Hi all,
I've written a set of REXX edit macros/execs that might be useful for
documentation/presentation purposes. They basically parse the member
(PL/I, REXX, assembler, JCL and, very limited, output of SuperC) being
viewed/edited and turn it into a format that is for all intents and
purposes the same as when the member is looked at in ISPF's Edit or
View using HI RESET + HI PAREN - there are a few small differences,
mainly because I (hopefully) correctly parsing JCL in-stream data, no-
operand assembler opcodes/directives and removed the SuperC bug where
the last two columns are displayed in green. :)
The set of (currently) eight macros consists of:
EHIASM - convert assembler to Edit HILITE'd HTML
EHIHELP - general help 'screen'
EHIJCL - convert JCL to Edit HILITE'd HTML
EHINONE - convert any text to Edit HTML (no HILITE)
EHIPLI - convert PL/I to Edit HILITE'd HTML
EHIREXX - convert REXX to Edit HILITE'd HTML
EHISUPC - convert SuperC to Edit HILITE'd HTML (as yet limited to what
I required...)
EHISUPP - general support (generate the actual HTML)
BTW, besides working as edit macro under ISPF (and as TSO command) on
z/OS, they also work on Regina.
EHISUPP contains Javascript code to faithfully emulate an 80-character
wide ISPF VIEW screen with (obviously non-working) action bars,
horizontal and vertical scrollbars (the horizontal one updates the
column numbers) and non-left/right scrolling '... *** Top/Bottom of
Data ***...' lines. At my request (I actually paid for it for the
code!), the developer of this code, Derek Pattenson <http://
www.smallofficesolutions.co.uk/>, released it under GPL V3 , which may
imply that if you use this fancy output (_it is also possible to
select plain HTML_) that your code will also be released under GPL
V3... (Only a "problem" if you put it on a _publicly_ accessible
website)
NB, the fancy output works OK in W3C standards based Firefox, but in
IE(6) it's slower than a snail on a very large dose of valium... (The
reason is that it uses 'white-space:pre' CSS rather than <pre>...</
pre> HTML tags)
As for the generating speed? EHIASM parses 'SYS1.MACLIB(CBDZPARS)' in
about 17 seconds and converting the parsed data into fancy HTML takes
another 15. A PL/I program of almost 11,000 lines takes around the
same total time. (Both on z/OS 1.7)
If you're interested in getting copies of the code (I think at almost
6,300 lines it's a bit to big post here) drop me a line, do not use
the bigfoot adress, but _ONLY_ use
x2c(88899389A3857C97998995964B9585A3) (EBCDIC), put '[HILITE]'
somewhere in the subject AND DO NOT SEND EMAIL IN HTML FORMAT!
Indicate if you want the code as eight zipped text files or as a PDS
in zipped XMIT format and make sure that any return message with
'[HILITE]' in the subject isn't stopped (or worse, returned to me) by
any spam filters.
I consider the code to be of RC-1 quality, but I would appreciate any
bug reports. And if you fancy any other EHIxxx'es, drop me a line, I'm
likely to do an EHIPAS as it's probably very similar to EHIPLI/
EHIREXX, but an EHICOBOL is probably trickier, as I don't know
Cobol...
Finally, the code is released under GPL V3, I plan to put it on
Sourceforge in the near(ish) future, but if you find any problems with
it before then, please drop me a line!
Robert
--
Robert AH Prins
x2c(88899389A3857C97998995964B9585A3)
|
