"xcmd" is no stage, it just is a command that STARMSG will issue just after
its start.
I've got an idea, I don't know if it is easily implemented: pass a command
to STARMSG with a known response. when that response arrives, send it to
some other stream where you uses SPECS to replace it by the command to
execute, that you pass to a COMMAND (or CP) stage that you give a secondary
output to trap the returncode.
Something like:
'!STARMSG comamndWithKnowResponse',
'!L: NLOCATE /the_expected_response',
.....
'?L:!SPEC /'xcmd'/ 1',
'!Cmd: COMMAND',
'?Cmd:!Var CmdRc',
...
2011/2/8 Tom Huegel <tehue...@gmail.com>
> Thanks David, I thought that was the flow, but I didn't get it to work..
> Maybe I was just tired. Between shoveling snow and watching the Super Bowl
> I'm ready for a vacation.
>
>
> On Tue, Feb 8, 2011 at 5:53 AM, Dave Jones <d...@vsoft-software.com>wrote:
>
>> Tom, if the "xcmd" stage has a secondary output stream defined,
>> Pipelines will write the return code from the cp/cms command there.
>>
>> Does that help?
>>
>> DJ
>>
>> On 02/07/2011 02:09 PM, Tom Huegel wrote:
>> > I am trying to replace a WAKEUP (IUCVMSG routine with a PIPE STARMSG.
>> > I have the following snippet that seems to work fine except for
>> capturing
>> > the return code from the cp/cms command (XCMD).
>> >
>> > Is there a simple fix to this? - Thanks
>> >
>> > /*
>> > */
>> >
>> > trace
>> > o
>> > cp set msg
>> > iucv
>> > arg
>> > xcmd
>> > 'pipe (endchar
>> > ?)',
>> > '| starmsg
>> > ',
>> >
>> > xcmd,
>> > '| specs 9-16 1 17-* strip nw x15
>> > nw',
>> > '| stem linecount.', /* save results
>> > */
>> > '?', /* start of second pipeline
>> > */
>> > 'literal +01', /* Delay time
>> > */
>> > '| delay', /* Delay
>> > */
>> > '| pipestop' /* Stop
>> > */
>> > say
>> > rc
>> >
>> > cp set msg
>> > on
>> > 'pipe stem linecount. |
>> > console'
>> > exit
>> >
>>
>> --
>> Dave Jones
>> V/Soft Software
>> www.vsoft-software.com
>> Houston, TX
>> 281.578.7544
>>
>
>
--
Kris Buelens,
IBM Belgium, VM customer support
