On Fri, 27 Mar 2020 17:42:30 +0000, Seymour J Metz <[email protected]> wrote:
>Even if you accept "the sky's the limit" for undefined behavior, there's still
>the undocumented issue of whether it is permissible to include a leading
>ampersand or a trailing period with the symbol, which is permissible in the
>SYMDEF statement. Also, I'd like to see whether the following behaves as
>documented:
>
> foo = SYSNAME
> say MVSVAR('SYMDEF','foo')
>
>
>--
>Shmuel (Seymour J.) Metz
>http://mason.gmu.edu/~smetz3
Straight from REXXTRY under TSO on a z/OS V2.2 system:
foo = 'SYSNAME'
...............................................................................
...................................................
Say MVSVAR('SYMDEF',foo)
SYS0
...............................................................................
...................................................
foo = 'INVALID'
...............................................................................
...................................................
Say MVSVAR('SYMDEF',foo)
...............................................................................
...................................................
say '"'MVSVAR('SYMDEF',foo)'"'
""
I did that last one to show that if the Symbol Name does not exist, then SYMDEF
returns a null value. &symbol is for use in JCL. SYMDEF needs the Name of the
Symbol.
--
Dale R. Smith
Factotum
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
