Re: [Groff] State of the MathML patch

[Eric S. Raymond]

Werner LEMBERG <address@hidden>:
> > Werner, I have a couple of minor improvements to the MathML patch;
> > notably, I've fixed a minor bug in the handling of inline equations
> > and another that attached an incorrect attribute to <mfrac>.  Would
> > you prefer to receive an incremental patch or a rollup?
> 
> Please send me the whole patch again; I haven't had time yet to look
> at it.

OK.  This time I'll add documentation patches.
 
> > Not having write access to the CVS is starting to be a pain.
> 
> No problem.  What's your Savannah account name?

esr

> > Teaching doclifter to use groff -TMathML has gone mostly quite
> > smoothly, but there turns out to be one nasty little issue.  A
> > couple of man pages embed numeric troff highlight escapes like \f2
> > in eqn formulas.  It's not at all clear what to do with these when
> > generating MathML.
> 
> A commented example, please.

Enclosed.  The relevant line is:

185:${log sub 2}( max (\f2width\fP,\f2height\fP))$.

See the \f2 and \fP instances in there?  One isn't really supposed to
do that; eqn has its own font keywords {roman, bold, italic}.  The
correct way to write the above would have been

        ${log sub 2}( max (italic width, italic height))$

that is, assuming font 2 is mapped to italic as usual.  The man page
author got away with it because in his environment they get passed
through and interpreted by troff.  doclifter tries to turn these 
into DocBook emphasis tags, and I end up with a tag mixture that doesn't
validate.

At the moment I am nuking these escapes and documenting that as a
doclifter limitation.  It's not clear that more effort is merited; 
I've only found three pages with this problem. But I'm open to
suggestions.
-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>

foobar.man
Description: Unix manual page