Hyphen in COBOL field name

[Gabor Markon]

Dear Support,

I've tried to run a sample application to read a (downloaded) mainframe file by using a COBOL data structure definition, something like this:

           01  TCQ76B-DATEN.                                           
               05 TCQ76B-SORT.                                             
                    10 TCQ76BXWK            PIC X(004).                  
                    10 TCQ76B-FT            PIC X(024).                   
                    10 TCQ76B-POS           PIC X(004).                       

It seems the Java application does not interpret the hyphen "-" as a part of the field name. Rather the name will be cut at "-" and only the part after ("FT", "POS"...) will be delivered by the code line "entry.getKey()".

Another issue is: the field names will be not delivered in the loop in the sequence defined in the copybook (here the corresponding part of my code).

AbstractLineReader reader = ioProvider.getLineReader(
    fileStructure, Convert.FMT_MAINFRAME,
    CopybookLoader.SPLIT_NONE, copybookName, salesFile
);

AbstractLineWriter writer = ioProvider.getLineWriter(fileStructure, salesFileOut);

while ((saleRecord = reader.read()) != null) {
    lineNum += 1;
    Map mymap = new HashMap();
    mymap = saleRecord.getLayout().getFieldNameMap();

    Set set = mymap.entrySet(); //Converting to Set so that we can traverse  
    Iterator itr = set.iterator();
    while (itr.hasNext()) {
        //Converting to Map.Entry so that we can get key and value separately  
        Map.Entry entry = (Map.Entry) itr.next();
      **  System.out.println("field...:" + entry.getKey() + " **" + entry.getValue().toString());
    }
    ...
}

And a last question. Is JRecord runnable on a mainframe (perhaps using JZOS to read the mainframe files)?

Best regards,
Gabor

[Bruce Martin]

Firstly it would be better to use the JRecordInterface1.COBOL.newIOBuilder(>>) interface.

It seems the Java application does not interpret the hyphen "-" as a part of the field name. Rather the name will be cut at "-" and only the part after ("FT", "POS"...) will be delivered by the code line "entry.getKey()".

• Depends on what attributes are set. The setDropCopybookNameFromFields(boolean)
  of the Iobuilder controls wether the Copybook name is dropped from the front of
  field-names.

       ICobolIOBuilder iob = JRecordInterface1.COBOL
            .newIOBuilder("G:/Users/Bruce01/RecordEditor_HSQL/CopyBook/Cobol/DTAR020.cbl")
                 .setDropCopybookNameFromFields(false)
                 .setFont("cp037");

Another issue is: the field names will be not delivered in the loop in the sequence defined in the copybook (here the corresponding part of my code).

• The getFieldNameMap() returns a hashmap. The getFieldIterator method is what you want

            final FieldIterator iter = saleRecord.getFieldIterator(0);
            while (iter.hasNext()) {
                final AbstractFieldValue field = iter.next();
                System.out.print("|" + field.getFieldDetail().getName() + "=" + field.asString());
            }

And a last question. Is JRecord runnable on a mainframe (perhaps using JZOS to read the mainframe files)?

• I have not tried, probably. It should be possible use JZOS to read/write files
  but you will need to write some interface code.

in JRecord you can use the IByteRecordReader, IByteRecordWriter interfaces with
newLineReader / newLineWriter methods.

To read a file using JZOS, You would need to create a class that implements IByteRecordReader
that uses JZOS to read the file and returns an array of bytes

     reader = iobuilder.newLineReader(IByteRecordReader byteReader);

The IByteRecordReader interface:

public interface IByteRecordReader {

    /**
     * Read one line from the input file as an array of bytes
     *
     * @return line read in
     *
     * @throws IOException io error
     */
    byte[] read() throws IOException;

    /**
     * Closes the file
     *
     * @throws IOException io error
     */
    void close() throws IOException;
}

[Gabor Markon]

Hello Bruce,

thanks for the prompt answer, I will try asap, and let you know.

Regards
Gabor

Am Fr., 13. März 2020 um 14:35 Uhr schrieb Bruce Martin bruce_a_martin@users.sourceforge.net:

Firstly it would be better to use the
JRecordInterface1.COBOL.newIOBuilder(>>) interface.

It seems the Java application does not interpret the hyphen "-" as a part
of the field name. Rather the name will be cut at "-" and only the part
after ("FT", "POS"...) will be delivered by the code line "entry.getKey()".

• Depends on what attributes are set. The
  setDropCopybookNameFromFields(boolean)
  of the Iobuilder controls wether the Copybook name is dropped from the
  front of
  field-names.

  ICobolIOBuilder iob = JRecordInterface1.COBOL
  .newIOBuilder("G:/Users/Bruce01/RecordEditor_HSQL/CopyBook/Cobol/DTAR020.cbl")
  .setDropCopybookNameFromFields(false)
  .setFont("cp037");

Another issue is: the field names will be not delivered in the loop in the
sequence defined in the copybook (here the corresponding part of my code).

• The getFieldNameMap() returns a hashmap. The getFieldIterator
  method is what you want

      final FieldIterator iter = saleRecord.getFieldIterator(0);
      while (iter.hasNext()) {
          final AbstractFieldValue field = iter.next();
          System.out.print("|" + field.getFieldDetail().getName() + "=" + field.asString());
      }
  

And a last question. Is JRecord runnable on a mainframe (perhaps using
JZOS to read the mainframe files)?

• I have not tried, probably. It should be possible use JZOS to
  read/write files
  but you will need to write some interface code.

in JRecord you can use the IByteRecordReader, IByteRecordWriter interfaces
with
newLineReader / newLineWriter methods.

To read a file using JZOS, You would need to create a class that
implements IByteRecordReader
that uses JZOS to read the file and returns an array of bytes

 reader = iobuilder.newLineReader(IByteRecordReader byteReader);

The IByteRecordReader interface:

public interface IByteRecordReader {

/**     * Read one line from the input file as an array of bytes     *     * @return line read in     *     * @throws IOException io error     */
byte[] read() throws IOException;

/**     * Closes the file     *     * @throws IOException io error     */
void close() throws IOException;}

---

Hyphen in COBOL field name
https://sourceforge.net/p/jrecord/discussion/678633/thread/b5a81928ae/?limit=25#5032

---

Sent from sourceforge.net because you indicated interest in
https://sourceforge.net/p/jrecord/discussion/678633/

To unsubscribe from further messages, please visit
https://sourceforge.net/auth/subscriptions/

[Gabor Markon]

Hello Bruce,

issues 1. (Java application does not interpret the hyphen "-" as a part
of the field name) and 2. (field names will be not delivered in the loop in the
sequence defined in the copybook ) are clear now, thanks.

As for JZOS: I'll try to implement your suggestion in the next time.

Regards
Gabor