Re: wrong-type-argument

[Bruce Korb]

"Gary V. Vaughan" wrote:

> > "ifnohtml" is not documented.
> 
> Sure it is.

No, it is not documented where I was looking for it.  :-)
I found it by grepping the info doc.

> > > @iftex
> > > @ifhtml
> > > THIS APPEARS IN THE HTML OUTPUT, BUT NOT PRINTED TEXT!!
> 
> I find this more than a little surprising!  Perhaps a bug report to
> the texinfo mailing list is in order?  I can dig the address out of my
> mail archive if you can't find it.
> 
> > > @end ifhtml
> > > @multitable ...
> > > @end iftex
> >
> > > @ifinfo
> > > @multitable ...
> > > @end ifinfo
> >
> > Which yielded the page you see here:
> >
> >   http://AutoGen.sourceforge.net/doc/autogen_7.html#SEC165
> >
> > (which is generated, of course  :-)
> 
> And looks much nicer -- I still see the mangled texinfo instructions
> in your web page though...

That is exactly the problem.  Attached is the generated .texi fragment
used to produce the above page.  Notice that the table stuff is
all properly wrapped with the @ifxxx stuff.  I have cc-ed the bug list.
The best fix would be to have the html-izer handle the multitable,
but I will settle for the @iftex/@end keeping the text out of the
web output.

> I still have an AutoOpts task languishing in my TODO list which would
> fix one of the few disadvantages you list.  I would really like to
> have a template that parses AutoOpts .def files, but generates a
> function that relies only on getopt.c and getopt1.c (and the function
> it generates ofcourse).

It would not be terribly hard.  I just don't really know what it
would buy you.  You would not have to "depend" on the presence of
libopts, but you would have to generate the equivalent functionality
and compile it into the program.  Not clear that that is a win.

The only "negative" in the list that I see is the
range check.  If someone really wanted it, then adding a "arg_range"
attribute with a generated callback procedure (a la the enumeration
type handler) would be trivial.  Nobody has asked.

> Also, out of interest, what is the `makefile' row for?

No idea.  The output for each "competitor" is static:

> <TR><TD>makefile</TD>[=
> FOR competitor =]
>     <TD>no</TD>[=
> ENDFOR =]</TR>

You'd have to ask Jim Van Zandt what he was thinking about.
He's the author of the autoproject stuff.  Perhaps he was
thinking about rule construction for building the parser from
the definitions?  Don't really know.

compete.texi
Description: TeXInfo document