Re: improved lyrics handling

[Han-Wen]

address@hidden writes:
> >         \property LyricsVoice.LyricSyllable \set #'X-extent-callback =
> >               #'(0.0 . 4.3)  
> >         #"\Guo\043\n\t"
> >             ..
> > 
> >     }
> > 
> > (So \Zhoung is not how it works in TeX, but you get the idea), perhaps
> > using Scheme, since it makes encoding the strings more standard.
> 
> You mean `\Zhong', don't you :-)
> 
> [Actually, the pinyin style file from my CJK package for LaTeX accepts
>  those macros.]

(Owpps :)

> Anyway, I don't fully understand your example.  Do you mean a
> two-stage compiling process similar to, say, bibtex handling during a
> LaTeX run?  Please elaborate.

I mean generate dimensions using LaTeX/Omega/whatever before actual
lilypond typesetting begins. This can be done using a preprocessor
(crude), or by invoking a recursive lambda from the \score block
(smart)

This function should collect all texts, be them in lyrics or in
scripts, dump them into a tex file, run latex, extract all dimensions
in one go, and add the dimension information to each text-event again.

This approach has the advantage that it is done in "user-space", and
that it is very efficient: only one Latex run to be done before
lilypond starts processing.  If it had to be done during typesetting,
it is much more difficult to determine which texts to run through
latex. Invoking TeX separately for each text is not an option.

(Please don't say that you also want accurately dimensioned barnumbers
in Hindi :-) 

-- 

Han-Wen Nienhuys   |   address@hidden   |   http://www.cs.uu.nl/~hanwen