Getting to Know the Demo HTML
The structure and content of the UI are defined in the search.html file. Open the search.html file using the
HTML editor. Familiarize yourself with the structure of the HTML.
Partly collapsed, search.html looks like this:
In the head area and at the end of the HTML, there is meta data and references to the various JavaScript libraries
and CSS used by the UI.
The body of the HTML contains the UI's header bar with the title text and the alert bar for displaying messages.
The rest of the body is the workarea. A workarea connects to a SAP HANA attribute view and serves as a
container for all other widgets on the page. The workarea is divided into the facetPane, the contentPane, and
the facetrepository.
The facetPane starts with the searchbox followed by a facetcolumn and the facetpane. The workareaitem
widgets inside the column and the pane are only placeholders for facets of the repository that the user can shuffle
around at runtime.
Inside the facetrepository, there are the actual workareaitem widgets. For switching between different
chart types, there are the switchbox widgets. The last widget in the repository is the resultlist.
400
PUBLI C

Enabling Search
12.3.5 Preparing Your Source Data
Before you start to build your information access application, you need to have a clear understanding of who will
use it and for what purpose. Once you are sure which content you need to enable for access, prepare this data as
described here. Then you define how the content is to be displayed and which interactive navigation attributes are
to be provided.
12.3.5.1 Ensuring Full Text Indexing
Make sure that a full-text index is created for each of your table columns that contains human-readable text data
that is suitable for a freestyle search. The index structure is attached to the column and is leveraged to identify
texts that contain the search terms. The full text index is automatically updated when new records are entered in
the table.
Along with the full text index, you can also trigger a text analysis that extracts entities, such as dates or people,
from the text and, therefore, enriches the set of attributes.
Related Links
indexes.
Text Analysis [page 298]
12.3.5.2 Checking the Settings in SAP HANA Studio
Technical search in SAP HANA may not be enabled. To check if it is enabled, proceed as follows:
1. In the SAP HANA studio menu, choose Window Preferences Modeler Search Options .
2. Check if the Enable Search Attributes option is selected. If not, select it and choose Apply and OK.
Note
You must make this setting in every SAP HANA studio that you use to model views for search.
12.3.5.3 Modeling Your Search Content
For each app user, you have created a named database user with the following authorizations:
Object privileges: SELECT on database schemas<_SYS_RT> and <_SYS_BI>.
Analytic privileges: For <_SYS_BI_CP_ALL> and for the attribute view you want to visualize in the app.
In the SAP HANA modeler, create an attribute view of type standard using the tables that you want to enable for
the search. Create joins and add the attributes you want to use for searching and displaying.
Enabling Search
PUBLI C

Note
You can also join additional attributes derived from the text analysis.
To enable certain attributes for the search, proceed as follows:
1. In the Output view, select the attributes.
Note
The searchable attributes must not be calculated attributes nor have lower case letters in their names.
2. In the lower area of the Properties tab, select the Information Access category.
3. Set the Freestyle Search property to true.
In the Information Access category, you can make additional search-specific settings.
The Weights for Ranking setting (between 0.0 and 1.0, default 0.5) influences the ranking of items in the results
list. The higher the weight of the attribute, the higher up in the list an item with a hit in this attribute is positioned.
The Fuzziness Threshold setting (between 0.0 and 1.0, default 0.8) defines the grade of error tolerance for a
search on this attribute. The higher the threshold, the more exact the search terms must hit the text to produce a
result.
Note
Do not forget to activate the attribute view.
Related Links
Analytic Views [page 126]
FUZZY Search [page 314]
Fuzzy Search is a fast and fault-tolerant search feature for SAP HANA. A fuzzy search returns records even if the
search term contains additional or missing characters or other types of spelling errors.
12.3.6 Connecting Your Source Data
You can connect any search-enabled SAP HANA attribute view to the app by referencing the view and its
attributes directly in your HTML file.
To reference the attribute view and its schema in the HTML, enter their names in the workarea widget:
<div data-sap-widget="workarea" data-title="searchandanalytics"
data-entityset="J_EPM_PRODUCT" data-packagename="epm"
data-aggregationdimension="PRODUCT_ID">
data-entityset is the view name and data-packagename the package this view resides in. The data-
aggregationdimension is used as basis for counting items with distinct values. Enter an attribute that has
unique values, for example, the primary key.
402
PUBLI C

Enabling Search
For chart facets that display the aggregates of attribute values, for example the grouped bar chart, enter the
attribute names here:
<div data-sap-widget="chart" data-title="PRICE-GROUPBAR"
data-dimension-line="CURRENCY_CODE" data-dimension-x="CATEGORY"
data-dimension-y="COUNT" data-charttype="groupbar">
Note
You cannot use attributes of data type TEXT_AE for charts.
In the grouped bar chart example, data-dimension-line is the attribute whose distinct value count is to be
displayed as the bars. The data-dimension-x attribute is displayed as an additional dimension inside the chart
as a group of bars. data-dimension-y is the measurement that is displayed on the y-axis defining the height of
the single bars. In the search app, the only measurement available is the distinct values count ($count) of the
attribute defined in data-dimension-line.
In a simple bar or pie chart, the data-dimension parameter contains the attribute whose distinct value count is
to be displayed
To define the attributes to be displayed in a result list entry, enter the attribute names here:
<div data-sap-widget="inaresultlist" data-maxresultslarge="6"
data-maxresultssmall="3" data-responseattributes="PRODUCT_ID,
WEIGHT_MEASURE,WEIGHT_UNIT,PRICE,CURRENCY_CODE,DEPTH,DIM_UNIT,CATEGORY,
CHANGED_AT,WIDTH,HEIGHT,THUMBNAIL"data-detailwidth="650"
Do not leave the data-responseattributes parameter empty. Only use attributes from the main table of the
view or attributes that are joined in 1:1 from other tables. Which response attributes are actually displayed in a
result entry depends on the result layout templates. There are CSS and HTML templates available for the small
result list, the large list, and the details view.
Related Links
workarea [page 404]
chart [page 406]
Defining the Layout of Result Lists and Details [page 408]
12.3.7 Defining the Page Layout
Now that your source data is prepared and connected, you can form your UI using the search.html template.
You may want to start by arranging the available widgets on your site. To place the column, the panes, and the
facets within them, copy, move, or remove the corresponding HTML blocks.
First, place the facet pane and its column(s) as well as the content pane. The panes and the columns inside are
displayed from left to right on the UI in the same order as in the HTML. Only one content pane is allowed.
Typically, a maximum of three facet columns are used. The facet repository is always last in the HTML.
The search box and the facets inside the facet columns are next. The search box can span the entire pane across
multiple columns. Place as many placeholder work area items as you want to have slots in a column. Inside the
content pane, only one placeholder is allowed. Make sure that the data-target-position count is continuously
set across the facet columns and the content pane.
Enabling Search
PUBLI C

The facet repository holds the actual facets, meaning the chart switch boxes and the result list. The data-
source-position parameter defines their position inside the facet repository as well as inside the columns and
the content pane at page load.
If the source position equals the target position, the respective facet replaces the placeholder in the columns and
pane. If there are two or more facets with the same source position number, only the first one in the HTML
replaces the placeholder. In the facet repository, all defined facets are available in the order of the source position
or their appearance in the HTML. At runtime, the user can drag and drop the facets.
At least one facet pane and content pane are required for the facet repository and free drag-and-drop. However,
you could also build a minimal static UI with a workarea containing just a search box and a result list, for example.
You can even integrate single widgets into other sites.
Related Links
workareaitem [page 405]
12.3.8 Configuring the Widgets
The widgets are based on jQuery UI. In the HTML, you can customize the single widgets to a certain degree using
parameters. The common parameter data-sap-widget defines the type of the widget.
Each widget type has its own set of parameters. The widgets and their parameters are listed in the order of
appearance in the demo HTML file.
12.3.8.1 header
Parameter Name Default Value Description
data-title none Title text in the header bar of the UI.
data-helphref none Reference to the documentation for the users of your application.
12.3.8.2 workarea
data-title none This title is not displayed on the UI.
data-
packagename
none Name of the database package in which the attribute view was created.
data-entityset none Name of the attribute view.
data-schema _SYS_BIC Name of the database schema (catalog) in which the activated attribute
view resides. The schema _SYS_BIC is default because your activated
attribute views usually end up in this schema.
404
PUBLI C

Enabling Search
data-
aggregationdime
nsion
none Primary key attribute of the items you want to search and display. The
unique key is used to identify items for the result list and to count items
with distinct values for the charts.
data-
aggregationsort
order
DESC Descending or ascending (ASC) order for sorting the counts in the charts.
This setting is inherited by all charts if you do not set a sort order there.
Related Links
Connecting Your Source Data [page 402]
12.3.8.3 searchbox
data-
enablesuggestions
false The system suggests possible search terms as the user types.
data-maxsuggestions 10 Maximum number of suggested terms to be displayed in the
drop-down list of the search box.
data-enablelivesearch false The system already performs a search with each letter a user
types.
data-maxinputlength 255 Maximum number of characters that users can enter in the
search box.
12.3.8.4 workareaitem
data-title none This title is displayed on the UI as the facet headline. If the
workareaitem is a placeholder, the title is used as
alternative text. The alternative is displayed if the
placeholder is not replaced by an actual workareaitem at
runtime.
data-target-position 0 Position of a placeholder workareaitem on the UI in the
facet or content pane. The count starts on the upper left of
the UI.
data-source-position 0 Position of the actual workareaitem on the UI in the facet
repository on the right. On page load, the workareaitem
placeholders are replaced by the actual workarea items
whose source positions match the target positions.
Enabling Search
PUBLI C

12.3.8.5 facetrepository
data-children-drag-
handle
body Defines the area where a user can grab a facet for drag and drop.
body means the complete facet, header means the facet's
header bar only.
12.3.8.6 switchbox
data-activeindex 0 Defines which chart contained in the switchbox is displayed on page
load. 0 means the first one, 1 the second, and so on.
12.3.8.7 chart
data-toplarge none Defines how many bars or pie segments are displayed when the
chart is enlarged.
data-dimension none Attribute whose distinct value count is to be displayed in a
simple bar or pie chart.
Note
You cannot use attributes of data type TEXT_AE for charts.
data-charttype none Type of the chart (bar, pie, groupbar, or line).
data-color none Color of the chart as hex triplet, for example #FFFFFF for white,
or as standard HTML names, for example blue or black.
data-dimension-line none Attribute whose distinct value count is to be displayed in
grouped bar or line chart.
data-dimension-x none Attribute that is displayed as additional dimension inside the
chart as group of bars or different lines.
data-dimension-y none Measurement that is displayed on the y-axis. In a search
scenario, the only measurement available is the distinct value
count ($count) of the attribute defined in data-dimension-
line.
406
PUBLI C

Enabling Search
data-
aggregationsortorder
ASC Descending (DESC) or ascending order for sorting the counts in
the charts. Here you can override the sort order defined
centrally in the workarea widget.
data-topsmall none Defines how many bars or pie segments are displayed when the
chart is small.
data-
animationduration
1000 Duration of the animation, in milliseconds, when a chart adapts
(default 1000). 0 means that animation is switched off.
Related Links
12.3.8.8 inaresultlist
data-maxresultslarge 10 Maximum number of result items displayed on one
page if the list is enlarged in the content pane.
data-maxresultssmall 5 Maximum number of result items displayed on one
page if the list is small in the facet column.
data-responseattributes none Attributes to be displayed in a result list entry. The
actual scope and layout of the response attributes on
the result list and details pane are defined in the files
referenced in the data-resulttemplate
parameters.
Note
Do not leave the data-responseattributes
parameter empty.
Only use attributes from the main table of the view
or attributes that are joined in 1:1 from other tables.
Attributes with data type blob are not supported
by the HTTP service.
data-detailwidth 600 Width of the details pop-up in pixels.
data-detailheight 800 Height of the details pop-up in pixels.
data-resulttemplate none Reference to HTML and CSS files defining the layouts
of the result lists (small and large) and of the details
pop-up.
data-resulttemplate-css none
data-resulttemplate-small none
Enabling Search
PUBLI C

data-resulttemplate-small-
css
none
data-resulttemplate-detail none
data-resulttemplate-detail-
css
none
data-scoresortorder DESC By default, the result list is sorted according to the
ranking by relevancy for the search term(s). Leave the
default sort order unchanged.
data-orderby none Attribute by whose values the list is sorted
alphabetically. First choice for sorting is always the
ranking by relevancy for the search term(s). The
data-orderby setting only steps in if there is no
relevancy ranking available, for example before the
first search, or if there are equal ranking values.
data-sortorder ASC Ascending or descending (DESC) order for the data-
orderby parameter.
Related Links
Defining the Layout of Result Lists and Details [page 408]
12.3.9 Defining the Layout of Result Lists and Details
You can select which attributes are displayed and how they are displayed in the result lists (small or centered) and
on the details pop-up. In the search HTML file, the result templates are referenced in the inaresultlist widget
using the data-resulttemplate parameters.
The templates folder contains the HTML and CSS templates of the demo app for adapting. You can create your
own result layouts from the templates and reference them in the parameters. Make sure that the path is correct.
In the HTML templates, you define what is in a result entry in the different views. You can enter fixed text and
reference the response properties and values in double braces.
For a deeper understanding of these HTML templates, familiarize yourself with the Tempo JSON rendering
engine.
Note
If you want to integrate graphics in result list templates, define the exact width and height in pixels.
Use the corresponding CSS files to define the result layouts and the behavior on hover events. To tweak the layout
of the details screen further, you can also use the widgets provided by jQuery UI.
408
PUBLI C

Enabling Search
13 Setting Up Roles and Authorizations
13.1 The Authorization Model
Authorization mechanisms ensure that authenticated users can do only those things they are allowed to do. You
can use the authorization mechanisms of the SAP HANA database to specify who is allowed to access which data
and for which activities.
Authorization mechanisms can also be used to help control who is allowed to execute stored procedures and
functions or execute system-level activities, for example: making backups, creating schema, users, roles, and so
on.
Authorizations are specified by granting privileges to principals. Principals are either users or roles. Roles
represent collections of privileges that can be granted to users or other roles, which enables you to create a
hierarchy of roles. Privileges are granted to principals by users. In SAP HANA, you can use roles to assign
permissions for different areas to users. For example, a role enables you to assign SQL privileges, analytic
privileges, system privileges, package privileges, and so on. To create and maintain artifacts in the SAP HANA
repository, you can assign application-development users pre-defined roles that provide access to the areas and
objects they require.
Note
A SAP HANA user with the necessary authorization can modify existing roles, for example, to remove or restrict
privileges in particular areas. Authorized users can also create new customized roles, for example, to provide
application developers with precisely the privileges they need to perform the every-day tasks associated with
application development.
SAP HANA database authorization mechanisms use the following privileges:
SQL privileges
Run SQL commands and access table data
System privileges
Perform system-level operations or administrative tasks
Object privileges
Perform specified actions on specified database objects
Analytic privileges
Allow selective access control for database views generated when modeled are activated
Package Privileges
Allow operations on packages, for example, creation and maintenance. Privileges can differ for native and
imported packages.
Other types of privileges enable you to provide specific additional authorizations, for example, to users who need
to perform administrative tasks on objects in the repository or to users and clients that need access to
applications:
Repository Privileges
Enable users to perform basic repository operations, for example: import and export packages, and maintain
delivery units (DU)
Application Privileges
Enable users and clients to access an application, configure what data to expose, and set rules for exposing
URLs.
Setting Up Roles and Authorizations
PUBLI C

13.2 Authentication Overview
Authentication is the process used to validate the identity of the users who connect to SAP HANA.
Secure authentication mechanisms ensure that the users requesting a connection really are who they claim to be.
Note
For connections to SAP HANA it is important to distinguish between SQL- and HTTP-based connections.
For SQL access to SAP HANA by means of client interfaces, for example, JDBC or ODBC, the following
authentication methods are supported:
External authentication:
Kerberos services
A standardized service for client authentication in networks. The client connects to the Kerberos server to
authenticate the user and to acquire an authentication ticket that proves the users identity to the SAP
HANA database. The client uses the ticket to connect to the SAP HANA database. Kerberos also ensures
the correct identity of the server.
SAML (bearer token)
SAP HANA can authenticate users who provide a valid standard-based SAML assertion issued by a
trusted identity provider.
Internal authentication
Users are authenticated by the SAP HANA database using the database user name and password.
SAP HANA maps the external identity approved by external authentication service to the identity of an internal
database user. This internal database user is then used for authorization checks during the database session.
For HTTP access to SAP HANA by means of SAP HANA XS, the following authentication methods are supported:
User name and password:
Using HTTP basic authentication
SAP logon tickets
13.3 Roles
Roles contain privileges that are used to define which data, application, or function can be accessed and in which
manner consumed. Roles are assigned either to users or to other roles at runtime.
SAP delivers some generic roles with SAP HANA that can be used as they are or as templates (during design time)
for creating roles that meet your company's specific requirements. More information about roles, their use at
runtime and in design time are explained in the subsequent sections.
Roles
Roles as Repository Objects
Creating Roles in the Repository
410
PUBLI C

13.3.1 Roles
A role is a collection of privileges that can be granted to either a user or another role in runtime.
A role typically contains the privileges required for a particular function or task, for example:
Business end users reading reports using client tools such as Microsoft Excel
Modelers creating models and reports in the modeler of the SAP HANA studio
Database administrators operating and maintaining the database and users in the Administration editor of the
SAP HANA studio
Privileges can be granted directly to users of the SAP HANA database. However, roles are the standard
mechanism of granting privileges as they allow you to implement complex, reusable authorization concepts that
can be modeled on business roles. Several standard roles are delivered with the SAP HANA database (for
example, MODELING, MONITORING). You can use these as templates for creating your own roles.
Roles in the SAP HANA database can exist as runtime objects only, or as design-time objects that become
runtime objects on activation.
Role Structure
A role can contain any number of the following privileges:
System privileges for administrative tasks (for example, AUDIT ADMIN, BACKUP ADMIN, CATALOG READ)
Object privileges on database objects (for example, SELECT, INSERT, UPDATE)
Analytic privileges on SAP HANA information models
Package privileges on repository packages (for example, REPO.READ, REPO.EDIT_NATIVE_OBJECTS,
REPO.ACTIVATE_NATIVE_OBJECTS)
Application privileges for enabling access to SAP HANA XS applications
Note
Application privileges cannot be granted to roles in the SAP HANA studio.
A role can also extend other roles.
Role Modeling
You can model roles in the following ways:
As runtime objects
As design-time objects in the repository of the SAP HANA database
It is recommended that you model roles as design-time objects for the following reasons.
Firstly, unlike roles created in runtime, roles created as design-time objects can be transported between systems.
This is important for application development as it means that developers can model roles as part of their
application's security concept and then ship these roles or role templates with the application. Being able to
PUBLI C

transport roles is also advantageous for modelers implementing complex access control on analytic content. They
can model roles in a test system and then transport them into a productive system. This avoids unnecessary
duplication of effort.
Secondly, roles created as design-time objects are not directly associated with a database user. They are created
by the technical user _SYS_REPO and granted through the execution of stored procedures. Any user with access
to these procedures can grant and revoke a role. Roles created in runtime are granted directly by the database
user and can only be revoked by the same user. Additionally, if the database user is deleted, all roles that he or she
granted are revoked. As database users correspond to real people, this could impact the implementation of your
authorization concept, for example, if an employee leaves the organization or is on vacation.
Caution
The design-time version of a role in the repository and its activated runtime version should always contain the
same privileges. In particular, additional privileges should not be granted to the activated runtime version of a
role created in the repository. Although there is no mechanism of preventing a user from doing this, the next
time the role is activated in the repository, any changes made to the role in runtime will be reverted. It is
therefore important that the activated runtime version of a role is not changed in runtime.
13.3.2 Roles as Repository Objects
The repository of the SAP HANA database consists of packages that contain design-time versions of various
objects. Being a repository object has several implications for a role.
Grantable Privileges
According to the authorization concept of the SAP HANA database, a user can only grant a privilege to a user
directly or indirectly in a role if the following prerequisites are met:
The user has the privilege him- or herself
The user is authorized to grant the privilege to others (WITH ADMIN OPTION or WITH GRANT OPTION)
A user is also authorized to grant SQL object privileges on objects that he or she owns.
The technical user _SYS_REPO is the owner of all objects in the repository, as well as the runtime objects that are
created on activation. This means that when you create a role as a repository object, you can grant the following
privileges:
Privileges that have been granted to the technical user _SYS_REPO and that _SYS_REPO can grant further.
This is automatically the case for system privileges, package privileges, analytic privileges, and application
privileges. Therefore, all system privileges, package privileges, analytic privileges, and application privileges
can always be granted in modeled roles.
Privileges on objects that _SYS_REPO owns.
_SYS_REPO owns all activated objects. Object privileges on non-activated runtime objects must be explicitly
granted to _SYS_REPO. It is recommended that you use a technical user to do this to ensure that privileges
are not dropped when the granting user is dropped (for example, because she leaves the company.
The following table summarizes the situation described above:
412
PUBLI C

Privilege Action Necessary to Grant in Repository Role
System privilege None
Package privilege None
Analytic privilege None
Application privilege None
SQL object on activated object (for example, attribute
view, analytic view)
None
SQL object privilege on runtime object (for example,
replicated table)
Grant privilege to user _SYS_REPO with WITH GRANT
OPTION
Granting and Revoking
It is not possible to grant and revoke activated design-time roles using the GRANT and REVOKE SQL statements.
Instead, roles are granted and revoked through the execution of the procedures GRANT_ACTIVATED_ROLE and
REVOKE_ACTIVATED_ROLE. Therefore, to be able to grant or revoke a role, a user must have the SQL object
privilege EXECUTE on these procedures.
WITH ADMIN OPTION and WITH GRANT OPTION
When you create a role on the basis of SQL statements (that is, as a runtime object), you can grant privileges with
the additional parameters WITH ADMIN OPTION or WITH GRANT OPTION. This allows a user who is granted the
role to grant the privileges contained within the role to other users and roles. However, if you are implementing
your authorization concept with privileges encapsulated within roles created in design time, then you do not want
users to grant privileges using SQL statements. Therefore, it is not possible to pass the parameters WITH ADMIN
OPTION or WITH GRANT OPTION with privileges when you model roles as repository objects.
Similarly, when you grant an activated role to a user, it is not possible to allow the user to grant the role further
(WITH ADMIN OPTION is not available).
Auditability
The auditing feature of the SAP HANA database allows you to monitor and record selected actions performed in
your database system. One action that is typically audited is changes to user authorization. If you are using roles
created in the repository to grant privileges to users, then it is important to note that the creation of runtime roles
through activation cannot be meaningfully audited.
PUBLI C

13.3.3 Creating Roles in the Repository
You model roles in the SAP HANA repository in a domain-specific language (DSL).
A shared project must exist with a suitable package for storing roles.
You have the package and system privileges required for modeling and activating objects in the repository.
Caution
Theoretically, a user with authorization to model and activate repository objects can change a role that he
has been granted. Once the role is activated, the user has the new privileges that he or she just added.
Therefore, it is important that roles in productive systems are imported from a test or development system
and changes to imported objects are not allowed. This danger is however not specific to roles but also
applies to other repository objects, for example, modeled views.
You have granted privileges on non-activated runtime objects that you want to grant in the new role to the
technical user _SYS_REPO.
The following general conventions apply when modeling a role definition using the role DSL:
Comments start with a double-slash (//) or double-dash (--) and run to the end of the line.
When specifying a reference to a design-time object, you must always specify the package name as follows:
package::object if you are referencing a design-time role
package:object.extension if you are referencing any other design-time object
When specifying multiple privileges on the same object or the same privilege on multiple objects, you can do
so individually line by line, or you can group them on a single line. Separate multiple objects and/or multiple
privileges using a comma.
1. From the main menu in the SAP HANA studio, choose File New Other SAP HANA Development Role
.
The New Role dialog box appears.
2. In the Container field, enter the path to the package where you want to create the role.
3. In the Role name field, enter the name of the new role.
4. Choose Finish.
The new role appears in the Project Explorer view and opens in the role editor as follows:
// an empty role
role <package_name>::<role_name> {
}
The role is now ready to be defined.
5. Optional: Specify the role(s) that you want to embed within the new role.
You can specify both roles created in runtime and repository roles as follows:
extends role <package_name>::<role_name>
extends catalog role "role_name"
Example:
role <package_name>::<role_name>
extends role sap.example::role1
extends catalog role "CATROLE2"
414
PUBLI C

{
}
6. Model the required privileges.
Note
Unlike when you create a role using SQL statements, it is not possible to grant ALL PRIVILEGES when
creating a role in the repository. You must model every privilege individually.
a) Model system privileges using the keyword "system privilege" as follows: system privilege:
PRIVILEGE;
Example:
{
// multiple privileges in one line are OK
system privilege: BACKUP ADMIN, USER ADMIN;
// you can also split lists into multiple entries
system privilege: LICENSE ADMIN;
}
b) Optional: Model object privileges on the design-time objects views and procedures using the keyword
"SQL object" as follows: sql object <package>:<object>.extension: PRIVILEGE;
Example:
{
sql object sap.example:MY_VIEW.attributeview: DROP;
// object privileges can be split across lines
// a single privilege can be given on multiple objects in a single line
sql object sap.example:MY_VIEW.attributeview,
sap.example:MY_OTHER_VIEW.analyticview,
sap.example:MY_THIRD_VIEW.analyticview: SELECT;
}
c) Optional: Model object privileges on catalog objects using the keyword "catalog SQL object" as follows:
catalog sql object "SCHEMA"."CATALOG_OBJECT": PRIVILEGE;
Note
You must always qualify catalog objects with the schema name. You must also reference catalog
objects within double quotes, unlike design-time objects.
Example:
{
// catalog objects must always be qualified with the schema name
catalog sql object "MY_SCHEMA"."MY_VIEW": SELECT;
}
d) Optional: Model schema privileges on schemas using the keywords "catalog schema" or "schema" as
follows depending on whether you are referring to an activated schema or a schema in the repository:
PUBLI C

catalog schema "SCHEMA": PRIVILEGE;
schema <package>:<schema>.schema: PRIVILEGE;
Example:
{
catalog schema "MY_SCHEMA": SELECT;
schema sap.example:MY_OTHER_SCHEMA.schema: SELECT;
}
e) Optional: Model package privileges using the keywords "package" as follows: package PACKAGE:
PRIVILEGE;
Example:
{
package sap.example: REPO.READ;
}
f) Optional: Model analytic privileges using the keywords "analytic privilege" or "catalog analytic privilege"
depending on whether you are referring to an activated analytic privilege or a runtime analytic privilege
analytic privilege: <package>:<analytic_priv_name>.analyticprivilege
catalog analytic privilege: "analytic_priv_name";
Example:
{
analytic privilege: sap.example:sp1.analyticprivilege,
sap.example:AP2.analyticprivilege;
catalog analytic privilege: "sp3";
}
g) Optional: Model application privileges using the keyword "application privilege" as follows: application
privilege: <application_priv_name>;
Example:
{
application privilege: sap.example::Execute;
}
Note
Application privileges are implemented using the application-privileges file (.xsprivileges).
7. From the main menu, choose File Save .
The role is saved as an .hdbrole file. After it has been saved, the file is committed to the repository
8. Activate the role by right-clicking it in the Project Explorer view and choosing Team Activate .
Note
Any changes made to a previously activated version of the role in runtime will be reverted on activation.
This is to ensure that the design-time version of a role in the repository and its activated runtime version
416
PUBLI C

contain the same privileges. It is therefore important that the activated runtime version of a role is not
changed in runtime.
The activated role is now visible in the Navigator view under Security Roles following the naming convention
package::role_name and can be granted to users as part of user provisioning.
Example
Complete Role Definition Example
extends role sap.example::role1
extends catalog role "CATROLE1", "CATROLE2"
{
// system privileges
system privilege: BACKUP ADMIN, USER ADMIN;
// schema privileges
catalog schema "SYSTEM": SELECT;
schema sap.example:app1.schema: INSERT, UPDATE, DELETE;
// sql object privileges
// privileges on the same object may be split up in several lines
catalog sql object "SYSTEM"."TABLE2": SELECT;
catalog sql object "SYSTEM"."TABLE2": INSERT, UPDATE, DELETE;
// or a list of objects may get a list of privileges (object = table, view,
procedure, sequence)
// SELECT, DROP for all objects in list
sql object sap.example:VIEW1.attributeview, sap.example:PROC1.procedure,
sap.example:SEQ1.sequence: SELECT, DROP;
// additional INSERT, UPDATE for TABLE1
// analytic privileges
analytic privilege: sap.example:sp1.analyticprivilege,
sap.example:AP2.analyticprivilege;
catalog analytic privilege: "sp3";
// design time privileges
package sap.example: REPO.EDIT_NATIVE_OBJECTS;
package sap.example, sap.co: REPO.READ;
application privilege: sap.example::Execute, sap.example::Save;
}
Related Links
http://help.sap.com/hana/hana_admin_en.pdf
PUBLI C

13.4 Privileges
SAP HANA offers various privileges that can be assigned to users. Users can be a system, an application, a
service, or a person, and each user is assigned specific roles. Privileges can be assigned to users, roles, or both.
There are several privileges that are required either by the SAP HANA studio or by development tools, such as
SQL. The following types of privileges are explained in the subsequent sections.
System privileges
Object privileges
Package privileges
Analytic privileges
This section also contains information for creating and granting privileges to users.
13.4.1 System Privileges
System privileges are required to perform system-level operations or administrative tasks.
The following lists shows some of the most common system-level tasks that administrators regularly must
perform:
Database schema
Creation and deletion of database schema
Users/Roles
Maintenance and management of users and roles
Monitor/Trace
Administration of monitoring and tracing
Backup/Restore
Performance of backup and restore operations
In the SAP HANA database the user SYSTEM has all system privileges. The SYSTEM user can grant system
privileges to other users and roles.
13.4.2 Object Privileges
Object privileges enable users to perform specified actions on specified database objects.
Object privileges are not specific to the SAP HANA database; they are part of the SQL standard. You can use the
SQL statements GRANT and REVOKE to manage privileges.
Note
. If object privileges (or any other privilege type) are granted to roles created in the repository, granting and
revoking actions happen through the execution of stored procedures.
418
PUBLI C

The set of actions that can be allowed by an object privilege depends on the object type. For tables, you can use
actions such as: drop, alter, select, insert and update. An execute action is available for privileges for procedures
and functions. Privileges on a schema level are used to enable actions such as: create objects, perform select,
update, delete or execute operations on all objects contained in the schema.
In the SAP HANA database object privileges are not only available for database catalog objects such as tables,
views and procedures. Object privileges can also be granted for non-catalog objects such as development objects
in the repository of the SAP HANA database.
Some database objects depend on other objects. Views, for example, are defined as queries on other tables and
views. The authorization for an operation on the dependent object (the queried tables and views) requires
privileges for the dependent object and the underlying object. In case of views, the SAP HANA database
implements the standard SQL behavior. A user has the authorization for an operation on a view if the following is
true:
The privilege for operations on the view has been granted to the user or a role assigned to the user.
The owner of the view has the corresponding privileges on the underlying objects with the option to grant
them to others.
This behavior can be used to grant selective access to a table.
13.4.3 Package Privileges
In the SAP HANA repository, authorizations can be assigned to individual packages.
Authorizations assigned to a repository package are implicitly assigned to the design-time objects in the package
as well as to all sub-packages. Users are only allowed to maintain objects in a repository package if they have the
necessary privileges for the package in which they want to perform an operation, for example to read or write to
an object in that package.
Note
Package authorizations can be set for a specific user or for a role.
If the user-authorization check establishes that a user does not have the necessary privileges to perform the
requested operation in a specific package, the authorization check is repeated on the parent package and
recursively up the package hierarchy to the root level of the repository. If the user does not have the necessary
privileges for any of the packages in the hierarchy chain, the authorization check fails and the user is not
permitted to perform the requested operation.
In the context of repository authorizations, there is a distinction to be made between native packages and
imported packages.
Native Package
A package that is created in the current system and expected to be edited in the current system. Changes to
package objects must be performed in the original development system where they were created and
transported into subsequent systems. The content of native packages are regularly edited by developers.
Imported Package
A package that is created in a remote system and imported into the current system. Imported packages
should not usually be modified, except when replaced by new imports during an update. Otherwise, imported
packages should only be modified in exceptional cases, for example, to carry out emergency repairs.
PUBLI C

To perform every-day, package-related, application-development tasks in the repository, developers typically
need the following privileges for the application packages:
REPO.READ
REPO.EDIT_NATIVE_OBJECTS
REPO.ACTIVATE_NATIVE_OBJECTS
working in
REPO.MAINTAIN_NATIVE_PACKAGES
To perform every-day application-development tasks on imported packages, developers need the following
privileges:
Note
It is not recommended to work on imported packages. Imported packages should only be modified in
exceptional cases, for example, to carry out emergency repairs.
REPO.EDIT_IMPORTED_OBJECTS
Authorization to modify design-time objects in packages originating in a system other than the one the user is
working in
REPO.ACTIVATE_IMPORTED_OBJECTS
Authorization to activate/reactivate design-time objects in packages originating in a system other than the
one the user is working in
REPO.MAINTAIN_IMPORTED_PACKAGES
Authorization to update or delete native packages, or create sub-packages of packages originating in a
system other than the one in which the user is working
13.4.3.1 SAP HANA Repository Privileges
In addition to the authorizations you can grant users for specific packages (and the design-time objects in those
packages), you can also assign general system privileges to users to enable them to perform basic repository
operations, for example: importing and exporting content, and maintaining delivery units (DU).
If your daily tasks include general administration tasks, you need basic system-level privileges in the repository.
To perform basic administration tasks in the repository, you typically need the following privileges:
REPO.EXPORT
Enables you to export delivery units; exporting a DU transports it to another system
REPO.IMPORT
Enables you to import delivery units; importing a DU transports it from another system
REPO.MAINTAIN_DELIVERY_UNITS
Enables you to maintain your own delivery units (the DU-vendor must equal the system-vendor)
REPO.WORK_IN_FOREIGN_WORKSPACE
Enables you to move objects from a "foreign" inactive workspace into your own workspace
420
PUBLI C

You can grant or revoke basic, repository privileges to users in the following ways:
SAP HANA Studio
The System Privileges tab page in the Modeler perspective: Catalog -> Authorization -> Roles or Users
Caution
If you want to access the repository from the SAP HANA studio, Eclipse, or other clients,you need the
EXECUTE privilege for SYS.REPOSITORY_REST, the database procedure through which the REST API is
tunneled.
13.4.3.2 Defining Repository Package Privileges
To set user (or role) authorizations for repository packages, perform the following steps:
1. Open the Navigator view in the SAP HANA studio's Modeler perspective.
2. In the Navigator view, expand the Security Roles/Users node for the system hosting the repository that
contains the packages you want to grant access to.
You can also define roles via source files; roles defined in this way can be assigned to a delivery unit and
transported to other systems.
3. Double click the user (or role) to whom you want to assign authorizations.
4. Open the Package Privileges tab page.
5. Choose [+] to add one or more packages. Press and hold the Ctrl key to select multiple packages.
6. In the Select Repository Package dialog, use all or part of the package name to locate the repository package
that you want to authorize access to.
7. Select one or more repository packages that you want to authorize access to; the selected packages appear
in the Package Privileges tab page.
8. Select the packages to which you want authorize access and, in the Privileges for screen page, check the
required privileges, for example:
a) REPO.READ
b) REPO.EDIT_NATIVE_OBJECTS
c) REPO.ACTIVATE_NATIVE_OBJECTS
working in
d) REPO.MAINTAIN_NATIVE_PACKAGES
PUBLI C

13.4.4 Analytic Privileges
SQL privileges implement coarse-grained authorization at object level only. Users either have access to an object,
such as a table, view or procedure, or they do not. While this is often sufficient, there are cases when access to
data in an object depends on certain values or combinations of values. Analytic privileges are used in the SAP
HANA database to provide such fine-grained control of which data individual users can see within the same view.
Note
Sales data for all regions are contained within one analytic view. However, regional sales managers should only
see the data for their region. In this case, an analytic privilege could be modeled so that they can all query the
view, but only the data that each user is authorized to see is returned.
Analytic privileges are intended to control access to SAP HANA information models, that is:
Attribute views
Analytic views
Calculation views
Therefore, all column views modeled and activated in the SAP HANA modeler automatically enforce an
authorization check based on analytic privileges. Column views created using SQL must be explicitly registered
for such a check (by passing the parameter REGISTERVIEWFORAPCHECK).
Note
Analytic privileges do not apply to database tables or views modeled on row-store tables. Access to database
tables and row views is controlled entirely by SQL object privileges.
You create and manage analytic privileges in the SAP HANA modeler.
Note
Some advanced features of analytic privileges, namely dynamic value filters, can only be implemented using
SQL. The management of such analytic privileges created in SQL also varies to those created in the SAP HANA
modeler.
13.4.4.1 Structure of Analytic Privileges
An analytic privilege consists of a set of restrictions against which user access to a particular attribute view,
analytic view, or calculation view is verified. Each restriction controls the authorization check on the restricted
view using a set of value filters. A value filter defines a check condition that verifies whether or not the values of
the view (or view columns) qualify for user access.
The specification of these restrictions is contained in an XML document that conforms to a defined XML schema
definition (XSD).
The following restriction types can be used to restrict data access:
View
422
PUBLI C

Activity
Validity
Attribute
The following operators can be used to define value filters in the restrictions.
Note
The activity and validity restrictions support only a subset of these operators.
IN <list of scalar values>
CP <pattern with *>
EQ (=), LE (<=), LT (<), GT (>), GE (>=) <scalar value>
BT <scalar value as lower limit><scalar value as upper limit>
IS_NULL
NOT_NULL
All of the above operators, except IS_NULL and NOT_NULL, accept empty strings (" ") as filter operands. IS_NULL
and NOT_NULL do not allow any input value.
The following are examples of how empty strings can be used with the filter operators:
For the IN operator: IN ("", "A", "B") to filter on these exact values
As a lower limit in comparison operators, such as:
BT ("", XYZ), which is equivalent to NOT_NULL AND LE "XYZ"GT "", which is equivalent to NOT_NULL
LE "", which is equivalent to EQ ""
LT "", which will always return false
CP "", which is equivalent to EQ ""
The filter conditions CP "*" will also return rows with empty-string as values in the corresponding attribute.
View Restriction
This restriction specifies to which column view(s) the analytic privilege applies. It can be a single view, a list of
views, or all views. An analytic privilege must have exactly one cube restriction.
Example: IN ("Cube1", "Cube2")
Note
When an analytic view is created in the SAP HANA modeler, automatically-generated views are included
automatically in the cube restriction.
Note
The SAP HANA modeler uses a special syntax to specify the cube names in the view restriction:
_SYS_BIC:<package_hierarchy>/<view_name>
For example:
<cubes>
<cube name="_SYS_BIC:test.sales/AN_SALES" />
PUBLI C

<cube name="_SYS_BIC:test.sales/AN_SALES/olap" />
</cubes>
Activity Restriction
This restriction specifies the activities that the user is allowed to perform on the restricted view(s), for example,
read data. An analytic privilege must have exactly one activity restriction.
Example: EQ "read", or EQ "edit"
Note
Currently, all analytic privileges created in the SAP HANA modeler are automatically configured to restrict
access to READ activity only. This corresponds to SQL SELECT queries. This is due to the fact that the
attribute, analytic, and calculation views are read-only views. This restriction is therefore not configurable.
Validity Restriction
This restriction specifies the validity period of the analytic privilege. An analytic privilege must have exactly one
validity restriction.
Example: GT 2010/10/01, 01:01:00.000
Note
All analytic privileges automatically become immediately valid and have unlimited validity when activated in the
SAP HANA modeler. This restriction is therefore not configurable.
Attribute Restriction
This restriction specifies the value range that the user is permitted to access. Attribute restrictions are applied to
the actual attributes of a view. Each attribute restriction is relevant for one attribute, which can contain multiple
value filters. Each value filter represents a logical filter condition.
Note
The HANA modeler uses different ways to specify attribute names in the attribute restriction depending on the
type of view providing the attribute. In particular, attributes from attribute views are specified using the syntax
"<package_hierarchy>/<view_name>$<attribute_name>", while local attributes of analytical views and
calculation views are specified using their attribute name only. For example:
<dimensionAttribute name="test.sales/AT_PRODUCT$PRODUCT_NAME">
<restrictions>
<valueFilter operator="IN">
424
PUBLI C

<value value="Car" />
<value value="Bike" />
</valueFilter>
</restrictions>
</dimensionAttribute>
Value filters for attribute restrictions can be static or dynamic.
A static value filter consists of an operator and either a list of values as the filter operands or a single value as
the filter operand. All data types are supported except those for LOB data types (CLOB, BLOB, and NCLOB).
For example, a value filter (EQ 2006) can be defined for an attribute YEAR in a dimension restriction to filter
accessible data using the condition YEAR=2006 for potential users.
Note
Only attributes, not aggregatable facts (for example, measures or key figures) can be used in dimension
restrictions for analytic views.
A dynamic value filter consists of an operator and a stored procedure call that determines the operand value
at runtime.
For example, a value filter (IN (GET_MATERIAL_NUMBER_FOR_CURRENT_USER())) is defined for the
attribute MATERIAL_NUMBER. This filter indicates that a user with this analytic privilege is only allowed to
access material data with the numbers returned by the procedure
GET_MATERIAL_NUMBER_FOR_CURRENT_USER.
Note
If you create an analytic privilege in the SAP HANA modeler, it is not possible to specify a procedure call
when defining restrictions. If you want to create analytic privileges with dynamic filter conditions, you must
do so using SQL.
It is possible to combine static and dynamic value filters as shown in the following example:
<dimensionAttribute name=" test.sales/AT_PRODUCT$PRODUCT_NAME ">
<restrictions>
<valueFilter operator="CP"> <value value="ELECTRO*"/> </valueFilter>
<valueFilter operator="IN"> <procedureCall schema="PROCEDURE_OWNER"
procedure="DETERMINE_AUTHORIZED_PRODUCT_FOR_USER" /> </valueFilter >
</restrictions>
<dimensionAttribute name=" test.sales/AT_TIME$YEAR ">
<restrictions>
<valueFilter operator="EQ"> <value value="2012"/> </valueFilter>
<valueFilter operator="IN"> <procedureCall schema="PROCEDURE_OWNER"
procedure="DETERMINE_AUTHORIZED_YEAR_FOR_USER" /> </valueFilter >
</restrictions>
An analytic privilege can have multiple attribute restrictions, but it must have at least one attribute restriction. An
attribute restriction must have at least one value filter. Therefore, if you want to permit access to the whole
content of a restricted view, then the attribute restriction must specify all attributes.
Similarly, if you want to permit access to the whole content of the view with the corresponding attribute, then the
value filter must specify all values.
The SAP HANA modeler automatically implements these two cases if you do not select either an attribute
restriction or a value filter. However, if you create the analytic privilege using SQL, you must make the
specification manually.
PUBLI C

Example 1: Specification of all attributes
<dimensionAttributes>
<allDimensionAttributes/ >
</dimensionAttributes>
Example 2: Specification of all values of an attribute
<dimensionAttribute name="PRODUCT">
<all />
Logical Combination of Restrictions and Filter Conditions
The result of user queries on restricted views is filtered according to the conditions specified by the analytic
privileges granted to the user as follows:
Multiple analytic privileges are combined with the logical operator OR.
Within one analytic privilege, all attribute restrictions are combined with the logical operator AND.
Within one attribute restriction, all value filters on the attribute are combined with the logical operator OR.
Example
You create two analytic privileges AP1 and AP2. AP1 has the following attribute restrictions:
Restriction R11 restricting the attribute Year with the value filters (EQ 2006) and (BT 2008, 2010)
Restriction R12 restricting the attribute Country with the value filter (IN ("USA", "Germany"))
Given that multiple value filters are combined with the logical operator OR and multiple attribute restrictions
are combined with the logical operator AND, AP1 generates the condition:
((Year = 2006) OR (Year BT 2008 and 2010)) AND (Country IN ("USA", "Germany"))
AP2 has the following restriction:
Restriction R21 restricting the attribute Country with the value filter (EQ "France")
AP2 generates the condition:
(Country = "France")
Any query of a user who has been granted both AP1 and AP2 will therefore be appended with the following
WHERE clause:
((Year = 2006) OR (Year BT 2008 and 2010)) AND (Country IN ("USA", "Germany"))) OR
(Country = "France")
426
PUBLI C

13.4.4.2 Dynamic Value Filters in the Attribute Restriction of
Analytic Privileges
The attribute restriction of an analytic privilege specifies the value range that the user is permitted to access using
value filters. In addition to static scalar values, stored procedures can be used to define filters. This allows user-
specific filter conditions to be determined dynamically in runtime, for example, by querying specified tables or
views. As a result, the same analytic privilege can be applied to many users, while the filter values for authorization
can be updated and changed independently in the relevant database tables. In addition, application developers
have full control not only to design and manage such filter conditions, but also to design the logic for obtaining the
relevant filter values for the individual user at runtime.
Procedures used to define filter conditions must have the following properties:
They must have the security mode DEFINER.
They must be read-only procedures.
A procedure with a predefined signature must be used. The following conditions apply:
No input parameter
Only 1 output parameter as table type with one single column for the IN operator
Only 1 output parameter of a scalar type for all unary operators, such as EQUAL
Only 2 output parameters of a scalar type for the binary operator BETWEEN
Only the following data types are supported as the scalar types and the data type of the column in the table
type:
Date/Time types DATE, TIME, SECONDDATE, and TIMESTAMP
Numeric types TINYINT, SMALLINT, INTEGER, BIGINT, DECIMAL, REAL, and DOUBLE
Character string types VARCHAR and NVARCHAR
Binary type VARBINARY
NULL as Operand for Filter Operators
In static value filters, it is not possible to specify NULL as the operand of the operator. The operators IS_NULL or
NOT_NULL must be used instead. In dynamic value filters where a procedure is used to determine a filter
condition, NULL or valid values may be returned. The following behavior applies in the evaluation of such cases
during the authorization check of a user query:
Filter conditions of operators with NULL as the operand are disregarded, in particular the following:
EQ NULL, GT NULL, LT NULL, LE NULL, and CP NULL
BT NULL and NULL
If no valid filter conditions remain (that is, they have all been disregarded because they contain the NULL
operand), the user query is rejected with a Not authorized error.
Example
Dynamic analytic privilege 1 generates the filter condition (Year >= NULL) and dynamic analytic privilege 2
generates the condition (Country EQ NULL). The query of a user assigned these analytic privileges (combined
with the logical operator OR) will return a Not authorized error.
PUBLI C

Example
Dynamic analytic privilege 1 generates the filter condition (Year >= NULL) and dynamic analytic privilege 2
generates the condition (Country EQ NULL AND Currency = USD). The query of a user assigned these
analytic privileges (combined with the logical operator OR) will be filtered with the filter Currency = USD.
In addition, a user query is not authorized in the following cases even if further applicable analytic privileges have
been granted to the user.
The BT operator has as input operands a valid scalar value and NULL, for example, BT 2002 and NULL or BT
NULL and 2002
The IN operator has as input operand NULL among the value list, for example, IN (12, 13, NULL)
Permitting Access to All Values
If you want to allow the user to see all the values of a particular attribute, instead of filtering for certain values, the
procedure must return "*" and '' '' (empty string) as the operand for the CP and GT operators respectively. These
are the only operators that support the specification of all values.
Implementation Considerations
When the procedure is executed as part of the authorization check in runtime, note the following:
The user who must be authorized is the database user who executes the query accessing a secured view. This
is the session user. The database table or view used in the procedure must therefore contain a column to
store the user name of the session user. The procedure can then filter by this column using the SQL function
SESSION_USER. This table or view should only be accessible to the procedure owner.
The user executing the procedure is the _SYS_REPO user. In the case of procedures activated in the SAP
HANA modeler, _SYS_REPO is the owner of the procedures. For procedures created in SQL, the EXECUTE
privilege on the procedure must be granted to the _SYS_REPO user.
If the procedure fails to execute, the users query stops processing and a Not authorized error is returned.
The root cause can be investigated in the error trace file of the indexserver,
indexserver_alert_<host>.trc.
When designing and implementing procedures as filter for dynamic analytic privileges, bear the following in mind:
To avoid a recursive analytic privilege check, the procedures should only select from database tables or views
that are not subject to an authorization check based on analytic privileges. In particular, views activated in the
SAP HANA modeler are to be avoided completely as they are automatically registered for the analytic
privilege check.
The execution of procedures in analytic privileges slows down query processing compared to analytic
privileges containing only static filters. Therefore, procedures used in analytic privileges must be designed
carefully.
428
PUBLI C

13.4.4.3 Runtime Authorization Check of Analytic Privileges
When a user requests access to data stored in an attribute, analytic, or calculation view, an authorization check
based on analytic privileges is performed and the data returned to the user is filtered accordingly. Access to a
view and the way in which results are filtered depend on whether the view is independent or associated with other
modeling views (dependent views).
Independent Views
The authorization check for a view that is not defined on another modeling view is as follows:
1. The user requests access to an individual view, for example with a SELECT query.
This is possible if the both of the following prerequisites are met:
The user has been granted the SQL privilege SELECT on the view or the schema in which it is located.
The user has been granted an analytic privilege that is applicable to the view. An analytic privilege is
applicable to a view if it contains the view in the view restriction and at least one filter on one attribute of
the view.
Note
The user does not require the SELECT privilege on the underlying base tables or views of the modeling
view.
2. The authorization check determines the analytic privileges that are relevant for the current user and view.
Relevant analytic privileges are those that met all of the following criteria:
Analytic privileges previously granted to the user, either directly or indirectly through a role
Analytic privileges with a view restriction that includes the accessed view
Analytic privileges with a currently valid validity restriction
Note
This check is always positive for analytic privileges created and activated in the SAP HANA modeler.
Analytic privileges with an activity restriction covering the activity requested by the query are considered
Note
This check is always positive for analytic privileges created and activated in the SAP HANA modeler.
Analytic privileges with dimension restrictions covering some of the views attributes
3. If no relevant analytic privileges are found, the users queries are rejected with a Not authorized error.
This means that even though the user has the SELECT privilege on the view, access is not possible.
If the user does have a relevant analytic privilege but does not have the SELECT privilege on the view, access
is also not possible. If relevant analytic privileges are found, the authorization check evaluates the value filters
specified in the dimension restrictions and presents the appropriate data to the user.
Note
PUBLI C

Multiple dimension restrictions and/or multiple value filters are combined as described in the section on
the structure of analytic privileges.
Dependent Views
Calculation views can be defined by selecting data from other column views, in particular, attribute views, analytic
views, and other calculation views. This can lead to a complex view hierarchy that requires careful design of row-
level authorization.
If a user requests access to such a calculation view that is dependent on another view, the behavior of the
authorization check and result filtering is performed as follows:
A user can access a calculation view based on other view(s) if both of the following prerequisites are met:
The user has been granted the SELECT privilege on the calculation view or the schema that contains the
calculation view.
The user has been granted analytic privileges that cover the calculation view itself and all the other column
views in the hierarchy.
Result filtering on the calculation view is performed as follows:
Individual views in the hierarchy are filtered according to their respective analytic privileges, which use the
logical OR combination.
The filtered result of the calculation view is derived from the filtered result of its underlying views. This
corresponds to a logic AND combination of the filters generated by the analytic privileges for the individual
views.
Note
If an analytic view designed in the SAP HANA modeler contains one of the elements listed below, the analytic
view will be activated automatically with an calculation view (with the name of the analytic view appended by
the suffix /olap) on top of it.
Currency or unit conversions
Calculated attributes
Calculated measures that use attributes, calculated attributes, or input parameters in their formulas
This represents a view hierarchy for which the prerequisites described above apply.
Note
An analytic view can be defined on attribute views. However, this in no way represents a view dependency or
hierarchy with respect to authorization check and result filtering. Analytic privileges are checked and combined
independently for the analytic view and their related attribute views. However, as the analytic view inherits and
exposes the same attributes from the attribute views, the same filters on the same attribute can be used in
analytic privileges to filter an analytic view and its related attribute views. This is convenient to define in a single
analytic privilege with the view restriction including both the analytic view and the related attribute views to
control access to the views together.
430
PUBLI C

13.4.4.4 Creation and Management of Analytic Privileges
Analytic privileges can be created, dropped, and changed in the SAP HANA modeler and using SQL statements.
The SAP HANA modeler should be used in all cases except if you are creating analytic privileges that use
dynamic procedure-based value filters.
To create analytic privileges, the system privilege CREATE STRUCTURED PRIVILEGE is required. To drop analytic
privileges, the system privilege STRUCTUREDPRIVILEGE ADMIN is required.
In the SAP HANA modeler, repository objects are technically created by the technical user _SYS_REPO, which by
default has the system privileges for both creating and dropping analytic privileges. To be able to create, activate,
drop, and redeploy analytic privileges in the SAP HANA modeler therefore, a database user requires the package
privileges REPO.EDIT_NATIVE_OBJECTS and REPO.ACTIVATE_NATIVE_OBJECTS for the relevant package.
Implications of Creating Analytic Privileges Using SQL
The SAP HANA modeler is the recommended method for creating and managing analytic privileges. However, it is
necessary to use SQL to implement those features of analytic privileges not available in the modeler, that is,
dynamic, procedure-based value filters as attribute restrictions.
In the SAP HANA modeler, analytic privileges are created as design-time repository objects owned by the
technical user _SYS_REPO. They must be activated to become runtime objects available in the database. Analytic
privileges created using SQL statements are activated immediately. However, they are also owned by the
database user who executes the SQL statements. This is the main disadvantage of using SQL to create analytic
privileges. If the database user who created the analytic privilege is deleted, all objects owned by the user will also
be deleted. Therefore, if you are using SQL to create analytic privileges, we recommend that you create a
dedicated database user (that is, a technical user) for this purpose to avoid the potential loss of complex modeled
privileges.
An additional disadvantage of creating analytic privileges using SQL is that these analytic privileges are not in the
SAP HANA repository and they cannot be transported between different systems.
Granting and Revoking Analytic Privileges
Analytic privileges are granted and revoked as part of user provisioning.
If the analytic privilege was created and activated using the SAP HANA modeler, the analytic privilege is owned by
the _SYS_REPO user. Therefore, to be able to grant and revoke the analytic privilege, a user needs the privilege
EXECUTE on the procedures GRANT_ACTIVATED_ANALYTICAL_PRIVILEGE and
REVOKE_ACTIVATED_ANALYTICAL_PRIVILEGE respectively.
If the analytic privilege was created using SQL, only the owner (that is, the creator) of the analytic privilege can
grant and revoke it.
Related Links
SAP HANA Administration Guide
PUBLI C

13.4.4.5 Example: Creating an Analytic Privilege with Dynamic
Value Filter
This example shows you how to create an analytic privilege that contains a dynamic procedure-based value filter
and a fixed value filter in the attribute restriction.
Assume you want to restrict access to product data in secured views as follows:
Users should only see products beginning with ELECTRO, or
Users should only see products for which they are specifically authorized. This information is contained in the
database table PRODUCT_AUTHORIZATION_TABLE in the schema AUTHORIZATION.
To be able to implement the second filter condition, you need to create a procedure that will determine which
products a user is authorized to see by querying the table PRODUCT_AUTHORIZATION_TABLE.
1. Create the table type for the output parameter of the procedure:
CREATE TYPE "AUTHORIZATION"."PRODUCT_OUTPUT" AS TABLE("PRODUCT" int);
2. Create the table that the procedure will use to check authorization:
CREATE TABLE "AUTHORIZATION"."PRODUCT_AUTHORIZATION_TABLE" ("USER_NAME"
NVARCHAR(128), "PRODUCT" int);
3. Create the procedure that will determine which products the database user executing the query is authorized
to see based on information contained in the product authorization table:
CREATE PROCEDURE "AUTHORIZATION"."DETERMINE_AUTHORIZED_PRODUCT_FOR_USER" (OUT
VAL "AUTHORIZATION"."PRODUCT_OUTPUT")
LANGUAGE SQLSCRIPT SQL SECURITY DEFINER READS SQL DATA AS
BEGIN
VAL = SELECT PRODUCT FROM "AUTHORIZATION"."PRODUCT_AUTHORIZATION_TABLE
WHERE USER_NAME = SESSION_USER;
END;
Note
The session user is the database user who is executing the query to access a secured view. This is
therefore the user whose privileges must be checked. For this reason, the table or view used in the
procedure should contain a column to store the user name so that the procedure can filter on this column
using the SQL function SESSION_USER.
4. Create the analytic privilege:
CREATE STRUCTURED PRIVILEGE '<?xml version="1.0" encoding="utf-8"?>
<analyticPrivilegeSchema version="1">
<analyticPrivilege name="AP2">
<cubes>
<allCubes />
</cubes>
<validity>
<anyTime/>
</validity>
<activities>
<activity activity="read" />
</activities>
<dimensionAttribute name="PRODUCT">
<restrictions>
432
PUBLI C

<valueFilter operator="CP"> <value value="ELECTRO*"/> </valueFilter>
<valueFilter operator="IN"> <procedureCall schema="AUTHORIZATION"
procedure="DETERMINE_AUTHORIZED_PRODUCT_FOR_USER"/> </valueFilter>
</restrictions>
</analyticPrivilege>
</analyticPrivilegeSchema>';
Now when a database user requests access to a secured view containing product information, the data returned
will be filtered according to the following condition:
(product LIKE "ELECTRO*" OR product IN
(AUTHORIZATION.DETERMINE_AUTHORIZED_PRODUCT_FOR_USER())
13.4.5 Creating Analytic Privileges
To create, activate and drop the privilege you have system privileges CREATE STRUCTURED PRIVILEGE and
STRUCTUREDPRIVILEGE ADMIN.
Make sure that both CREATE STRUCTURED PRIVILEGE and STRUCTUREDPRIVILEGE ADMIN are correctly
owned by SYS_REPO user.
To activate and redeploy analytic privileges in the Modeler, a database user requires corresponding repository
privileges, namely REPO.EDIT_NATIVE_OBJECTS and REPO.ACTIVATE_NATIVE_OBJECTS
You apply analytic privileges when business users access values with certain combinations of dimension
attributes. You can use analytic privileges to partition data among various users sharing the same data
foundation. You can define restrictions for a selected group of models or apply them to all content models across
packages.
After activation, an analytic privilege needs to be assigned to a user before taking any effect. The user views the
filtered data based on the restrictions defined in the analytic privilege. If no analytic privilege applicable for models
is assigned to a user, he cannot access the model. If a user is assigned to multiple analytic privileges, the
privileges are combined with OR conditions.
Remember
In addition to the analytic privileges, a user needs SQL Select privileges on the generated column views.
The generated column views adhere to the following naming conventions:
For a view MyView in package p1.p2 (i.e. subpackage p2 of package p1) the generated column view lies in
schema _SYS_BIC and is named _SYS_BIC.p1.p2/MyView. Ensure that the users who are allowed to see the
view have select privileges on the view (or the entire schema _SYS_BIC).
Note
Multiple restrictions applied on the same column are combined by OR. However, restrictions across several
columns are always combined by AND.
1. Set Parameters for the Analytic Privilege
a) In the Modeler perspective, expand the system node from the Navigator view.
PUBLI C

b) Expand the Content node.
c) Select the required package.
d) From the context menu, choose New Analytic Privilege .
1. Enter a name and description for the analytic privilege, and choose Next.
2. Select the information models that you want to use for applying restrictions.
3. Choose Finish.
Tip
You can choose to add more models in the editor pane.
2. Add Attributes
a) If you want to add restrictions for all models, select the Applicable to all Information Models option.
Note
If you do not select this option, the restrictions you create apply only to the secured list of models
available in the Reference Models panel that you selected above.
b) In the Associated Attributes Restrictions panel, choose Add to select the attributes for defining
restrictions.
c) Choose OK.
Note
If you do not add any attributes for restrictions there will be unrestricted access to the selected models
or to all the models (if Applicable to all the Information Models option is selected).
3. Assign Restrictions
a) In the Assign Restrictions pane, choose Add to add value restriction for the selected attributes.
b) Select the required operator and enter a value (manually or via Value Help dialog).
Remember
To activate the analytic privilege, you must assign a minimum of one restriction to each attribute.
4. Activate the analytic privilege using the Save and Activate option in the editor.
Note
Select the Save and Activate All option to activate the privilege along with all the required objects.
5. Assign the privilege to a user
a) In the Navigator view, navigate the Security Authorizations Users node.
b) Select a user.
c) In the context menu, choose Open.
d) Choose Analytic Privileges tab page, and add the privilege.
e) From the editor toolbar, choose Deploy.
Example
434
PUBLI C

Consider an analytic view (without fields coming from attribute views) or a calculation view SALES, which is
added as part of an analytic privileges secured models having the following data.
CUST_ID CUST_GROUP SALES
1 GRP1 1000
2 GRP2 1500
3 GRP3 1200
1 GRP4 1300
If you create a restriction on column CUST_ID to filter data for CUST_ID 1 and 2, the conditions are
combined with OR and the data available for a user is:
1 GRP1 1000
2 GRP2 1500
1 GRP4 1300
If you create restrictions on columns CUST_ID and CUST_GROUP such as CUST_ID = 1 and CUST_GROUP
= 1, the conditions are combined with AND, and the data available for a user is:
1 GRP1 1000
Note
The technical name used for attributes of calculation views and private attributes of analytic views,
is the same as that of the attribute name. Hence any restriction applied to a private attribute of an
analytic or calculation view attribute is also applied to any other private attribute of an analytic view
and calculation view attribute having the same name.
In the above example, if there is any other analytic view or calculation view, which is part of a
privileges secured list of models, and has a field called CUST_ID (not coming from any attribute
view), the data for these privileges also gets restricted.
If Applicable to all information models is selected, any analytic view/calculation view (even if not part
of the secured models) which has a (private) attribute called CUST_ID, the data for these
privileges also get restricted.
The behavior for the calculation view is the same as that of the analytic view described above.
Consider an attribute view CUSTOMER which is part of an analytic privileges secured list of models having
the following data.
CUST_ID COUNTRY MANDT
1 IN 1
2 IN 1
3 US 1
1 DE 2
If you create a restriction on column CUST_ID to filter data for CUST_ID 1 and 2, the conditions are
combined with OR and the data is shown as follows:
PUBLI C

1 IN 1
2 IN 1
1 DE 2
If you create restrictions on columns CUST_ID and COUNTRY such as CUST_ID = 1 and COUNTRY = IN, the
conditions are combined with AND, and the data available for a user is:
1 IN 1000
Note
The technical name used for an attribute view attribute is <package name>/<attribute view name>
$<attribute name>. Hence, in the above example, the technical name for CUST_ID is mypackage/
CUSTOMER$CUST_ID. This implies that if there is any other attribute view STORE which is a part
of the analytic privilege and has CUST_ID as its attribute, it will not get restricted.
Any analytic view that is part of the privileges secured list of models and has this attribute view as
its required object, gets restricted using the technical name. In the example above, if an analytic
view contains the attribute views CUSTOMER and STORE, both CUST_ID attributes are handled
independently, because their internal technical name used for the privilege check are mypackage/
CUSTOMER$CUST_ID and myotherpackage/STORE$UST_ID.
If Applicable to all information models is selected, any analytic view (even if not part of the secured
models) having this attribute view as its required object, also gets restricted.
Related Links
Structure of Analytic Privileges [page 422]
An analytic privilege consists of a set of restrictions against which user access to a particular attribute view,
analytic view, or calculation view is verified. Each restriction controls the authorization check on the restricted
view using a set of value filters. A value filter defines a check condition that verifies whether or not the values of
the view (or view columns) qualify for user access.
Runtime Authorization Check of Analytic Privileges [page 429]
When a user requests access to data stored in an attribute, analytic, or calculation view, an authorization check
based on analytic privileges is performed and the data returned to the user is filtered accordingly. Access to a
view and the way in which results are filtered depend on whether the view is independent or associated with other
modeling views (dependent views).
Dynamic Value Filters in the Attribute Restriction of Analytic Privileges [page 427]
The attribute restriction of an analytic privilege specifies the value range that the user is permitted to access using
value filters. In addition to static scalar values, stored procedures can be used to define filters. This allows user-
specific filter conditions to be determined dynamically in runtime, for example, by querying specified tables or
views. As a result, the same analytic privilege can be applied to many users, while the filter values for authorization
can be updated and changed independently in the relevant database tables. In addition, application developers
have full control not only to design and manage such filter conditions, but also to design the logic for obtaining the
relevant filter values for the individual user at runtime.
Creation and Management of Analytic Privileges [page 431]
436
PUBLI C

Analytic privileges can be created, dropped, and changed in the SAP HANA modeler and using SQL statements.
The SAP HANA modeler should be used in all cases except if you are creating analytic privileges that use
dynamic procedure-based value filters.
Example: Creating an Analytic Privilege with Dynamic Value Filter [page 432]
This example shows you how to create an analytic privilege that contains a dynamic procedure-based value filter
and a fixed value filter in the attribute restriction.
13.4.6 Granting Privileges to Users
Privileges can be granted to database users either directly, or indirectly through roles that they have been
granted. In this case, the privileges are inherited. Roles are the standard mechanism of granting privileges to
users.
To be able to grant and revoke privileges and roles to and from users and roles, the following prerequisites must
be met:
Action Prerequisite
Grant system privilege, object privilege, or package
privilege to user or role
Granting user must have the privilege being granted
and be authorized to grant it to other users and roles
Grant analytic privilege to user or role Granting user must have the object privilege EXECUTE
on the procedure
GRANT_ACTIVATED_ANALYTICAL_PRIVILEGE
Grant object privilege on activated modeled objects,
such as calculation views, to user or role
Granting user must have the object privilege EXECUTE
on the procedure
GRANT_PRIVILEGE_ON_ACTIVATED_CONTENT
Grant role created in runtime to user or role
Granting user must have the role being granted
and be authorized to grant it to other users and
roles, or
Granting user must be have the system privilege
ROLE ADMIN
Grant role created in the repository to user or role Granting user must have the object privilege EXECUTE
on the procedure GRANT_ACTIVATED_ROLE
Grant application privilege Granting user must have the object privilege EXECUTE
on the procedure GRANT_APPLICATION_PRIVILEGE
Grant object privilege on schema containing activated
modeled objects, such as calculation views, to user or
role
Granting user must have the object privilege EXECUTE
on the procedure
GRANT_SCHEMA_PRIVILEGE_ON_ACTIVATED_CONT
ENT
Revoke system privilege, object privilege, or package
privilege from user or role
Revoking user must be the user who granted the
privilege
Revoke object privilege on schema activated
containing modeled objects, such as calculation views,
from user or role
Revoking user must have the object privilege EXECUTE
on the procedure
PUBLI C

Action Prerequisite
REVOKE_SCHEMA_PRIVILEGE_ON_ACTIVATED_CON
TENT
Revoke analytic privilege from user or role Revoking user must have the object privilege EXECUTE
on the procedure
REVOKE_ACTIVATED_ANALYTICAL_PRIVILEGE
Revoke role created in runtime from user or role Revoking user must be the user who granted the role
Revoke role created in the repository from user or role Revoking user must have the object privilege EXECUTE
on the procedure REVOKE_ACTIVATED_ROLE
Revoke application privilege Revoking user must have the object privilege EXECUTE
on the procedure REVOKE_APPLICATION_PRIVILEGE
Revoke object privilege on activated modeled objects,
such as calculation views from user or role
Revoking user must have the object privilege EXECUTE
on the procedure
REVOKE_PRIVILEGE_ON_ACTIVATED_CONTENT
1. In the Navigator view, choose Security Users .
2. Open the relevant user and grant the required roles and privileges (object privileges, analytic privileges,
system privileges, and package privileges) to the user.
To grant a role or privilege, choose the button and search for the required role or privilege.
To allow the user to pass on his or her privileges to other users, select Grantable to other users and roles.
Note
This option is not available if you are granting a role that was created in the repository.
Note
You cannot grant application privileges directly to users in the SAP HANA studio. It is recommended that
you grant application privileges to roles created in the repository and then grant the role to the user.
3. Choose the (Deploy) button to save the changes.
Related Links
http://help.sap.com/hana/hana1_imdb_studio_admin_en.pdf
13.5 Application Access
When you develop and deploy applications in the context of SAP HANA Extended Application Services (SAP
HANA XS), you must define the application descriptors, which describe the framework in which the application
runs. This application framework includes the root point in the package hierarchy where content is to be served to
client requests, whether or not the application is permitted to expose data to client requests, what kind of access
to the data is allowed, and what if any privileges are required to perform actions on application-related packages
and package content. The application descriptors include the following files:
438
PUBLI C

Application-descriptor file
The location of the application-descriptor file defines the root path of the resources exposed to client
requests by the application you develop. You can further restrict access using the application-access and
application-privileges files.
Application-Access File
a SAP HANA XS application package and what content they are allowed to see. You can use keywords in the
application-access file to set authentication rules, define package-privilege levels (for example, EXECUTE or
ADMIN), specify the connection security level (for example, SSL/HTTPS), allow or prevent the creation of
entity tags (Etags), and so on. You can also define rewrite rules for URLs exposed by an application, for
example, to hide internal details of URL paths from external users, clients, and search engines.
Application-Privileges File
The application-privileges file enables you to define the authorization privileges required for access to an SAP
HANA XS application, for example, to start the application (EXECUTE) or to perform administrative actions on
an application (ADMIN). The privileges defined here are activated for a particular application in the
application-access file. These privileges can be checked by an application at runtime. Privileges defined apply
to the package where the privileges file is located as well as any packages further down the package hierarchy
unless an additional privileges file is present, for example, in a subpackage.
Note
An additional access-related file, the SQL-connection configuration file, enables you to execute SQL
statements from inside your SAP HANA XS JavaScript application with credentials that are different to the
credentials of the requesting user. Each SQL-connection configuration object has a unique name, and the
administrator can assign specific, individual database users to this configuration.
Related Links
The Application-Descriptor File [page 55]
XS) must have an application descriptor file. The application descriptor is the core file that you use to describe an
The Application-Access File [page 56]
SAP HANA XS enables you to define access to each individual application package that you want to develop and
deploy.
PUBLI C

14 Implementing Lifecycle Management
14.1 SAP HANA Delivery Units
In SAP HANA, the delivery unit (DU) is the vehicle that lifecycle management (LCM) uses to ship one or more
software components from SAP (or a partner) to a customer .
Delivery Units correspond to an "application" with versions, for which support packages and patches are
delivered. Delivery units are not an equivalent to development classes; they are similar to Software Component
Versions in the Product Availability Matrix (PAM). You can also use the delivery unit to transport repository
content between SAP HANA systems, for example, between development systems or between development and
productive systems.
Note
A governance plan exists for the naming of delivery units and the package structure.
A delivery unit is identified by the following key properties:
Vendor
Used primarily to define the identity of the company producing the software component to deliver, for
example, "sap". However, vendor can also be used to specify any customer implementing SAP HANA. To
create a delivery unit, it is a prerequisite to maintain a vendor in your system.
Name
Used to identify the software component to be delivered
Although a vendor might provide multiple versions of a delivery unit in parallel, only one version of a delivery unit
can be installed in a HANA system at any one time. In addition, duplicate names are not allowed for delivery units.
SAP HANA treats delivery units and their versions as software components and software-component versions.
Software components and software-component versions are usually shipped as part of a product (product
version).
Note
If you try to create a delivery unit using a name that already exists in your system, you will receive a validation
error. A validation error also occurs if the check for a valid vendor ID does not find a suitable entry.
14.2 The SAP HANA Delivery-Unit Lifecycle
In SAP HANA, lifecycle management (LCM) includes all the activities you need to plan and perform to ensure that
the software components you develop for SAP HANA Application Services are produced and shipped in a
regulated way that meets the requirements laid out for the SAP HANA platform.
For example, you need to plan, manage, and maintain the application versions you want to ship, and you need to
ensure that a stable plan exists to provide support for existing and all future versions of your application.
440
PUBLI C

Implementing Lifecycle Management
SAP HANA lifecycle management uses the delivery unit (DU) is a vehicle to ship one or more software
components from SAP (or a partner) to a customer. Delivery units correspond to an "application" with versions,
for which support packages and patches are delivered. You can also use the delivery unit to transport content
between SAP HANA systems, for example, between development systems or between development and
productive systems.
Note
Delivery units are not to be confused with development classes; delivery units are similar to software
component versions in the Product Availability Matrix (PAM).
The Application-Development Lifecycle
Application development on SAP HANA requires a server-centric lifecycle for design-time objects, which are the
development artifacts that you store in the SAP HANA repository. As an application developer, you check out
design-time content from the repository and edit a copy of the checked-out artifact in the local file system on your
personal computer (PC). The following steps provide a brief, high-level overview of the development lifecycle for
design-time content:
1. Check out design-time content.
Check out the package containing the design-time artifacts you want to work on (if the package already
exists).
2. Edit the design-time content.
Edit the copies of the design-time artifacts, which are stored in your SAP HANA repository "workspace" on
your local file system; the local copies of the design-time artifacts are created during the checkout process.
3. Commit changes to design-time content.
Committing the changes you have made to the design-time artifacts creates new versions of the artifacts in
the SAP HANA repository. Note that identical (unchanged) versions of a file are not committed.
4. Activate changes to design-time content.
Activating the changes you have made to the design-time artifacts makes these changes available to
applications and other users, creating runtime objects where necessary.
14.3 Exporting Delivery Units
Exporting a delivery unit (with all packages and dependencies) is the mechanism SAP HANA uses to enable the
transfer of delivery units between systems, for example, to transport application content between development
systems of from a development system to a consolidation system.
The SAP HANA studio includes a mechanism that enables you to export and import delivery units. You can use the
export feature to create an image of a delivery unit, which you can store either on the SAP HANA server or in the
file system of a selected client for future import to another system. You can also export development objects as
an archive that you can send to SAP support or share with other members of the development team.
To export a delivery unit, perform the following steps:
PUBLI C

Ensure that the SAP HANA studio is connected to the SAP HANA system hosting the delivery unit you want to
export. If this is not the case, choose Select System... and change the system.
2. In the File menu, choose Export....
3. Select the export destination.
The export destination enables you to specify the purpose for which you want to export the packages that
make up the export archive. The option you choose determines the contents of the export archive. In most
cases, you would choose the Delivery Unit option, which includes everything you need to transport the
application-development structure to another system . However, you can choose from the following
additional export destination options:
Developer Mode
Includes most objects and data, but does not include translatable text bundles
SAP Support Mode
Includes only the active version of objects and any data required by SAP Support to reproduce an issue
you have encountered and reported
Note
By default, the package is saved in the directory: /usr/sap/<SID>/HDB<clientNr>/backup/
Specify the Delivery Unit option, for example, SAP HANA Content > Delivery Unit
4. Select the source system; the system you want to export the delivery unit from.
In the Systems for export screen area of the Delivery Unit dialog, select the SAP HANA system hosting the
repository in which the delivery unit is located that you want to export and choose Next to confirm the
selection..
5. Select the delivery unit to export.
For example, HANA_XS_DOCS
6. Filter the contents of the export package according to a specified time stamp.
This step is optional. You can choose to export a complete delivery unit, or only those parts of a delivery unit
that were changed during a specified time interval, which you define in the Filter by Time screen area.
7. Select a location for the exported delivery unit:
a) Server
The SAP HANA studio automatically selects a file location to export the delivery unit to.
b) Client
Browse to the location in the local file-system where you want to store the exported delivery unit.
Note
Some export operations (for example, SAP Support Mode) do not allow you to change the export
location.
8. Choose Next to confirm the settings.
9. Choose Finish to start the export operation.
10. Confirm the export operation succeeded.
You can browse to the location in the file system and examine the package contents in a file explorer to make
sure the package contains the required content.
442
PUBLI C

Note
For more information about a specific export operation, right-click the specified export in the Job Log view
and choose Open Job Details in the context-sensitive popup menu.
14.4 Importing Delivery Units
Importing a delivery-unit archive is the mechanism SAP HANA uses to enable you to transport delivery units
between systems, for example, to upload a delivery unit that was exported from another development system or
saved as an archive on a local file system.
The SAP HANA studio includes a mechanism that enables you to export and import delivery units. To import a
delivery unit, perform the following steps:
Ensure that the SAP HANA studio is connected to the SAP HANA system hosting the delivery unit you want to
export. If this is not the case, choose Select System... and change the system.
2. In the File menu, choose Import...
3. Select the import mode. The choice you make determines the information you need to enter in the next steps.
The import mode enables you to specify the purpose for which you want to import the packages that make up
the import archive. In most cases, you would choose the Delivery Unit option, which includes everything you
need to transport the application-development structure from another system . However, you can choose
from the following additional import destination options:
Data from local file
Imports data from a comma-separated-list (CSV) file or a Microsoft Excel file stored on the local file
system into a new or an existing table (and schema).
Developer Mode
Includes most objects and data, but does not include translatable text bundles. The import operation
obeys the following rules regarding the contents of the imported archive:
If a package with the same name as the one to be imported already exists on the target system, but
the package languages differ, objects are imported with a warning message.
If a package with the same name as the one to be imported does not already exist on the target
system, then the respective package with its original language (fetched from the properties file) is
created.
For old objects where no related properties file exists, a package is created with its language set to
the logon language of the current import session.
Import SAP NetWeaver BW Models
Imports SAP NetWeaver Business Warehouse (SAP NetWeaver BW) models that are InfoCubes
optimized forSAP HANA, SAP HANA-optimized DataStore objects, and Query Snapshot InfoProviders to
the SAP HANA modeling environment.
Note
With this option, you must maintain connection details for the BW system you want to import from, for
example, host name, instance ID, client number and so on. You can test the connection before
continuing with the import operation.
PUBLI C

Specify the mode in which you want the data to be imported, for example, SAP HANA Content > Delivery
Unit .
4. Select the target system; the system you want to import the delivery unit into.
In the Systems for export screen area of the Delivery Unit dialog, select the SAP HANA system hosting the
repository into which you want to import the delivery unit and choose Next to confirm the selection.
5. Select the file containing the delivery unit you want to import.
a) Server
The SAP HANA studio automatically displays a list of available files stored in the default location; use the
drop-down list to choose the file you want to import.
b) Client
Browse to the location in the local file-system where the file you want to import is stored.
Before starting the import operation, the SAP HANA studio compares the contents of the delivery unit
selected for imported with the current status of the repository on the target system and displays the results in
the Object import simulation screen area. Hover the cursor over the icons in the Status column for a brief
description of the indicated status.
6. Choose Finish to start the import operation.
Use the Actions screen area to specify what action to take if an imported object already exists in the target
repository, for example:
Overwrite inactive versions
Overwrite inactive versions in your inactive workspace with objects from the imported delivery unit.
Activate objects
Activate objects imported from the delivery unit.
Bypass validation
Do not perform the client-side validation check on imported objects.
7. Check the status bar in the SAP HANA studio to follow the progress of the import operation.
You can also check progress in the Current tag page of the Job Log view.
8. Confirm the import operation succeeded.
14.5 Translating Delivery Units
SAP HANA includes features that enable you to translate package-related metadata texts, for example labels and
menu items in the user interface. The translation typically occurs on a separate dedicated system.
If you plan to translate the contents of the new delivery unit you create, you must maintain translation details. The
details include the connection to the translation system that the translation team use to store the localized
versions of the translated data. The text to be translated is uploaded from the SAP HANA repository to the
translation system and then reimported to the repository after the translated text are complete. As part of the
translation process for delivery units, you typically perform the following tasks
1. Maintain translation details for each delivery unit
Translation details includes both mandatory and optional information. Mandatory details to maintain include
the terminology domain, for example, Financial Accounting (FI), or Production Planning (PP) and the name
space for the text collection, which is used by the translation systems.
444
PUBLI C

2. Maintain the connection with the remote translation system
You need to be able to exchange the translation texts between SAP HANA and the translation system.
3. Upload translation data to the translation system
4. Download translated text to the SAP HANA repository
14.5.1 SAP HANA Delivery-Unit Translation Details
The SAP HANA repository includes features for translating package-related metadata texts. If you plan to
translate the contents of the new delivery unit you create, you must maintain translation details.
texts.
Note
Translation text can include the following elements: UI strings (menus, dialog titles), error messages, and
documentation, amongst other things. Package names and descriptions are also relevant for translation
purposes as are the names and descriptions of database views, tables, columns, and so on.
To make things easier for the translation team, you should maintain the following metadata for translation
purposes; the metadata must be maintained at the package level:
Terminology domain (mandatory)
The technical/product area with which your delivery unit (application) is associated, for example: AC
(Accounting General), FI (Financial Accounting), or PP (Production Planning and Control). This enables
translator teams to assign work to the most suitable experts, and allows the translators to make use of
existing terminology databases. The terminology domain is used by SAP internal translation systems.
Text collection (mandatory)
The name space for the text objects in the translation system. Technically, you can use any value, but the
text-collection name space is similar to the package namespace used in SAP HANA, for example:
"sap.hana.xs.docs". The value assigned to the Text collection field is used by SAP internal translation systems.
Hint
An optional field that customers and partners can use to pass on notes concerning the text objects tagged for
translation.
Text status
An optional field that customers and partners can use to pass on information indicating the current stage in
the translation-review process of the text objects, for example: "New", "In review", "Reviewed", "Approved",
"Published", and so on.
14.5.2 Maintaining Translation Details
texts.
PUBLI C

If you plan to translate the contents of a delivery unit, you must maintain translation details. You can maintain the
translation details when you create a new delivery unit or for existing delivery units. It is mandatory to maintain
the following translation-related details:
Terminology domain
The technical/product area with which your delivery unit (application) is associated, for example: AC
(Accounting General), FI, or PP.
Text collection
The name space for the text objects in the translation system, for example, "sap.hana.xs.docs".
1. In the SAP HANA studio, start the Modeler perspective.
2. In the Navigator view, select the SAP HANA instance where the repository containing the delivery unit is
located.
3. Expand the Content node until you can see the package whose translation details you want to maintain.
4. Maintain the translation details for the selected delivery unit, as follows:
a) Right-click the package whose translation details you want to maintain and choose Edit.
b) Choose Translation >> to display the Translation Details screen area at the bottom of the Edit Package
Details dialog.
c) Select a Terminology Domain from the drop-down list.
d) Enter a Text collection name.
e) Save the changes to the translation details.
5. Save the changes to the delivery unit.
6. Commit and activate the changes in the SAP HANA repository.
446
PUBLI C

15 Using Database Client Interfaces
SAP HANA provides client interfaces for connecting applications so they can retrieve and update data.
SAP HANA provides many methods for applications to access its data. For example, a spreadsheet application
could use ODBO to consume analytic views and enable users to create pivot tables, or a Web application could use
OData interfaces to access the data and display it.
SAP HANA can expose data with the following interfaces:
Client interfaces, which are available as long as the SAP HANA clients are installed:
ODBC
ODBO
JDBC
Web-based interfaces, which must be defined by the application developer, who determines what data to
expose and to whom:
OData
XMLA
Server-Side JavaScript
Related Links
SAP HANA Database - Client Installation and Update Guide
Defining Web-based Data Access [page 191]
SAP HANA extended application services (SAP HANA XS) provide applications and application developers with
15.1 Connecting via ODBC
SAP HANA provides an ODBC driver for connecting applications to the database.
Note
Make sure to use the 32-bit ODBC driver for 32-bit applications, and the 64-bit driver for 64-bit applications.
1. Install the ODBC driver. The driver is installed as part of the SAP HANA client installation.
2. Write code to create a connection to the database. You can use one of the following methods:
Connection String (SQLDriverConnect): Use a connection string in the form:
DRIVER={<driver>};UID=<username>;PWD=<password>; SERVERNODE=<server>:<port>;
<driver> should be one of the following:
HDBODBC: For 64-bit applications
HDBODBC32: For 32-bit applications
<port> should be 3<instance number>15, for example, 30015, if the instance is 00.
Using Database Client Interfaces
PUBLI C

The following is an example of a connection string:
DRIVER={HDBODBC};UID=myUser;PWD=myPassword;SERVERNODE=myServer:30015;
DSN (SQLConnect): Create a data source by running the odbcad32.exe tool. You can run this via a
command line or via the Control Panel /Administrative Tools /Data Sources (ODBC) .
Note
The application is located in the system32 directory. To create a data source for the 32-bit driver
on a 64-bit Microsoft Windows machine, run the tool from the SysWOW64 directory.
In the DSN tab, choose Add, select the SAP HANA driver, and select Finish. The following appears:
You cannot enter a user name and password here. For the server and port, you can either enter a key
created using the SAP HANA user store (which defines the server, port, user name and password), or you
can enter a server and port (e.g., myServer:30015). If you enter a server and port, then the application
must supply the user name and password when connecting.
Related Links
Using the User Store [page 448]
15.1.1 Using the User Store
The SAP HANA user store enables you to store connection information for connecting to an SAP HANA system.
Instead of entering connection information each time you make a connection, you store the information, assign it
a key, and use this key when making connections.
This makes it easier to move between systems (for example, when executing SQL from the command line), and
also keeps connection information, including user names and passwords, in a secure place.
The SAP HANA user store is part of the client installation.
1. In a command line, run the following:
hdbuserstore.exe set <key> <server>:<port> <user> <password>
448
PUBLI C

The server, port, user name and password are now stored in the user store. The key is a string you use to refer
to this set of connection information.
2. Use the key to reference a connection to a particular system. Always precede the key with an @, for example,
@mykey.
Use the key in the following ways:
In the connection string, use the key for the SERVERNODE parameter, and do not include a user name and
password, for example:
SERVERNODE={@mykey};DRIVER={hdbodbc};
In creating a DSN, enter the key (for example, @mykey) for the server and port.
In testing your connection (by running odbcreg -t hdbodbc), use the key (for example, @mykey) for
the server and port. If the connection is successful, you get something like this:
The response includes a sample connection string using the key.
Related Links
Testing the ODBC Installation [page 449]
You can test the installation of the ODBC driver and your ability to connect by using the odbcreg tool, which is
part of the ODBC installation.
15.1.2 Testing the ODBC Installation
You can test the installation of the ODBC driver and your ability to connect by using the odbcreg tool, which is
part of the ODBC installation.
1. Open a command window.
2. Start the odbcreg tool by enter a command in the form: odbcreg -t hdbcodbc (for 64-bit driver) or
odbcreg32 -t hdbcodbc32 (for 32-bit driver).
If the driver is installed properly, you should get the ODBC login screen.
Note
PUBLI C

You can also run the command odbcreg -g or odbcreg32 -g to get a list of installed drivers. The SAP
HANA driver is called HDBODBC.
3. Test your ability to connect by entering connection information (system, port, user name and password) and
select OK. The tool closes and the results of the test are printed in the command window.
You can also run odbcreg -g to get a list of installed ODBC drivers, and check to make sure that the SAP HANA
driver (either HDBODBC or HDBODBC32) is installed.
15.2 Connecting via JDBC
SAP HANA provides a JDBC driver for connecting Java applications to the database.
1. Install the JDBC driver.
The driver (ngdbc.jar) is installed as part of the SAP HANA client installation and is located at:
C:\Program Files\sap\hdbclient\ on Microsoft Windows platforms
/usr/sap/hdbclient/ on Linux and UNIX platforms
2. Add ngdbc.jar to your classpath.
3. If you are on a version of Java earlier than Java 6, load the JDBC driver class, which is called
com.sap.db.jdbc.Driver.
4. Write Java code to create a connection to the database and execute SQL commands. Use a connection string
in the form of jdbc:sap://<server>:<port>[/?<options>]. For example:
jdbc:sap://myServer:30015/?autocommit=false
The port should be 3<instance number>15, for example, 30015, if the instance is 00.
450
PUBLI C

You can specify one or more failover servers by adding additional hosts, as in the following example:
jdbc:sap://myServer:30015,failover1:30015,failover2:30015/?autocommit=false
Example
The following is an example of connecting to an SAP HANA server called myhdb, which was installed as instance
07, with user name myname and password mysecret. Make sure to change these for your system, and add the
JDBC driver (ngdbc.jar) to your classpath.
import java.sql.*;
public class jdemo {
public static void main(String[] argv) {
Connection connection = null;
try {
connection = DriverManager.getConnection(
"jdbc:sap://myhdb:30715/?
autocommit=false",myname,mysecret);
} catch (SQLException e) {
System.err.println("Connection Failed. User/Passwd Error?");
return;
}
if (connection != null) {
try {
System.out.println("Connection to HANA successful!");
Statement stmt = connection.createStatement();
ResultSet resultSet = stmt.executeQuery("Select 'hello world' from
dummy");
resultSet.next();
String hello = resultSet.getString(1);
System.out.println(hello);
} catch (SQLException e) {
System.err.println("Query failed!");
}
}
}
}
Related Links
15.2.1 Tracing JDBC Connections
You can activate the JDBC trace to find errors while your application is connected to a database via JDBC.
You must be logged on as the operating system user who started (or will start) the JDBC application.
Note
You always activate the JDBC trace for all JDBC applications that the current operating system user has
started.
Configuration changes have an effect on all JDBC applications that the current operating system user has
started.
When the JDBC trace is activated, the JDBC driver logs on the client the following information:
PUBLI C

JDBC API calls called by the JDBC application
JDBC API call parameters
Executed SQL statements and their results
The location of the trace file is determined by the trace options.
15.2.1.1 Tracing via GUI
You can start tracing by running the tracing configuration tool that includes a graphical user interface (GUI).
Tracing via the GUI enables you to start and configure tracing without stopping and restarting your application
that is connected via JDBC.
1. On the client, enter the following command on the command line: java -jar <installation_path>
\ngdbc.jar.
2. Select Trace enabled.
3. Select and modify the tracing options.
4. Select OK.
15.2.1.2 Tracing via Command Line
You can start tracing by running the tracing configuration tool and sending commands via the command line.
Tracing via the command line enables you to start and configure tracing without stopping and restarting your
application that is connected via JDBC.
1. Display the current configuration by running the command java -jar <installation_path>
\ngdbc.jar.
2. Select trace options by running the command java -jar <installation_path>\ngdbc.jar
<option>.
3. Start tracing by running the command java -jar <installation_path>\ngdbc.jar TRACE ON.
Related Links
Trace Options
15.2.1.3 Tracing via Connection String
You can start tracing by adding an option in the connection string when creating a JDBC connection.
Tracing via the connection string requires you to stop and restart your application that is making the JDBC
connection. Also, with the connection string, you cannot turn off tracing or set any options except setting the
trace filename.
Add the trace option to the connection when creating a JDBC connection.
452
PUBLI C

Here is an example connection string that starts tracing:
jdbc:sap://localhost:30015/?autocommit=false&trace=traceFile.txt
The location of the trace file is determined by the trace options.
15.2.1.4 Trace Options
Options when enabling JDBC tracing.
The first column shows the field name in the GUI-based tracing configuration tool, and the second column shows
the command to enter when using the command-line tool.
Table 18: Trace Options
Option Command Line Option Description
Trace enabled TRACE ON | OFF Starts and stops tracing
Trace file folder No command-line option. The folder can be specified
with the FILENAME option.
Sets the directory where the system
writes the trace files.
When no folder is specified, the files
are saved in the working directory of
the application.
Trace file name TRACE FILENAME [<path>]<file_name> Sets the name of the trace file.
The system assigns each trace file
an additional unique ID, so the file
name is:
<file_name>_<id>.prt
The default file name is jdbctrace.
Limit file size TRACE SIZE <size> [KB|MB|GB]
To remove the size limit, use the following option:
TRACE SIZE UNLIMITED
Limits the size of each trace file.
Stop on error TRACE STOP ON ERROR <error_code>
To no longer stop on the specified error, use the
following option:
TRACE STOP ON ERROR OFF
Stops writing the JDBC trace when
the specified error code occurs.
15.2.2 Valid Java-to-SQL Conversions
SAP HANA allows each Java object to be converted to specific SQL types using the JDBC method
PreparedStatement.setObject or RowSet.setObject.
PUBLI C

Some conversions may fail at runtime if the value passed is invalid.
15.3 Connecting via ODBO
SAP HANA provides an ODBO driver for connecting applications to the database and executing MDX statements.
1. Install the ODBO driver. The driver is installed as part of the SAP HANA client installation.
2. Specify in your client the provider name: SAPNewDBMDXProvider
3. Create a connection string in the form of:
<host of HANA>;User ID=<your user>;Password=<your
password;SFC_USE_ROWCACHE=true;SFC_INSTANCE_NUM=<instance number>
For example:
localhost;User
ID=system;Password=mypassword;SFC_USE_ROWCACHE=true;SFC_INSTANCE_NUM=00
If the server instance is 00, then you can omit the SFC_INSTANCE_NUM parameter.
SFC_USE_ROWCACHE is optional. It enables backward and forward navigation through rowsets.
Related Links
15.3.1 Connecting with Microsoft Excel
You can use Microsoft Excel and its PivotTables as a practical way to access and analyze SAP HANA data,
connecting with ODBO.
SAP HANA supports Microsoft Excel 2007 and 2010.
454
PUBLI C

1. Start the Data Connection Wizard, and select Other/Advanced as the type of data source.
2. Select the SAP HANA MDX Provider as the OLE DB Provider.
3. In the Connection tab of the Data Link Properties window, enter the SAP HANA server name, instance number,
user name and password.
PUBLI C

4. Select the cube from which you want to import data. SAP HANA analytic and calculation views are exposed as
cubes.
456
PUBLI C

SAP HANA supports the following Microsoft Excel features:
Drilling down
Selection filtering
Top/bottom filters
Report filters
Member properties
Refresh cube
Convert PivotTable into formulas
Server formatting
Pre-modeled calculated members
Show/hide fields
Enhanced value and label filters
Insert slicer
Text search in report filter
PivotTable filter
Creation of named sets
15.3.2 Multidimensional Expressions (MDX)
Multidimensional Expressions (MDX) is a language you can use to query multidimensional data stored in OLAP
cubes.
PUBLI C

MDX uses a multidimensional data model to enable navigation in multiple dimensions, levels, and up and down a
hierarchy. With MDX, you can access pre-computed aggregates at specified positions (levels or members) in a
hierarchy.
Note
MDX is an open standard. However, SAP has developed extensions to MDX that are designed to enable faster
and more efficient access to multidimensional data, for example, to serve specific SAP HANA application
requirements and to optimize the resultset for SAP HANA clients.
MDX is implicitly a hierarchy-based paradigm. All members of all dimensions must belong to a hierarchy. Even if
you do not explicitly create hierarchies in your SAP HANA data model, the SAP HANA modeler implicitly
generates default hierarchies for each dimension. All identifiers that are used to uniquely identify hierarchies,
levels and members in MDX statements (and metadata requests) embed the hierarchy name within the identifier.
In SAP HANA, the standard use of MDX is to access SAP HANA models (for example, analytical and attribute
views) that have been designed, validated and activated in the modeler in the SAP HANA studio. The studio
provides a graphical design environment that enables detailed control over all aspects of the model and its
language-context-sensitive runtime representation to users.
MDX in SAP HANA uses a runtime cube model, which usually consists of an analytical (or calculation) view that
represents data in which dimensions are modeled as attribute views. You can use the analytical view to specify
whether a given attribute is intended for display purposes only or for aggregation. The key attributes of attribute
views are linked to private attributes in an analytical view in order to connect the entities. One benefit of MDX in
SAP HANA is the native support of hierarchies defined for attribute views.
Note
MDX in SAP HANA includes native support of hierarchies defined for attribute views. SAP HANA supports level-
based and parent-child hierarchies and both types of hierarchies are accessible with MDX.
SAP HANA supports the use of variables in MDX queries; the variables are a SAP-specific enhancement to
standard MDX syntax. You can specify values for all mandatory variables that are defined in SAP HANA studio to
various modeling entities. The following example illustrates how to declare SAP HANA variables and their values:
MDX
Select
From [MINI_C1_VAR]
Where [Measures].[M2_1_M3_CONV]
[VAR_TARGET_CURRENCY] including 'EUR',
15.3.3 MDX Functions
MDX in SAP HANA supports a variety of standard MDX functions.
458
PUBLI C

Table 19: Supported MDX Functions
Aggregate Returns a calculated value using the appropriate aggregate function, based on the
aggregation type of the member.
Ancestor Returns the ancestor of a member at a specified level or at a specific distance away
in the hierarchy.
Ancestors Returns a set of all ancestors of a member at a specified level or at a specific
Ascendants Returns the set of the ascendants of the member.
Avg Returns the average value of a numeric expression evaluated over a set.
BottomCount Returns a specified number of items from the bottom of a set, optionally sorting
the set first.
Children Returns the children of a member.
ClosingPeriod Returns the last sibling among the descendants of a member at a specified level.
Count Counts the number of members in the tuple.
Cousin Returns the child member with the same relative position under a parent member
as the specified child member.
Crossjoin Returns the cross product of two sets.
CurrentMember Returns the current member along a hierarchy.
DefaultMember Returns the default member of a hierarchy.
Descendants Returns the set of descendants of a member at a specified level or at a specific
Dimension Returns the hierarchy that contains a specified member or level.
Dimensions Returns a hierarchy specified by a numeric or string expression.
Distinct Returns a set, removing duplicate tuples from a specified set.
DistinctCount Returns the number of distinct tuples in a set.
DrillDownLevel Drills down the members of a set one level below the lowest level represented in the
set, or to one level below an optional level of a member represented in the set.
DrillDownLevelBottom Drills down the members of a specified count of bottom members of a set, at a
specified level, to one level below.
DrillDownLevelTop Drills down a specified count of top members of a set, at a specified level, to one
level below.
DrillDownMember Drills down the members in a specified set that are present in a second specified
set.
DrillDownMemberBottom Drills down the members in a specified set that are present in a second specified
set, limiting the result set to a specified number of bottommost members.
DrillDownMemberTop Drills down the members in a specified set that are present in a second specified
set, limiting the result set to a specified number of topmost members.
PUBLI C

DrillUpLevel Drills up the members of a set that are below a specified level.
DrillUpmember Drills up the members in a specified set that are present in a second specified set.
Except Finds the difference between two sets, optionally retaining duplicates.
Filter Returns the set resulting from filtering a set based on a search condition.
FirstChild Returns the first child of a specified member.
FirstSibling Returns the first child of the parent of a specified member.
Generate Applies a set to each member of another set, and then joins the resulting sets by
union. Alternatively, this function returns a concatenated string created by
evaluating a string expression over a set.
Head Returns the first specified number of elements in a set.
Hierarchize Orders the members of a specified set in a hierarchy in natural or, optionally, post-
natural order.
Hierarchy
Instr The InStr function finds the starting location of a substring within a specified string.
Intersect Returns the intersection of two sets, optionally retaining duplicates.
IsAncestor Returns true if the first member specified is an ancestor of the second member
IsGeneration Returns true if the member specified is a leaf, else returns false.
IsLeaf Returns true if the first member specified is an ancestor of the second member
IsSibling Returns true if the first member specified is an sibling of the second member
Item If an integer is specified, the Item function returns the tuple that is in the zero-
based position specified by Index.
IIF Returns one of values determined by a logical test.
Lag Returns the member that is a specified number of positions prior to a specified
LastChild Returns the last child of a specified member.
LastPeriods Returns a set of members prior to and including a specified member.
LastSibling Returns the last child of the parent of a specified member.
Lead Returns the member that is a specified number of positions following a specified
Leaves If a dimension name is specified, returns a set that contains the leaf members of
the key attribute for the specified dimension.If a dimension name is not specified,
the function returns a set that contains the leaf members of the entire cube.
460
PUBLI C

Left The Left function returns a string of a specified number of characters from the left
side (beginning) of a specified string.
Level Returns the level of a member.
Levels Returns the level whose zero-based position in a dimension is specified by a
numeric expression.
Max Returns the maximum value of a numeric expression evaluated over a set.
Member_caption Returns the caption of a member
Members Returns the set of all members in a specified hierarchy.
MembersAscendantsDesc
endants
Returns the set of specified members in a given hierarchy.
Mid The Mid function returns a substring of a string argument.
Min Returns the minimum value of a numeric expression evaluated over a set
MTD Returns a set of members from the Month level in a Time dimension starting with
Name Returns the name of a specified hierarchy or member.
NextMember Returns the next member in the level that contains a specified member.
NOT Performs a logical negation on a numeric expression.
OpeningPeriod Returns the first sibling among the descendants of a specified level, optionally at a
specified member.
OR Performs a logical disjunction on two numeric expressions.
Ordinal Returns the zero-based ordinal value associated with a specified level.
ParallelPeriod Returns a member from a prior period in the same relative position as a specified
member.
Parent Returns the parent of a specified member.
PeriodsToDate Returns a set of members (periods) from a specified level starting with the first
member and ending with a specified member.
PrevMember Returns the previous member in the level that contains a specified member.
Properties Returns a string containing the value of the specified member property.
QTD Returns a set of members from the Quarter level in a Time dimension starting with
Range Performs a set operation that returns a naturally ordered set, with the two
specified members as endpoints, and all members between the two specified
members included as members of the set
Right The Right function returns a string of a specified number of characters from the
right side (end) of a specified string.
Siblings Returns the set of siblings of a specified member, including the member itself.
PUBLI C

StrToMember Returns a member from a string expression in MDX format.
StrToSet Constructs a set from a specified string expression in MDX format.
StrToTuple Constructs a tuple from a specified string expression in MDX format.
StrToValue Returns a value from a string expression
Subset Returns a subset of tuples from a specified set.
Sum Returns the sum of a numeric expression evaluated over a specified set.
Tail Returns the last specified number of elements in a set.
TopCount Returns a specified number of items from the topmost members of a specified set,
optionally ordering the set first.
Union Performs a set operation that returns a union of two sets, removing duplicate
members.
UniqueName Returns the unique name of a specified hierarchy.
WTD Returns a set of members from the Week level in a Time dimension starting with
YTD Returns a set of members from the Year level in a Time dimension starting with the
first period and ending with a specified member.
15.3.4 MDX Extensions
SAP HANA supports several extensions to the MDX language, for example, additional predefined functions and
support for variables.
15.3.4.1 Sibling_Ordinal Intrinsic Property
The object Member includes a property called Sibling_Ordinal, that is equal to the 0-based position of the member
within its siblings.
Example
WITH
MEMBER [Measures].[Termination Rate] AS
[Measures].[NET_SALES] / [Measures].[BILLED_QUANTITY]
SELECT
{
[Measures].[NET_SALES],
[Measures].[BILLED_QUANTITY],
[Measures].[Termination Rate]
} ON COLUMNS,
Descendants
(
[MDX_TEST_10_DISTRIBUTION_CHANNEL].[MDX_TEST_10_DISTRIBUTION_CHANNEL].[All].
462
PUBLI C

[(all)],
1,
SELF_AND_BEFORE
)
DIMENSION PROPERTIES SIBLING_ORDINAL ON ROWS
FROM MDX_TEST_10_ITELO_SALES_DATA
15.3.4.2 MembersAscendantsDescendants Function
SAP HANA includes a new function called MembersAscendantsDescendants that enables you to get, for
example, all ascendants and descendants of a specific member. This function improves on the standard MDX
functions Ascendants and Descendants.
The function can be called as follows:
MembersAscendantsDescendants (<set>, <flag>)
set: A set of members from a single hierarchy
flag: Indicates which related members to return, and can be one of the following:
MEMBERS_AND_ASCENDANTS_AND_DESCENDANTS
MEMBERS_AND_ASCENDANTS
MEMBERS_AND_DESCENDANTS
ASCENDANTS_AND_DESCENDANTS
ONLY_ASCENDANTS
ONLY_DESCENDANTS
Example
SELECT
ON COLUMNS,
NON EMPTY
[JUICE_TIME].[TimeHier].[QUARTER].[4], MEMBERS_AND_ASCENDANTS_AND_DESCENDANTS )) }
ON ROWS
FROM [JUICE]
Example
SELECT
ON COLUMNS,
NON EMPTY
[JUICE_TIME].[TimeHier].[QUARTER].[4], ONLY_ASCENDANTS )) }
ON ROWS
FROM [JUICE]
PUBLI C

15.3.4.3 Variables
An MDX SELECT statement in SAP HANA enables you to send values for variables defined within modeling views.
Analytic and calculation views can contain variables, that can be bound to specific attributes. When calling the
view, you can send values for those variables. These variables can be used, for example, to filter the results.
SAP HANA supports an extension to MDX whereby you can pass values for variables defined in views by adding a
SAP Variables clause in your select statement. Here is the syntax for a Select statement:
<select_statement>: [WITH <formula_specification> ]
SELECT [<axis_specification>[,<axis_specification>...]]
FROM <cube_specification>
[WHERE <slicer_specification>
SAP VARIABLES: <sap_variable> [[,] <sap_variable>]]
<sap_variable>: <variable_name> <sign> [<option>] <variable_value>
<sign>: INCLUDING | EXCLUDING
<option>: = | > | >= | < | <= | <>
<variable_value>: <unique_member_name>
| <unsigned_numeric_literal>
| <string_value_expression>
| <member> : <member>
| <character_string_literal> : <character_string_literal>
| <unsigned_numeric_literal> : <unsigned_numeric_literal>
Example
The following specifies a single value for variables VAR_KAT, VAR_K2, and VAR_TARGET_CURRENCY.
SELECT
FROM [MINI_C1_VAR]
WHERE [Measures].[M2_1_M3_CONV]
[VAR_TARGET_CURRENCY] including 'EUR'
Example
The following specifies an interval for variable VAR_K2.
SELECT NON EMPTY
{
[K2].[K2].Members
}ON ROWS
FROM [MINI_C1_VAR_SIMPLE]
WHERE [Measures].[M3_CONV]
SAP VARIABLES [VAR_K2] including [K2].[K2].&[122]:[K2].[K2].&[221]
Metadata on Variables in Views
SAP HANA includes the following set of tables that contain information about the variables defined for views:
BIMC_VARIABLE
464
PUBLI C

BIMC_VARIABLE_ASSIGNMENT
BIMC_VARIABLE_VALUE
BIMC_VARIABLE_ODBO (virtual table)
The tables enable, for example, an application to retrieve the variables defined for a view and create a user
interface so the user can enter values.
PUBLI C

16 SAP HANA Developer References
The SAP HANA Developer's Guide comes with the following language, API and other references.
Name Link
SAP HANA SQL Reference http://help.sap.com/hana/html/sqlmain.html
SAP HANA SQLScript Reference http://help.sap.com/hana/hana_dev_sqlscript_en.pdf
SAP HANA System Tables and Monitoring Views
Reference
http://help.sap.com/hana/html/monitor_views.html
SAP HANA R Integration Guide http://help.sap.com/hana/hana_dev_r_emb_en.pdf
SAP HANA Business Function Library (BFL) Reference http://help.sap.com/hana/hana_dev_bfl_en.pdf
SAP HANA Predictive Analysis Library (PAL) Reference http://help.sap.com/hana/hana_dev_pal_en.pdf
SAP HANA XS JavaScript Reference http://help.sap.com/hana/jsapi/index.html
466
PUBLI C

SAP HANA Developer References
www.sap.com/contactsap

No part of this publication may be reproduced or transmitted in any
form or for any purpose without the express permission of SAP AG.
The information contained herein may be changed without prior
notice.
Some software products marketed by SAP AG and its distributors
contain proprietary software components of other software
vendors. National product specifications may vary.
These materials are provided by SAP AG and its affiliated
companies ("SAP Group") for informational purposes only, without
representation or warranty of any kind, and SAP Group shall not be
liable for errors or omissions with respect to the materials. The only
warranties for SAP Group products and services are those that are
set forth in the express warranty statements accompanying such
products and services, if any. Nothing herein should be construed as
constituting an additional warranty.
SAP and other SAP products and services mentioned herein as well
as their respective logos are trademarks or registered trademarks
of SAP AG in Germany and other countries.
Please see http://www.sap.com/corporate-en/legal/copyright/
index.epx for additional trademark information and notices.

Hana Dev en

Uploaded by

Copyright:

Available Formats

Hana Dev en

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hana Dev en

Uploaded by

Copyright:

Available Formats

What is SAP HANA?

What is SAP HANA?

What are the main components of SAP HANA?

What are the main components of SAP HANA?

PUBLIC

SAP HANA Appliance Software SPS 05

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 3

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 5

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 7

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 9

2013 SAP AG or an SAP affiliate company. All rights reserved.

. It explains how to use the SAP HANA development tools to

2013 SAP AG or an SAP affiliate company. All rights reserved. 11

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 13

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 15

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 17

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 19

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 21

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 23

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 25

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 27

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 29

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 31

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 33

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 35

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 37

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 39

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 41

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 43

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 45

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 47

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 49

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 51

2013 SAP AG or an SAP affiliate company. All rights reserved.

2013 SAP AG or an SAP affiliate company. All rights reserved. 53