Reinhold Kainhofer a écrit : > Here's a further update of the open issues with texi2html for lilypond's > documentation:
Very cool! > Now, we can simply use a @translationof macro in each section to > give the original section title (to be used for the filename). E.g.: > @node Translated chapter title > @chapter Translated chapter title > @translationof Spacing issues > will create a file and anchor from "Spacing issues", but display the > translated chapter title. > In the init file I'll then simply ignore the whole macro, so for texi2html we > wouldn't even need it defined at all, but of course, for all other output > programs, we need it. > > The problem that I see in general are cross references (they are a problem > even with multiple nodes in one file... How do we get the correct file name > for a cross-reference, if it cannot be deduced from the node title at > all????). How should be define the @r* macros to work Here's a possible solution -- I don't speak a word of Perl and haven't looked at the code, maybe something similar is already implemented. @translationof is parsed just like @node, and (@node,@translationof) pairs are stored in a map or a list, or whatever the right Perl data structure for maps. Then, it's easy to deduce the desired file name when writing HTML output for a x-ref. > The language passed to texi2html as --lang=$(ISOLANG) is now added to the > extension as .$(ISOLANG).html. I'm not sure this will work well with add_html_footer.py hacking, especially extensions stripping for automatic language selection: does your init script writes hrefs with .$(LANG).html too? If it does, add_html_footer.py will likely need some hacking to handle file names which already have a language suffix. Thanks for the news and the good work John _______________________________________________ lilypond-devel mailing list lilypond-devel@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-devel
