Hello Karl. > I doubt it would be possible to parse texinfo with lex and yacc, by > the way.
Because it's line-oriented? I do have doubts about yacc being useful here. As for tokenization, Perl's regexes are more than enough for the job. > Also, any changes to the language (or the Info output) must be 100% > backward compatible. That's why I thought of adding an extra @node arg. I noticed that makeinfo ignored it. Still have to check with texi2dvi. > Regarding Perl, it would make some things easier, but I doubt rms would > appreciate it. Which language do you think he would appreciate? > I should mention that the texi2html processor (not part > of the > Texinfo package, but independently and "friendly" maintained) is > > written in Perl. I didn't know it existed. I've had a 3 minute look at it and I have the impression that it doesn't build a parse tree either. Do you think it would be more amenable to my needs? Hmm. I'm sorely tempted to have a shot at a new converter. I would have a module that builds an AST; then converters that use multimethods like generate(command, output) to make it easier to take an existing converter and overried part of its behavior. The HTML converter would use HTML templates instead of hard-coded HTML... Ah, time... -- Jean-Louis Leroy Sound Object Logic http://www.soundobjectlogic.com _______________________________________________ Texinfo home page: http://www.gnu.org/software/texinfo [email protected] http://lists.gnu.org/mailman/listinfo/help-texinfo
