The problem is that ISPF/ISREDIT evaluates/substitutes any ISPF variable, which
in this case is "&PDQR".
So:
ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE' "
is the same as
"ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR' "
and is the same as
"ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR.' "
In the latter two cases (apparently) "&PDQR" resolves into "" (empty) and
"&PDQR." also. (The dot "." is used as an end-of-variable-name marker, IIRC.)
In e g this:
"ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR.(' "
- resolves to: "ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE(' "
wich is no to be found in the dataset.
Etc.
Regards,
Thomas Berg
_______________________________________________________
Thomas Berg Specialist AM/SM&S SWEDBANK AB (publ)
> -----Ursprungligt meddelande-----
> Från: IBM Mainframe Discussion List [mailto:[email protected]]
> För John Mattson
> Skickat: den 18 juli 2012 18:48
> Till: [email protected]
> Ämne: REXX ISPF edit FIND failing
>
> I have a little dataset which contains
> //REMOTE DD DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR.(&SYSTEM)
> //REMOTE DD DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR.(SYSTEM)
> //REMOTE DD DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR(SYSTEM)
> //SYSIN DD DISP=SHR,DSN=&PDQ.ALC.UNVLIB(&UCMIN)
> X/MYSCRIPT DD DISP=SHR,DSN=&PDQ.ALC.UNVLIB(&MY)
>
> I have a little REXX which I want to "FIND" these strings
> /* REXX */
> TRACE I
> Address ISPEXEC
> "ISREDIT MACRO (MEM) NOPROCESS"
> "CONTROL ERRORS RETURN"
> "ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE' "
> "ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR' "
> "ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR.' "
> "ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR.(' "
> "ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR.(&' "
> "ISREDIT F ALL 'DISP=SHR,DSN=MSYS.UCMD.REMOTE&PDQR.(&SYSTEM)' "
> "ISREDIT F ALL 'DISP=SHR,DSN=&PDQ.ALC.UNVLIB(&UCMIN)' "
> "ISREDIT F ALL 'DISP=SHR,DSN=&PDQ.ALC.UNVLIB(&MY)' "
> EXIT
>
> The First Three FIND's work fine. Starting from the fourth
> find, they all get RC=4, not found. Even tho I can clearly see that the
> strings exist in the dataset. What in the whirled is going on here?? Is
> there something "special" about two &'s in a string? I have tried
> removing the second & from both the find and my dataset and the find
> still fails. I am at wits end here.
> I have also tried "extracting" the actual finds like: F ALL
> 'DISP=SHR,DSN=&PDQ.ALC.UNVLIB(&MY)'and executing them in TSO, and they
> all work. What is special here. FYI, the final " is in col 71
>
> ----------------------------------------------------------------------
> 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
