Jeff writes:
> This is accomplished via a new parameter: job=job_id, where job_id can be:
>    [#]Jnnn
...
> I would like to extend job_id to include all jobs for a user.acct.  I imagine
> some of you already have developed a syntax to accomplish this.  My idea
> is:
>    [@J | @S:]user.acct
>   "[@J | @S:]jobname,user.acct"   <-- quotes needed for same reason
...
> With multiple job queues, this could be extened as:
>   [@J | @S:]user.acct[:jobQ]
> where a colon seperates the jobQ from the user.acct ID.
 
What about the obvious:
 
   #J@:user.acct
 
which maintains the reasonable standard of having the pattern match character
("@" in this case) appear where the non-pattern text would be.
 
The next enhancement is to allow any MPE pattern.  For example, I want to
wait for all (or is it *any one of*?) jobs in the range 120..129:
 
   #J12?:@.@
or, jobs in the range 120..129 with an account of SIELER:
 
   #J12?:@.SIELER
 
or, any job/session in range 120..129 with a user name matching ?STAN?,
and an account with "Z" in the name:
 
   #?12#:?STAN?.@Z@
 
In effect, the syntax could be:
 
   [optional leading "#"] [J, S, or ?] [digits or pattern]
 
   optionally followed by:  [:  [jobname, | jobname_pattern, ]
                                 [username | username_pattern] .
                                 [acctname | acctname_pattern] ]
 
--
Stan Sieler                                          [log in to unmask]
                                     http://www.allegro.com/sieler.html