On Mon, 18 Feb 2008, Zack Weinberg wrote:
> I'm not aware of any such problems - it could well be that I haven't
> tried to do whatever it was that was a problem, of course.
Regrettably, I can't remember, but it might have been quotes in node
names, or possibly underscores. It was a while ago.
> I sympathize with your desire not to change formats; unfortunately,
> no, an automated conversion is not feasible, because Texinfo is a much
> higher-level syntax than "man". For instance, the existing
> pcrepattern.3 uses a whole lot of tables done with simple spaces -
> that all has to get converted to explicit markup. (How do you do that
> for HTML?)
I leave them as simple spaces! :-) For example:
<pre>
(*CR) carriage return
(*LF) linefeed
(*CRLF) carriage return, followed by linefeed
(*ANYCRLF) any of the three above
(*ANY) all Unicode newline sequences
</pre>
> And there are a bunch of places where punctuation-type
> metacharacters aren't marked up at all in pcrepattern.3, but have to
> be in Texinfo.
Surely a simple Perl script can deal with that?
> I'm attaching my pcrepattern.texi so you can see what it looks like.
> This was about four hours of nontrivial human effort.
Apart from anything else, there's a practical reason for not doing
anything in a hurry, which is that my time is currently limited (even
though I'm retired). But alongside that, I'm not convinced that moving
to Texinfo as a basic format is "right".
What do others on this list think?
Philip
--
Philip Hazel
