Streaming newline-delimited JSON (NDJSON) parser and encoder for ReactPHP.

NDJSON can be used to store multiple JSON records in a file to store any kind of (uniform) structured data, such as a list of user objects or log entries. It uses a simple newline character between each individual record and as such can be both used for efficient persistence and simple append-style operations. This also allows it to be used in a streaming context, such as a simple inter-process commmunication (IPC) protocol or for a remote procedure call (RPC) mechanism. This library provides a simple streaming API to process very large NDJSON files with thousands or even millions of rows efficiently without having to load the whole file into memory at once.

• Standard interfaces - Allows easy integration with existing higher-level components by implementing ReactPHP's standard streaming interfaces.
• Lightweight, SOLID design - Provides a thin abstraction that is just good enough and does not get in your way. Builds on top of well-tested components and well-established concepts instead of reinventing the wheel.
• Good test coverage - Comes with an automated tests suite and is regularly tested in the real world.

Table of contents

• Support us
• NDJSON format
• Usage
  • Decoder
  • Encoder
• Install
• Tests
• License
• More

We invest a lot of time developing, maintaining and updating our awesome open-source projects. You can help us sustain this high-quality of our work by becoming a sponsor on GitHub. Sponsors get numerous benefits in return, see our sponsoring page for details.

Let's take these projects to the next level together! 🚀

NDJSON ("Newline-Delimited JSON" or sometimes referred to as "JSON lines") is a very simple text-based format for storing a large number of records, such as a list of user records or log entries.

{"name":"Alice","age":30,"comment":"Yes, I like cheese"}
{"name":"Bob","age":50,"comment":"Hello\nWorld!"}

If you understand JSON and you're now looking at this newline-delimited JSON for the first time, you should already know everything you need to know to understand NDJSON: As the name implies, this format essentially consists of individual lines where each individual line is any valid JSON text and each line is delimited with a newline character.

This example uses a list of user objects where each user has some arbitrary properties. This can easily be adjusted for many different use cases, such as storing for example products instead of users, assigning additional properties or having a significantly larger number of records. You can edit NDJSON files in any text editor or use them in a streaming context where individual records should be processed. Unlike normal JSON files, adding a new log entry to this NDJSON file does not require modification of this file's structure (note there's no "outer array" to be modified). This makes it a perfect fit for a streaming context, for line-oriented CLI tools (such as grep and others) or for a logging context where you want to append records at a later time. Additionally, this also allows it to be used in a streaming context, such as a simple inter-process commmunication (IPC) protocol or for a remote procedure call (RPC) mechanism.

The newline character at the end of each line allows for some really simple framing (detecting individual records). While each individual line is valid JSON, the complete file as a whole is technically no longer valid JSON, because it contains multiple JSON texts. This implies that for example calling PHP's json_decode() on this complete input would fail because it would try to parse multiple records at once. Likewise, using "pretty printing" JSON (JSON_PRETTY_PRINT) is not allowed because each JSON text is limited to exactly one line. On the other hand, values containing newline characters (such as the comment property in the above example) do not cause issues because each newline within a JSON string will be represented by a \n instead.

One common alternative to NDJSON would be Comma-Separated Values (CSV). If you want to process CSV files, you may want to take a look at the related project clue/reactphp-csv instead:

name,age,comment
Alice,30,"Yes, I like cheese"
Bob,50,"Hello
World!"

CSV may look slightly simpler, but this simplicity comes at a price. CSV is limited to untyped, two-dimensional data, so there's no standard way of storing any nested structures or to differentiate a boolean value from a string or integer. Field names are sometimes used, sometimes they're not (application-dependant). Inconsistent handling for fields that contain separators such as , or spaces or line breaks (see the comment field above) introduce additional complexity and its text encoding is usually undefined, Unicode (or UTF-8) is unlikely to be supported and CSV files often use ISO 8859-1 encoding or some variant (again application-dependant).

While NDJSON helps avoiding many of CSV's shortcomings, it is still a (relatively) young format while CSV files have been used in production systems for decades. This means that if you want to interface with an existing system, you may have to rely on the format that's already supported. If you're building a new system, using NDJSON is an excellent choice as it provides a flexible way to process individual records using a common text-based format that can include any kind of structured data.

The Decoder (parser) class can be used to make sure you only get back complete, valid JSON elements when reading from a stream. It wraps a given ReadableStreamInterface and exposes its data through the same interface, but emits the JSON elements as parsed values instead of just chunks of strings:

{"name":"test","active":true}
{"name":"hello w\u00f6rld","active":true}

$stdin = new ReadableResourceStream(STDIN, $loop);

$stream = new Decoder($stdin);

$stream->on('data', function ($data) {
    // data is a parsed element from the JSON stream
    // line 1: $data = (object)array('name' => 'test', 'active' => true);
    // line 2: $data = (object)array('name' => 'hello wörld', 'active' => true);
    var_dump($data);
});

ReactPHP's streams emit chunks of data strings and make no assumption about their lengths. These chunks do not necessarily represent complete JSON elements, as an element may be broken up into multiple chunks. This class reassembles these elements by buffering incomplete ones.

The Decoder supports the same optional parameters as the underlying json_decode() function. This means that, by default, JSON objects will be emitted as a stdClass. This behavior can be controlled through the optional constructor parameters:

$stream = new Decoder($stdin, true);

$stream->on('data', function ($data) {
    // JSON objects will be emitted as assoc arrays now
});

Additionally, the Decoder limits the maximum buffer size (maximum line length) to avoid buffer overflows due to malformed user input. Usually, there should be no need to change this value, unless you know you're dealing with some unreasonably long lines. It accepts an additional argument if you want to change this from the default of 64 KiB:

$stream = new Decoder($stdin, false, 512, 0, 64 * 1024);

If the underlying stream emits an error event or the plain stream contains any data that does not represent a valid NDJson stream, it will emit an error event and then close the input stream:

$stream->on('error', function (Exception $error) {
    // an error occured, stream will close next
});

If the underlying stream emits an end event, it will flush any incomplete data from the buffer, thus either possibly emitting a final data event followed by an end event on success or an error event for incomplete/invalid JSON data as above:

$stream->on('end', function () {
    // stream successfully ended, stream will close next
});

If either the underlying stream or the Decoder is closed, it will forward the close event:

$stream->on('close', function () {
    // stream closed
    // possibly after an "end" event or due to an "error" event
});

The close(): void method can be used to explicitly close the Decoder and its underlying stream:

$stream->close();

The pipe(WritableStreamInterface $dest, array $options = array(): WritableStreamInterface method can be used to forward all data to the given destination stream. Please note that the Decoder emits decoded/parsed data events, while many (most?) writable streams expect only data chunks:

$stream->pipe($logger);

For more details, see ReactPHP's ReadableStreamInterface.

The Encoder (serializer) class can be used to make sure anything you write to a stream ends up as valid JSON elements in the resulting NDJSON stream. It wraps a given WritableStreamInterface and accepts its data through the same interface, but handles any data as complete JSON elements instead of just chunks of strings:

$stdout = new WritableResourceStream(STDOUT, $loop);

$stream = new Encoder($stdout);

$stream->write(array('name' => 'test', 'active' => true));
$stream->write(array('name' => 'hello wörld', 'active' => true));

{"name":"test","active":true}
{"name":"hello w\u00f6rld","active":true}

The Encoder supports the same parameters as the underlying json_encode() function. This means that, by default, unicode characters will be escaped in the output. This behavior can be controlled through the optional constructor parameters:

$stream = new Encoder($stdout, JSON_UNESCAPED_SLASHES | JSON_UNESCAPED_UNICODE);

$stream->write('hello wörld');

"hello wörld"

Note that trying to pass the JSON_PRETTY_PRINT option will yield an InvalidArgumentException because it is not compatible with NDJSON.

If the underlying stream emits an error event or the given data contains any data that can not be represented as a valid NDJSON stream, it will emit an error event and then close the input stream:

$stream->on('error', function (Exception $error) {
    // an error occured, stream will close next
});

If either the underlying stream or the Encoder is closed, it will forward the close event:

$stream->on('close', function () {
    // stream closed
    // possibly after an "end" event or due to an "error" event
});

The end(mixed $data = null): void method can be used to optionally emit any final data and then soft-close the Encoder and its underlying stream:

$stream->end();

The close(): void method can be used to explicitly close the Encoder and its underlying stream:

$stream->close();

For more details, see ReactPHP's WritableStreamInterface.

The recommended way to install this library is through Composer. New to Composer?

This project follows SemVer. This will install the latest supported version:

$ composer require clue/ndjson-react:^1.2

See also the CHANGELOG for details about version upgrades.

This project aims to run on any platform and thus does not require any PHP extensions and supports running on legacy PHP 5.3 through current PHP 8+ and HHVM. It's highly recommended to use PHP 7+ for this project.

To run the test suite, you first need to clone this repo and then install all dependencies through Composer:

$ composer install

To run the test suite, go to the project root and run:

$ php vendor/bin/phpunit

This project is released under the permissive MIT license.

Did you know that I offer custom development services and issuing invoices for sponsorships of releases and for contributions? Contact me (@clue) for details.

• If you want to learn more about processing streams of data, refer to the documentation of the underlying react/stream component.

• If you want to process compressed NDJSON files (.ndjson.gz file extension), you may want to use clue/reactphp-zlib on the compressed input stream before passing the decompressed stream to the NDJSON decoder.

• If you want to create compressed NDJSON files (.ndjson.gz file extension), you may want to use clue/reactphp-zlib on the resulting NDJSON encoder output stream before passing the compressed stream to the file output stream.

• If you want to concurrently process the records from your NDJSON stream, you may want to use clue/reactphp-flux to concurrently process many (but not too many) records at once.

• If you want to process structured data in the more common text-based format, you may want to use clue/reactphp-csv to process Comma-Separated-Values (CSV) files (.csv file extension).