On Fri, 20 Jun 2014 19:51:27 -0500, Peter X. DeFabritus wrote:
>Actually, SYSSYM=ALLOW is only for system symbols. Your job stream does work
>for me, though, so I'm not sure what's wrong.
>
OK. My problem was with JESRECFM=V,JESLRECL=254 and //SYSIN DD *
where IEBGENER requires F 80. DUMMY worked much better.
How did you submit your job stream? What was the appearance of SYSUT2
when it ran?
>On Fri, 20 Jun 2014 18:59:53 -0500, Peter X. DeFabritus wrote:
>
>>You should also set SYSSYM=ALLOW for the job class you are using. Also, as
>>was discussed a while ago, if the long substitution causes the data to flow
>>past the end of the record, you will get a 001-5 ABEND.
>>
"End of the record" is a slippery concept when RECFM=V. Apparently,
it is possible to extend a record in this case, but to what limit?
I need to submit an RCF requesting documentation on what determines
the attributes (RECFM, LRECL, BLKSIZE) merged into the DCB by OPEN.
I suspect any or all of:
o LRECL on the SYSIN DD statement
o Attributes with which INTRDR is allocated
o JESRECFM and JESLRECL in FTP commands with FILETYPE=JES
o Attributes coded in the DCB macro
o Attributes modified by the DCB OPEN exit
o Operation of SDB
o Length of records appearing in the instream data set
o Length of records modified by symbol substitution
o JES2 versus JES3 (I suspect their behaviors differ).
This seems complicated enough that English prose is probably an
inadequate medium to communicate it. Better would be a table
or a flow chart. The information should appear in one place, not
redundantly under "*" and "DATA", lest updates leave them out
of sync.
My revised job as submitted via FTP:
//
//JCLSYM JOB 505303JOB,'Paul Gilmartin',
// MSGLEVEL=(1,1),REGION=16385K
//*
//* Doc: experiment with instream symbol substitution.
//*
//USERC OUTPUT JESDS=ALL,DEFAULT=YES,
// CLASS=R,PAGEDEF=V0648Z,CHARS=GT12
//*
// EXPORT SYMLIST=*
// SET X='This is a long symbol value.'
//*
//CONTROL EXEC PGM=IEFBR14,PARM='This &X. that.'
//*
//*.+....|....+....|....+....|....+....|....+....|....+....|....+....|....+....|
//STEP EXEC PGM=IEBGENER
//SYSPRINT DD SYSOUT=(,)
//SYSIN DD DUMMY
//SYSUT2 DD SYSOUT=(,)
//SYSUT1 DD *,SYMBOLS=CNVTSYS
Prime instream data set with a long line:
Lots of data to fill up a line, followed by a long symbol to see &X.
//*
//*.+....|....+....|....+....|....+....|....+....|....+....|....+....|....+....|
//
:w ! JESRECFM=V JESLRECL=254 submit $ZOS21_HOST
... and SYSUT2 from SDSF (manually reconstructed by scraping an 80-column
screen):
SDSF OUTPUT DISPLAY JCLSYM JOB03210 DSID 104 LINE 0 COLUMNS 01- 80
COMMAND INPUT ===> SCROLL ===> CSR
----+----1----+----2----+----3----+----4----+----5----+----6----+----7----+----8
********************************* TOP OF DATA **********************************
Prime instream data set with a long line:
Lots of data to fill up a line, followed by a long symbol to see This is a
long symbol value.
******************************** BOTTOM OF DATA ********************************
... Similar results when I write directly to INTRDR with an EDIT macro. When
I SUBmit the identical job from ISPF VIEW, I get RC=12 and in SYSPRINT:
********************************* TOP OF DATA **********************************
DATA SET UTILITY - GENERATE
IEB352I WARNING: ONE OR MORE OF THE OUTPUT DCB PARMS COPIED FROM INPUT
IEB351I I/O ERROR ,JCLSYM ,STEP ,JES ,I,SYSUT1 ,READ ,WRONG LEN RECRD,****
******************************** BOTTOM OF DATA ********************************
ISPF SUBmit needs to be liberated from its card reader mindset.
Thanks,
gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
