Hello Sri, Argh!! Yes too many assumptions... 1) True. Not thinking of DSN on a separate line. 2) True. Step names can be single characters. 3) A temporary Dataset still has a Dataset Name, or a Dataset Name reference, right? 4) I did not know DMBATCH could have SYSIN2. 5) True. This was my attempt at ignoring any sequence numbers that might be in 73-80.
The reason for this exercise is that in some situations, the DMNETMAP and DMMSGFIL DataSet Names need to be modified. The Dataset Names need to have a newer value when certain conditions are met in the NDM control statements. I first tried scanning all of our Control Statement libraries and found over 80,000 possibilities. Then I thought I could check the JCL and PROCS, and grab a list with all the JCL/PROC Member Names and their associated DMNETMAP, DMMSGFIL, and SYSIN dataset names. Then from this list I could ignore the JOBs where the Dataset Names have already been updated. This gives me a much smaller list. Several hundred items. Then I will need to check each of the associated SYSIN members manually to determine if any changes may be required. Anyway, thanks for taking some time to explain this. On Thu, Oct 29, 2020 at 7:53 PM Sri h Kolusu <[email protected]> wrote: > > The control statements below work fine. They do what I wanted. > > Today my question is if these SORT Control statements can be optimized? > > This is not urgent. > > > Cameron, > > Your job definitely can be optimized. But I want to make a observation > that you might be skipping some results. > > You are assuming that the ddnames DMNETMAP and DMSGFIL have the dataset > name on the same line. > > 1. ) For example if your input has the following Input, you will NOT pick > the DSN names > > > > //DMNETMAP DD DISP=SHR, > // DSN=$HLQ.NETMAP > //DMPUBLIB DD DISP=SHR, > // DD DSN=$HLQ.PROCESL > //DMMSGFIL DD DISP=OLD, > // DSN=$HLQ.MSG4 > > > 2.) Similarly the stepnames can any where from 1 byte to 8 bytes, however > your picking bytes from 4 thru 8 . Assume the following input > > //S1 EXEC PGM=DMBATCH,REGION=1024K,PARM=(YYSLYNN) . > > > > 3.) What happens if SYSIN cards are generated from previous step as a temp > dataset and passed over to DMBATCH step? > > ex: > > https://www.ibm.com/support/pages/how-submit-instream-process-jcl-using-dgadbatc-dmbatch > > Check how &&TEMP is generated. > > How do you plan to handle the &&TEMP contents? > > > 4.) Also the program DMBATCH can have SYSIN2 also, do you plan to extract > the information? > > > 5.) Technically SYSIN can be a PDS with member name. So the maximum length > of PDS+Member = 54 bytes ( 44 Dsn name+ 8 byte member name + 2 byte for > opening and closing parenthesis). However your control cards don't seem to > be handling them. > > Let me know the answers to be above questions and I will show you the > optimized control cards. > > > Thanks, > Kolusu > DFSORT Development > IBM Corporation > > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
