IBM Mainframe Forum

by **chillmo** » Thu Aug 11, 2022 3:54 am

Team,

I hope everyone is doing well. I have a requirement to create a couple records based on one field from the input file (customer number).

Sample input file:

Select all

11177777777777Y
11199999999999Y
11188888888888Y
 

Desired output file:

Select all

11177777777777YCOKE
11177777777777NSPRITE
11199999999999YCOKE
11199999999999NSPRITE
11188888888888YCOKE
11188888888888YSPRITE
 

I made this work by using OUTFIL, creating two different files while using the same input file, with different BUILD parms, then merging the files.

I want to know if this can be done in one pass using sort.

Any assistance would be greatly appreciated.

Thanks!

by **sergeyken** » Thu Aug 11, 2022 5:34 pm

You need to use JOINKEYS for two separate inputs, to join the records all-to-all:

Select all

//......
//LEFT     DD  *
11177777777777Y
11199999999999Y
11188888888888Y
//*
//RIGHT    DD  *
YCOKE
NSPRITE
//*
//SYSIN    DD  *
 JOINKEYS F1=LEFT,FIELDS=(81,1,A)    JOIN ON THE ADDED FIELD
 JOINKEYS F2=RIGHT,FIELDS=(81,1,A)    JOIN ON THE ADDED FIELD
 REFORMAT FIELDS=(F1:1,14,F2:1,14)     11177777777777 <-> YCOKE
 INREC OVERLAY=(80:X)         EXTEND TO NORMAL SIZE
 SORT FIELDS=COPY
 END
//*
//JNF1CNTL DD  *
 INREC OVERLAY=(81:X)   ADD THE FIELD TO JOIN ON
//*
//JNF2CNTL DD  *
 INREC OVERLAY=(81:X)   ADD THE FIELD TO JOIN ON
//*
 

by **chillmo** » Thu Aug 11, 2022 9:54 pm

Thanks sergeyken for the prompt response.

This worked to perfection and I was able to get rid of my two-step approach. However, the business added a new wrinkle.

New input file:

Select all

11177777777777YN 777
11199999999999YN 999
11188888888888YN 888
 

New output file:

Select all

Account Number, Invoice Number, Sale, Word, New Value 
Hello,111,77777777777,,SPRITE, ,777
Hello,111,77777777777,,COKE, ,Y
Hello,111,99999999999,,SPRITE, ,999
Hello,111,99999999999,,COKE, ,Y
Hello,111,88888888888,,SPRITE, ,888
Hello,111,88888888888,,COKE, ,Y
 

I can add the header via the HEADER parm on the OUTFIL statement in addition to adding the word Hello to each record but NOT sure how to add the new value from input file to only the first record.

by **sergeyken** » Fri Aug 12, 2022 1:02 am

What is "new value from input file"?
Where it comes from?
How it looks like?

Etc. etc. etc...

by **chillmo** » Fri Aug 12, 2022 2:09 am

It's the 3-digit value at the end of the 1st file:

Select all

11177777777777YN 777 <--------
11199999999999YN 999 <--------
11188888888888YN 888 <--------
 

Desired output file:

Select all

Msg, Account Number, Invoice Number, Sale, Word, New Value
Hello,111,77777777777,,SPRITE,777
Hello,111,77777777777,,COKE,Y
Hello,111,99999999999,,SPRITE,999
Hello,111,99999999999,,COKE,Y
Hello,111,88888888888,,SPRITE,888
Hello,111,88888888888,,COKE, ,Y
 

by **sergeyken** » Fri Aug 12, 2022 2:23 am

1. Renumber the records from F2 (using SEQNUM)

2. Join two input streams as before

3. Use IFTHEN in either OUTREC, or OUTFIL, - to reformat the final record based on record number of F2. For number 1 append the “new field”, for all other numbers (if more than two) do not append the “new field”

IBM Mainframe Forum

Create multiple records based on one record within a file.

Create multiple records based on one record within a file.

Re: Create multiple records based on one record within a fil

Re: Create multiple records based on one record within a fil

Re: Create multiple records based on one record within a fil

Re: Create multiple records based on one record within a fil

Re: Create multiple records based on one record within a fil