MS SQLServer jdbc PreparedStatement bug

Question

I want to insert rows into a table with 118 column. I experience that it works perfectly with mssql-jdbc-8.2.2.jre8.jar but unfortunately with any later issue (I testes all from 9.2.1 to 12.8.1 ) gives back java.sql.BatchUpdateException, with a note that the column length at the 118. position is invalid. It happens only if the 118. column is VARCHAR or NVARCHAR. No error in the case of FLOAT or other numeric types. So if anybody has already faced with this, or know any solution, please let me know. Here is my PreparedStatement's command:

insert into tt
(CUSTOMER_ID,Value_date,_default,J_termek1,J_termek2,J_termek1_appl_pont,J_termek1_beh_pont,J_termek2_appl_pont,J_termek2_beh_pont,J_t
ermek1_appl_intercept,J_termek1_appl_beta,J_termek1_beh_intercept,J_termek1_beh_beta,J_t
ermek1_appl_weight,J_termek1_beh_weight,J_REL_def_STD,J_macro_correction,J_termek2_appl_intercept,J_termek2_appl_beta,J_termek2_beh_in
tercept,J_termek2_beh_beta,J_termek2_appl_weight,J_termek2_beh_weight,J_termek1_weight,J_termek2_weight,L_termek1,L_
termek2,L_termek1_appl_pont,L_termek1_beh_pont,L_termek2_appl_pont,L_termek2_beh_pont,L_termek1_appl_intercept,L_termek1_appl_beta,L_t
ermek1_beh_intercept,L_termek1_beh_beta,L_termek1_appl_weight,L_termek1_beh_weight,L_REL_def_STD,L_macro_correction,
L_termek2_appl_intercept,L_termek2_appl_beta,L_termek2_beh_intercept,L_termek2_beh_beta,L_termek2_appl_weight,L_termek2_beh_weight,L_t
ermek1_weight,L_termek2_weight,S_termek1,S_termek2,S_termek1_appl_pont,S_termek1_beh_pont,S_termek2_appl_pont,S_term
ek2_beh_pont,S_termek1_appl_intercept,S_termek1_appl_beta,S_termek1_beh_intercept,S_termek1_beh_beta,S_termek1_appl_weight,S_termek1_b
eh_weight,S_REL_def_STD,S_macro_correction,S_termek2_appl_intercept,S_termek2_appl_beta,S_termek2_beh_intercept,S_te
rmek2_beh_beta,S_termek2_appl_weight,S_termek2_beh_weight,S_termek1_weight,S_termek2_weight,O_termek1,O_termek2,O_termek1_appl_pont,O_
termek1_beh_pont,O_termek2_appl_pont,O_termek2_beh_pont,O_termek1_appl_intercept,O_termek1_appl_beta,O_termek1_beh_i
ntercept,O_termek1_beh_beta,O_termek1_appl_weight,O_termek1_beh_weight,O_REL_def_STD,O_macro_correction,O_termek2_appl_intercept,O_ter
mek2_appl_beta,O_termek2_beh_intercept,O_termek2_beh_beta,O_termek2_appl_weight,O_termek2_beh_weight,O_termek1_weigh
t,O_termek2_weight,H_termek1,H_termek2,H_termek1_appl_pont,H_termek1_beh_pont,H_termek2_appl_pont,H_termek2_beh_pont,H_termek1_appl_in
tercept,H_termek1_appl_beta,H_termek1_beh_intercept,H_termek1_beh_beta,H_termek1_appl_weight,H_termek1_beh_weight,H_
REL_def_STD,H_macro_correction,H_termek2_appl_intercept,H_termek2_appl_beta,H_termek2_beh_intercept,H_termek2_beh_beta,H_termek2_appl_
weight,H_termek2_beh_weight,H_termek1_weight,H_termek2_weight,H_atlag_pd,spare3,BEH_PD,behpd,txt1) VALUES
(?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?
,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?)

I tried to use NVARCHAR instead of VARCHAR without success. I also tried adding one more VARCHAR column to the insert (I created a new table before that with that additional column) and in this case the error message said that the error happens at 119. column (the new one). Stack trace:

java.sql.BatchUpdateException: A rendszer érvénytelen oszlophosszt kapott a BCP-ügyféltől a következő oszlopazonosítóhoz: 118.
   at com.microsoft.sqlserver.jdbc.SQLServerPreparedStatement.executeBatch(SQLServerPreparedStatement.java:2042)
   at SAS2SQL.writeToOutput(SAS2SQL.java:1098)
   at SAS2SQL.main(SAS2SQL.java:121)

The error message is in Hungarian. In English it is:

The system received an invalid column length from the BCP client for the following column ID: 118

The part of the code arranging the insert:

        String str = stdin.readLine();
        while (str != null) {
            counter++;
            //log(str);
            split(str);
            //log(arrayToText(elemek));
            int k;
            for (int j = 0; j < fields.length; j++) {
                tip=fieldtypes[j];
                textelem=elemek[j];
                k=j+1;
                if (textelem==null)pstmt.setString(k,null);
                else {
                    if (tip==Types.VARCHAR || tip==Types.NVARCHAR ) {
                    pstmt.setString(k, textelem);
                    
                    //if (insertedRows<15) log("TXT: "+textelem);
                    }
                if (tip==Types.INTEGER ||
                    tip==Types.BIGINT ||
                    tip==Types.SMALLINT ||
                    tip==Types.TINYINT 
                         ){
                         //log2("INT:"+Float.parseFloat());
                         if (textelem.trim().equals(".")) pstmt.setNull(k, java.sql.Types.FLOAT);
                         else pstmt.setFloat(k, Float.parseFloat(textelem));
                         }
                if (tip==Types.FLOAT ||
                    tip==Types.NUMERIC ||
                    tip==Types.DOUBLE ||
                    tip==Types.DECIMAL ) {
                         if (textelem.trim().equals(".")) pstmt.setNull(k, java.sql.Types.DOUBLE);
                         else pstmt.setDouble(k, Double.parseDouble(textelem));
                        }
                if (tip==Types.DATE) {
                        if (textelem.trim().equals(".")) pstmt.setNull(k, java.sql.Types.DATE);
                        else pstmt.setDate(k, java.sql.Date.valueOf(textelem));
                        //log2("DATE:"+SASResultSet.getDate(j).toString());
                     }
                if (tip==Types.TIMESTAMP) {
                        if (textelem.trim().equals(".")) pstmt.setNull(k, java.sql.Types.TIMESTAMP);
                        else pstmt.setTimestamp(k,Timestamp.valueOf(textelem.trim()));
                     }
                if (tip==Types.TIME) {
                        if (textelem.trim().equals(".")) pstmt.setNull(k, java.sql.Types.TIME);
                        else pstmt.setTime(k,Time.valueOf(textelem.trim()));
                        //log2("TIME:"+SASResultSet.getTime(j).toString());
                     }
                    }
                 }
            str = stdin.readLine(); 
            insertedRows++;
            pstmt.addBatch();
            bulkcounter++;  
            if (bulkcounter==batchSize){
                pstmt.executeBatch();
                pstmt.clearBatch();
                bulkcounter=0;
                if (monitor)log(getTime()+" / "+insertedRows);
                }
             //pstmt.executeUpdate();
            }
        if (bulkcounter!=0) {
            pstmt.executeBatch();
            pstmt.clearBatch();
            if (monitor)log(getTime()+" / "+insertedRows);
            }
        //if (siker==0) log(""+counter+". sor beszúrása sikertelen");   
        sqlServerConnection.commit();
        //outputConnection.commit();
        stdin.close();
        pstmt.close();
        log("Number of inserted rows: "+insertedRows); 

Answer 1

So the problem is that the bulkcopy client has its narrower limitation than the original engine.

setUseBulkCopyForBatchInsert=false solved the problem.