Paul Gilmartin writes:
<begin snippet>
I haven't coded a WTO lately. IIRC (much of) the cruft mentioned arises from
the unforgivable misdesign of placing options after a variable-length text
argument. It would have been so easy to do it right.
</end snippet>
If this means what it appears to mean it reflects 'radical' ignorance of HLASM
syntax.
The HLASM imposes no requirement that keyword-parameter values follow (or, of
course, precede) positional-parameter ones. It does impose the obvious,
essential requirement that positional-parameter values appear in a macro
instruction in the same sequence in which they appear in its macro definition;
but that is another matter.
Consider the skeletal macro definition
| macro
| EXAMPLE &p1,&k1=,&p2,&k2=
| . . .
| mend
EXAMPLE macro instructions may take such forms as
| EXAMPLE k2=<value>,<p1 value>,k1=<value>,<p2 value>
or
| EXAMPLE k1=<value>,k2=<value>,<p1 value>,<p2 value>
What may NOT be written is, for example,
| EXAMPLE p2,p1, . . .
or in general any macro instruction in which for i < j positional parameter j
precedes positional parameter i; but this is a sui generis restriction on
positional parameters.
In short, no unforgivable misdesign occurred here.
That said, Mr Gilmartin's comment is not without interest. I suspect that
others share his misapprehension, and an irreal restriction is just as daunting
to someone who thinks he must comply with it as a real one.
Or again, his complaint may be against such syntax as
TEXT=(('<text>',linetype)[,('<text>',linetype),]. . . )
If so, while it may be conceded that parenthesized lists comprised of
parenthesized sublists of positional parameters are not for the faint of heart,
two-element sublists (which can be split among lines for readability) should be
manageable by anyone who has any business writing a WTO macro instruction. The
task of ordering of two elements, variable in length or not, should be feasible
in almost any context, even for the severely challenged in that context:
Skinner would have us believe that pigeonms can do it reliably.
I conclude that that the best verdict Mr Gilmartin can hope for is the Scots
one, not proven.
John Gilmore Ashland, MA 01721-1817 USA
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
