Here is one solution, good old gener ...
//ICEGENER EXEC PGM=ICEGENER
//SYSPRINT DD SYSOUT=*
//SYSIN DD DUMMY
//SYSUT1 DD DISP=SHR,DSN=AFICRWD.EPSON.VBSX
//SYSUT2 DD DISP=(NEW,CATLG),DSN=AFICRWD.EPSON.U,
// UNIT=DISK,SPACE=(CYL,(1,1),RLSE),
// DCB=(RECFM=U,LRECL=32760,BLKSIZE=0)
AFICRWD.EPSON.VBSX has LRECL=X,RECFM=VBS and recs > 32k
The AFICRWD.EPSON.U dataset has all of the data with the >32K
logical records split into one or more following physical records. Since
the data has easily identifiable markers at the start of each logical
record, it is easy to create logical records in a program. Bonus is that
you can browse the RECFM=U dataset, and that makes the programmers happy.
They dislike programming what they can't see, for some reason.
I found that I could create the file with LRECL>32K using
DCB=(LRECL=X,BLKSIZE=32000,RECFM=VBS), and I suspect that my programmers
program could read this data because the data is really cut into 32K
chunks inside the file anyhow. I have requested that he test it, we will
see. Might have to change BLKSIZE to + or - 8, we'll see.
And FINALLY, it is possible to write a script in Unix (or Windows,
I assume) to insert new line chars where necessary, and then you get the
records automatically segment when you send them to the mainframe.
I suspect that this will become a more common problem as we spend
more time talking back and forth between systems.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
