#9 Editing cobol copybook

[Anonymous]

Can I insert a new copybook field to a ExternalRecord instance ?

[Bruce Martin]

You can add a new field into a ExternalRecord e.g.

 xRec.getRecordFields().add(...);

But -

• It will not change the position of other fields; you need to do that your self
• If you are using IOBuilders, you need to create a new IOBuilder (see bellow) from the updated ExternalRecord.

Constructing an IOBuilder from an ExternalRecord:

  JRecordInterface1.SchemaIOBuilderProvider
                       .newIOBuilder(xRecord.asLayoutDetail());

[Anonymous]

Ok I would like to insert the field in the middle so that the offsets get adjusted automatically. Can this be provided as an enhancement ?

[Bruce Martin]

Ok, I will accept that, not sure of the time-table, may be a week; I have just started some other changes;
also it is easier if you login to sourceforge

[Bruce Martin]

How do you want it to work

• Do you want to provide a field and have it inserted in the correct place with fields shuffled down
• Do you want to provide an Index and name/type/length

[Bruce Martin]

I have created a updated version of JRecord (JRecord_Version_0.81.1_update1.7z) with the requested changes. You can download from:

https://sourceforge.net/projects/jrecord/files/jrecord/Version_0.81.1

Subversion ( https://sourceforge.net/p/jrecord/code/HEAD/tree/ ) has also been updated

[Immanuel Stephen]

Bruce,
Thanks for providing the update. I will try it
To answer your questions
providing both of the below will be useful
---- provide a field and have it inserted in the correct place with fields shuffled down
---- provide an Index and name/type/length
Also one more use case that I can think which may be useful is merging of two copybooks
essentially a application of add field operation

• Append fields from 2 ExternalRecord

  public ExternalRecord merge(ExternalRecord ext1, ExternalRecord ext2)

• Append 2 field lists

  public ExternalRecord merge(ExternalField []ef1, ExternalField []ef2)

Thanks
Immanuel

[Bruce Martin]

I have provided

• provide a field and have it inserted in the correct place with fields shuffled down. It provides three shuffle options
  • No shuffle
  • Shuffle by field-length
  • Calculate shuffle amount based on the insert option

With regards to the first option - Append 2 ExternalRecords, there is the newMultiCopybookIOBuilder builder. It allows multiple Cobol copybooks to be loaded into the one ExternalRecord. Usage:

ioBulder = JRecordInterface1.COBOL
                         .newMultiCopybookIOBuilder("Merged-Copybooks")
                             .addCopyBook("Copybook1.cbl")
                                  ....
                             .addCopyBook("Copybook2.cbl")
                                   ....
                             .addCopyBook("Copybook3.cbl")
                                   ....;

It will create a multi-record copybook
not sure if anyone has used that option though..

---

Not sure I understand the usage case of "merging" 2 lists or how the merge would work. I can understand adding fields / records
at a certain point

---

I will think about provide an Index and name/type/length

[Immanuel Stephen]

"Append 2 field lists" is similar to newMultiCopybookIOBuilder mentioned in your response.
say ef1 has 10 fields and ef2 has 5 fields, then the method will create a new ExternalRecord having 15 fields with ef2 added to the end of ef1 with their positions adjusted. But for now I think the newMultiCopybookIOBuilder should suffice for my requirements

[Bruce Martin]

Can you explain the situations you are facing ???

I would like to think about
the best way to handle it before I start coding ???

[Immanuel Stephen]

Bruce,
My ask is simple. A single record in a file is represented by 2 copybooks ie., 1 to 200 bytes is copybook A and 201 to 900 bytes is copybook B. So when I have this file as input, I would require to combine these 2 copybooks to format the records. Let me know if you want me to explain further

[Bruce Martin]

It depends,

if you have a filler at the start of the second copybook:

    01  Cobpybook-1.
       ...

:::cobol
    01  Cobpybook-2.
        03 filler                     Pic x(200).

Then use newMultiCopybookIOBuilder

---

If the second copybook does not contain the 200 byte filler but starts with 03 level
Read both copybooks into a string variables, concentan

:::cobol
03 Cobpybook-2.

:::java

  String copybook1 = loadFile("copybook1.cbl");
  String copybook2 = loadFile("copybook2.cbl");

  JRecordInterface1.COBOL
                   .newIOBuilder(
                        new StringReader(
                               copybook1 + "\n" + copybook2),
                        "Combined-Copybook") ....

if the 2nd copybook starts with 01 and does not contain the filler at the start. Let me know
and I will look at changes. But You could do

• use newMultiCopybookIOBuilder

• Then update all the fields in the second record

  ioBulder = JRecordInterface1.COBOL
  .newMultiCopybookIOBuilder("Merged-Copybooks")
  .addCopyBook("Copybook1.cbl")
  ....
  .addCopyBook("Copybook2.cbl")
  ....;
  xRecord = ioBuilder.getExternalRecord();

  ExternalField[] fields = xRecord.getRecord(1).getRecordFields();

  for (ExternalField f : fields) {
  f.setPos(f.getPos() + 200);
  }

[Immanuel Stephen]

my case is "if the 2nd copybook starts with 01 and does not contain the filler at the start."

[Bruce Martin]

I will update the newMultiCopybookIOBuilder so that you can set an initial position for the copybook
Initially I allow setting an integer starting position; but will ultimately support

• setStartingPosition(int pos)
• setStartingPosition(String copybookName, String fieldName)
• setStartingPositionFromCopybookLength(String copybookName)

What you some times see in cobol is

 01  Copybook
     03 Common-Header.
        ....
     03  Variable-Part             Pic x(...)

Where the variable part is described by multiple copybooks.
You seem to have a variation on this. The above changes should
handle this situation.

---

Hopefully have something next week.

[Immanuel Stephen]

Regarding newMultiCopybookIOBuilder you had mentioned that "It will create a multi-record copybook". But I want to create a single record copybook created out of these multiple copybooks. is that possible ?

[Bruce Martin]

It returns a multiple record layout, but it should not matter if you are accessing by field / field-name.

you can do

 v1 = line.getFieldValue("field-in-record-1").asString();
 v3 = line.getFieldValue("field-in-record-3").asString();

Internally JRecord tries to follow Cobol rather than Java. It encodes/decodes fields when they are accessed.

[Immanuel Stephen]

ok I will try using newMultiCopybookIOBuilder and see if it meets my requirements. Have outlined my requirement in my previous response

[Bruce Martin]

I have added JRecord_Version_0.81.1_update2.7z at https://sourceforge.net/projects/jrecord/files/jrecord/Version_0.81.1/

This version allows you to set Starting-Position of a Cobol copy e.g.

   ICobolMultiCopybookIOBuilder iob = JRecordInterface1.COBOL
             .newMultiCopybookIOBuilder("MultiCopybook")
                 .setDialect(ICopybookDialects.FMT_MAINFRAME)
                 .addCopyBook("DTAR020.cbl")
                 .addCopyBook("SecondRecord.cbl"))
                     .setStartingPosition(120);

or using a field:

    ICobolMultiCopybookIOBuilder iob = JRecordInterface1.COBOL
             .newMultiCopybookIOBuilder("MultiCopybook")
                 .setDialect(ICopybookDialects.FMT_MAINFRAME)
                 .addCopyBook("Master.cbl")
                 .addCopyBook("Child1.cbl")
                     .setStartingPositionToField("Master.cbl", "Record-Detail");

---

Do you really need to insert fields and have subsequent fields shuffle down ???
I do not understand why you need this ???

[Immanuel Stephen]

Will this upgrade (JRecord_Version_0.81.1_update2.7z) provide the following which was mentioned in one of your previous answers

ioBulder = JRecordInterface1.COBOL
         .newMultiCopybookIOBuilder("Merged-Copybooks")
             .addCopyBook("Copybook1.cbl")
                  ....
             .addCopyBook("Copybook2.cbl")
                  ....;
xRecord = ioBuilder.getExternalRecord();
ExternalField[] fields = xRecord.getRecord(1).getRecordFields();
for (ExternalField f : fields) {
     f.setPos(f.getPos() + 200);
}

How does setStartingPosition help to achieve this ?

[Bruce Martin]

What you would do is:

ioBulder = JRecordInterface1.COBOL
         .newMultiCopybookIOBuilder("Merged-Copybooks")
             .addCopyBook("Copybook1.cbl")
                  ....
             .addCopyBook("Copybook2.cbl")
                  .setStartingPosition(200);      // <<<====
                  ....;

[Immanuel Stephen]

"Do you really need to insert fields and have subsequent fields shuffle down ???" ..... this is required in a use case where I visually display a copybook and allow the user insert fields.

[Immanuel Stephen]

I tried something like this :

public static ExternalRecord appendCopyBooks(String copy1, String copy2) {

        ICobolMultiCopybookIOBuilder iob = JRecordInterface1.COBOL
                .newMultiCopybookIOBuilder("multicopy")
                .setDialect(ICopybookDialects.FMT_MAINFRAME)
                .addCopyBook(copy1)
                .addCopyBook(copy2);

        ExternalRecord rec=null;
        try {
            rec = iob.getExternalRecord();
            ExternalRecord rec0 = rec.getRecord(0);
            ExternalRecord rec1 = rec.getRecord(1);

            int cnt1 = rec0.getRecordFields().length;
            ExternalField last = rec0.getRecordFields()[cnt1 - 1];
            int len1 = last.getPos() - 1 + last.getLen();

            ExternalField[] ef = rec1.getRecordFields();
            for (ExternalField efs : ef) {
                efs.setPos(efs.getPos() + len1);
                rec0.addRecordField(efs);
            }
        } catch (IOException e) {
            e.printStackTrace();
        }

        return rec;

    }

[Bruce Martin]

That should pretty much work, The one potential problem is if you have a Copybook like

 01 My-Record
        ....
        03 redefined-field         pic x(500).
               ....
         03 filler redefines redefined-field.
                05 last-field          pic xx.

in which case the length is determines by redefined-field which is not the last field.
you would be better to search for the length

int length = 0;
ExternalField[] flds = rec0.getRecordFields();
for (ExternalField f : flds) {
    length = Math.max(length, f.getPos + f.getLength - 1);
}

with Update2 you can do

ISchemaIOBuilder iob = externalRecord.asIOBuilder();

to convert the ExternalRecord back into a IOBuilder.