2

Does anyone have a quick method for finding a COBOL copybook start-positions for fields? For example:

000100 01  BGG-FILE-REC.                                                        
000200     03  BGG-RCD-KEY.                                                     
000300         05  BGG-DUDENAME            PIC XXXX. 
000400         05  BGG-DUDEADDR            PIC XX.   
000500         05  BGG-HAIRCOLOR           PIC X(71). 
000600         05  BGG-EYECOLOR            PIC X(8).

The BGG-HAIRCOLOR column begins at column 7. However, the actual file I am handling has about 250 variables totalling to 3400 bytes. Slogging through and adding the PIC values with a hand calculator is hideous and not happening in my lifetime. The number would be nice to have in vi for when I am examining output from the system; i.e, does BGG-EYECOLOR value fall in the correct column?

I could write a perl script to annotate such a copybook but before I do it I wondered if anyone has a smoother way to get the same value?

Bill Woodger
  • 12,968
  • 4
  • 38
  • 47
noogrub
  • 872
  • 2
  • 12
  • 20

5 Answers5

2

If it is for the mainframe, you could use cb2xml to convert the copybook to xml (with field Start).

Alternatively you can use the RecordEditor. Import the Cobol Copybook into the RecordEditor (see RecordEditor-Cobol Notes). The RecordEditor will calculate the field positions. You can copy and paste the fields from the RecordEditor-Layout to either Excell or a Text Editor, Alternatively you can export the record layout as Xml

The RecordEdityor willwork for several Cobol Dialects (Mainframe, OpenCobol).

If you are reading files in java, you should look at JRecord. JRecord will let you read a Cobol Data file in Java using a Cobol Copybook.

Also Legstar may be worth a look. Legstar originally start with dealloy with online data transfer but they have been branching out.

Bruce Martin
  • 10,358
  • 1
  • 27
  • 38
  • RecordEditor seems a good shot, from the recent question. Otherwise, if on Mainframe the Data Map, there are probably/possibly other options for other compilers. – Bill Woodger Mar 22 '13 at 15:42
  • In fact I am in Java on a Unix server, writing code to convert COBOL files using their copybook as my guide to value locations. However this answer lets me tell my mainframe crew what I need from them, and how to obtain it. Thank you! – noogrub Mar 22 '13 at 19:57
1

disclaimer : I maintain https://www.cobolcopybook.co.in

Hi, check the site https://www.cobolcopybook.co.in, This site is designed specially for the analyze COBOL copybooks.

for Eg. Your input copybook is (as mentioned in question): 

000100 01  BGG-FILE-REC.
000200     03  BGG-RCD-KEY.
000300         05  BGG-DUDENAME            PIC XXXX.
000400         05  BGG-DUDEADDR            PIC XX.
000500         05  BGG-HAIRCOLOR           PIC X(71).
000600         05  BGG-EYECOLOR            PIC X(8).

Then the output will be:

SR#  LEVEL   FIELD NAME                 PICTURE   TYPE      START      END   LENGTH

0  1  BGG-FILE-REC.                 @         AN            1       85       85
1  3  BGG-RCD-KEY.                  @         AN            1       85       85
2  5  BGG-DUDENAME                  XXXX.     AN            1        4        4
3  5  BGG-DUDEADDR                  XX.       AN            5        6        2
4  5  BGG-HAIRCOLOR                 X(71).    AN            7       77       71
5  5  BGG-EYECOLOR                  X(8).     AN           78       85        8

I hope this will solve your problem.

Nilesh
  • 69
  • 1
  • 11
  • Thank you for notifying this.. Will solve this issue and let you know. Please feel free to write on nrcmoney@gmail.com if you want to see new features in the website. :) – Nilesh Sep 23 '20 at 07:28
0

If you're on a mainframe, use option 8 of FileAid.

mswanberg
  • 1,285
  • 8
  • 20
  • My company is the cheapest of the cheap and even we have FileAid... But you are correct. A compile listing would give the desired info, as long as it starts on an 01 level. – mswanberg Mar 23 '13 at 18:08
  • I don't believe that's true. If the copybook starts on a level higher than 01 and has data before it, then the entire copybook will be off by the number of bytes before it. You could subtract, but it's just easier to make sure it starts on an 01 level if you're using a compile solely to see field offsets. – mswanberg Mar 16 '15 at 23:17
  • Sure, the subtract is required in that case, but hardly onerous. These days with the latest IBM compiler, you can even get the offsets in decimal. Just makes it too easy. Anyway, let's tidy away the comments. I don't think they're vital here. OP's not on the Mainframe, a comment on Bruce's answer later revealed. – Bill Woodger Mar 16 '15 at 23:41
0

You are worried about how hard it is to line up the PIC clauses in the same column?

Open it in one of the many Eclipse variants and control-shift-F or whatever.

You know it compiles nicely regardless of columns, and has for almost 30 years, they really don't matter any more.

The low rent way would be to compile it and cut&paste your cross reference listing over your copybook.

Joe Zitzelberger
  • 4,238
  • 2
  • 28
  • 42
0

If you are using Mainframes you can use "File aid - view". From ISPF Primary Option Menu goto

G ==> General Utility F ==> File-Aid Products 1 ==> File-AID 8 ==> VIEW

In the screen the below options will come:

Specify Record Layout Dataset to View:                                 

   Dataset name  ===> 'XXXXXX.XXX.XXXX'                
   Member name   ===> COPYZXY   (Blank or pattern for member list)

Give your "Dataset name" and "Member name" (Dataset name : Your PDS and Member name : Your copy book name)

After pressing enter you will get a VIEW LAYOUT screen that describes about Variable name, Clause, number, Start, end and length of each variable present in that copy book.

E P
  • 11
  • 1
  • 4