>> To sum up: Keywords like `orientation' are far too frequent in >> normal text to be handled without protection if convert-ly is >> applied to lilypond-book (or texinfo) input. > > I'm fine with this being a command-line option to convert-ly. > Actually, this could even be the default -- as long as there's a > (CG-documented) way to apply convert-ly to the entire document, I'm > happy with this change.
Unfortunately, I'm not a python guru, and handling balanced text blocks with regular expressions is notoriously difficult. Contrary to Perl, which has support for that since version 5.10 (a special syntax extension of the `+' operator and the `(?n)' construct, cf. http://www.perlmonks.org/?node_id=660316), Python lacks this, AFAIK. There is the highly-praised pyparsing package, however, its license forbids inclusion into lilypond. Werner _______________________________________________ lilypond-devel mailing list [email protected] http://lists.gnu.org/mailman/listinfo/lilypond-devel
