Hi David--

    On the morning of 06-12, I sent you a message that I had found the
reference to the PL/I macro that emulates the REXX PARSE statemen (in
Issue #1 [December, 1994] of "The PL/I Connection"), and that piece
stated that a copy of Eberhard Sturm's macro could be obtained from the
author of the reference: Dave Jones (Member, Team PL/I) at
usqvhwvh{at}ibmmail.com.  On that same day, I did send the request to that
other Dave.  Lo and Behold!  On the morning of 06-13, I downloaded an
e-mail message from him appending the source code for the macro.
According to the signature, Dave Jones works for Velocity Software in
Houston, TX.  The following is an excerpt from that transmittal:

      "The PARSE PLI macro is attached to the end of this note.  While
   it was written for use on host (VM and MVS) PLI/I compilers, it
   should work unchanged on OS/2 PL/I, although I haven't tested it.
   Good luck, and enjoy."

And here is the intro comment in the source code:

 /*--------------------------------------------------------
  The PARSE macro is a full implementation of the REXX parse statement
  as far as is applicable. Only the keywords UPPER, VALUE, and WITH are
  allowed. The only difference to REXX is that items in the template are
  separated by commas instead of by blanks; e.g.,
        parse value (String) with (Fname, . , Fmode) upper;
  All of the following is supported in the templates:
       - Comments
       - literal patterns    ....'('
       - parsing into words  .... word1 word2 word3 ....
       - placeholders        .... the "dot" '.'
       - positional patterns .... +3, -1, .... 
       - variable patterns   .... (var_name)

  Written by:
      Eberhard Sturm                          +49-251-83-2609
      sturm{at}dmswwu1a.bitnet             sturm{at}uni-muenster.de
      Computing Center of the University of Muenster, Germany
  ---------------------------------------------------------*/

    I haven't printed out the source code yet to study it, nor have I
separated it from the message and tried to see if it would compile.  If
you would like a copy, either let me know and I will e-mail one to you,
or make t'other Dave's day (again) by requesting it directly from him.

    Regards,

          --Murray    (mlesser{at}ibm.net)

___
 * MR/2 2.25 #120 * Never send a PM program to do a character-mode job

--- Maximus/2 2.02