The SYSOUTTRAP will trap TEST OUTPUT but *ONLY* at the beginning of *TEST* Coding AT +148 ( EX MY.CLIST(TRAP))
Where the clist TRAP contain SYSOUTTRAP and SYSOUTLINE to trap output won't work -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Bill Godfrey Sent: Monday, July 4, 2016 2:56 PM To: [email protected] Subject: Re: TSO TEST and SYSOUTTRAP/SYSOUTLINE The clist in my first post was run on z/OS 2.2. The 4-character string "amp:" if present, is an artifact of the listserv web interface and should not be in the clist. The clist has also run on MVS 3.8, without the TSOEXEC in front of TEST and with different SUBSTR columns as shown in comments in my earlier post. The different SUBSTR columns are due to the 6-digit hex address on 24-bit MVS 3.8, instead of an 8-digit hex address. The column dependency could be eliminated, if somebody wanted to, by using an equated symbol on LIST, as below. If the EQUATE was done after the SET, the output of LIST would end up in SYSOUTLINE2. TSOEXEC TEST 'SYS1.LINKLIB(IEFBR14)' EQUATE CVTDATE 10.%+38 SET SYSOUTTRAP = 10 LIST CVTDATE SET SYSOUTTRAP = 0 SET LINE = &STR(&&SYSOUTLINE1) SET DOY = &SUBSTR(14:16,&LINE) SET YYY = &SUBSTR(11:13,&LINE) SET YEAR = &EVAL(1900 + &YYY) WRITE TODAY IS DAY &DOY OF &YEAR CONTROL END(FOO) /* NEXT END IS FOR TEST, NOT CLIST */ END Bill On Mon, 4 Jul 2016 13:33:18 -0400, Joseph Reichman wrote: >What version of z/OS MVS was this code run I think Rupert Reynolds >mentioned in a earlier Post on this thread that capturing TEST output >is different then when he tried it in the early 90's > > > > >> On Jul 4, 2016, at 12:28 PM, Bill Godfrey wrote: >> >> Here's an example of capturing TEST output into CLIST variables. >> >> TSOEXEC TEST 'SYS1.LINKLIB(IEFBR14)' >> SET SYSOUTTRAP = 10 >> LIST 10.%+38 /* CVTDATE */ >> SET SYSOUTTRAP = 0 >> SET LINE = &STR(&&amp;SYSOUTLINE1) SET DOY = &SUBSTR(16:18,&LINE) >> /* 14:16 ON MVS 3.8 */ SET YYY = &SUBSTR(13:15,&LINE) /* 11:13 ON MVS >> 3.8 */ SET YEAR = &EVAL(1900 + &YYY) WRITE TODAY IS DAY &DOY OF &YEAR >> CONTROL END(FOO) /* NEXT END IS FOR TEST, NOT CLIST */ END >> >> Bill >> >>> On Sun, 3 Jul 2016 13:58:51 -0400, Joseph Reichman wrote: >>> >>> I have tried to capture test subcommand output I.E list under clist >>> with no success Dont know about Rexx I am now thinking the AT >>> subcommsnd in TEST leaves you the ability to to exec a CLIST as part >>> of the parameters maybe I'll try the SYSOUTTRAP/SYSOUTLINE in that >>> instance >>> >>> >>> >>>> On Jul 3, 2016, at 1:51 PM, Jesse 1 Robinson <[email protected]> >>>> wrote: >>>> >>>> I was sloppy. AFAIK the functions are equivalent. In Rexx, you use this >>>> syntax: >>>> >>>> x = OUTTRAP("cmdl.") >>>> "LISTALC STATUS" >>>> line_count = cmdl.0 >>>> dsn = cmdl.i >>>> ... >>>> >>>> Output goes into the named array. Lines are identified by the value of i. >>>> >>>> Equivalent in Clist: >>>> >>>> SET SYSOUTTRAP = 600 >>>> LISTALC STATUS >>>> SET SYSOUTTRAP = 0 >>>> SET &LINE_COUNT = &SYSOUTLINE >>>> SET DSN = &STR(&&amp;SYSOUTLINE&I) ... >>>> >>>> Output always goes into complex variable &SYSOUTLINE. Lines are identified >>>> by the value of &I. >>>> >>>> The real difference for OP is the scope of output capture in Rexx vs. >>>> Clist. When Rexx switches into subcommand mode, I don't think output is >>>> captured from the original OUTTRAP. >>>> >>>> . >>>> . >>>> . >>>> J.O.Skip Robinson >>>> Southern California Edison Company >>>> Electric Dragon Team Paddler >>>> SHARE MVS Program Co-Manager >>>> 323-715-0595 Mobile >>>> 626-302-7535 Office >>>> [email protected] >>>> >>>> >>>> -----Original Message----- >>>> From: IBM Mainframe Discussion List >>>> [mailto:[email protected]] On Behalf Of Joseph Reichman >>>> Sent: Sunday, July 03, 2016 9:54 AM >>>> To: [email protected] >>>> Subject: (External):Re: TSO TEST and SYSOUTTRAP/SYSOUTLINE >>>> >>>> You use SYSOUTTRAP in Rexx thought it's a clist command unless you >>>> are calling clist from Rexx >>>> >>>>> On Jul 3, 2016, at 12:48 PM, Jesse 1 Robinson <[email protected]> >>>>> wrote: >>>>> >>>>> I use SYSOUTTRAP in Rexx all the time for ordinary commands. There is >>>>> indeed a huge difference between Rexx and Clist vis a vis 'subcommand >>>>> mode'. An example of subcommand mode is the SEND command under OPER. >>>>> Another is LIST under ACCOUNT. TEST runs in practice almost entirely in >>>>> subcommand mode. >>>>> >>>>> In a Clist, the exec retains control when running in subcommand mode. In >>>>> Rexx, the exec loses control in subcommand mode, so successive commands >>>>> must be 'stacked' ahead of time via QUEUE. Trouble is that in TEST, >>>>> subsequent command operands usually depend on the result of previous >>>>> commands, which makes QUEUEing commands ahead of time nearly impossible. >>>>> >>>>> I don't recall ever trying SYSOUTTRAP in TEST. >>>>> >>>>> . >>>>> . >>>>> . >>>>> J.O.Skip Robinson >>>>> Southern California Edison Company Electric Dragon Team Paddler >>>>> SHARE MVS Program Co-Manager >>>>> 323-715-0595 Mobile >>>>> 626-302-7535 Office >>>>> [email protected] >>>>> >>>>> -----Original Message----- >>>>> From: IBM Mainframe Discussion List >>>>> [mailto:[email protected]] On Behalf Of Joseph Reichman >>>>> Sent: Sunday, July 03, 2016 9:18 AM >>>>> To: [email protected] >>>>> Subject: (External):Re: TSO TEST and SYSOUTTRAP/SYSOUTLINE >>>>> >>>>> I don't think the clist SYSOUTTRAP/SYSOUTTRAP Work to trap TEST >>>>> output maybe Rexx OUTTRAP does didn't try that >>>>> >>>>> >>>>> >>>>>> On Jul 3, 2016, at 9:16 AM, Rupert Reynolds wrote: >>>>>> >>>>>> TEST behaves differently (or did, when I used it in the 90s) in CLIST. >>>>>> As I remember it, the CLIST that starts TEST keeps running for >>>>>> the TEST session, so the next line of a CLIST after the TEST >>>>>> command itself can be subcommands such as AT and GO. >>>>>> >>>>>> As I remember it, there were important differences in SYSOUTTRAP >>>>>> as well. I had a small collection of REXX and CLIST that could be >>>>>> run in TEST to do things such as single step and redisplay >>>>>> registers, storage, code, PSW and so on, almost half way to a full >>>>>> screen debugger. >>>>>> >>>>>> I found the differences meant I had to mix REXX and CLIST to get >>>>>> the job done. >>>>>> >>>>>> Roops >>>>>>> On 3 Jul 2016 12:53, "Joseph Reichman" wrote: >>>>>>> >>>>>>> Thanks so much 🤗 >>>>>>> >>>>>>> Joe Reichman >>>>>>> 8045 Newell St Apt 403 >>>>>>> Silver Spring MD 20910 >>>>>>> Home (240) 863-3965 >>>>>>> Cell (917) 748 -9693 >>>>>>> >>>>>>>>> On Jul 3, 2016, at 7:38 AM, Mike Shorkend >>>>>>>> wrote: >>>>>>>> >>>>>>>> I am not sure about CLIST but for REXX, the following will >>>>>>>> place each output line into a the stem variable 'test'. >>>>>>>> >>>>>>>> /* REXX */ >>>>>>>> x=outtrap('test.') >>>>>>>> push "end" >>>>>>>> PUSH "go" >>>>>>>> PUSH "at +0" >>>>>>>> "test 'sys1.linklib(iefbr14)'" >>>>>>>> say "number of trapped lines is" test.0 do i = 1 to test.0 >>>>>>>> say test.i end >>>>>>>> >>>>>>>> HTH >>>>>>>> >>>>>>>> Mike >>>>>>>> >>>>>>>> >>>>>>>>> On 29 June 2016 at 15:21, Joseph Reichman >>>>>>> wrote: >>>>>>>>> >>>>>>>>> Hi >>>>>>>>> >>>>>>>>> Does any one know if capturing TEST output Into CLIST >>>>>>>>> variables is possible >>>>>>>>> >>>>>>>>> >>>>>>>>> Thanks >> ---------------------------------------------------------------------- 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
