On 01/07/13 at 05:24pm, Glen Larsen wrote:
> Very cool, David.
> Felix, I knew you'd want something like this made out of that:
> Simple python wrapper for this scheme code <https://gist.github.com/4480245>
Yeah, I think it still needs some processing. IMO, The header fields should be
checked for their validity and the remaining fields written to a preliminary
(incomplete) .rdf-file. (They are only required for the .rdf aren't they?)
Your wrapper looks like a good start.
>From reading David's mail I realised that header fields can also contain
>newlines,
which could mess up parsing. I'd replace those with nulls in the scheme and
change
them back with the python. (If there are nulls in header fields this will go
wrong,
but I don't really see an application for having nulls in header fields.) So
varying David's command line a bit
lilypond -e '(set! print-book-with-defaults (lambda (parser book)
(module-for-each (lambda (sym var) (if (markup? (variable-ref var))
(format #t "~a=~a~%" sym
(string-map (lambda (c) (if (eq? c #\newline) #\nul c))
(markup->string (variable-ref var)))))) (ly:book-header book))))'
file.ly
replaces newlines by nulls and prints header fields literally without escaping
characters. (Otherwise the nulls would be displayed as "\x00".)
Python's xml.sax.saxutils.escape looks useful for producing the .rdf format.
Felix
PS: Is supporting multiple files really necessary for the script? If there is a
multiple
file piece the relevant fields should be equal for any of the files (which are
not
only for inclusion in other files). Ok, one could compare the output for the
different
files and check whether that's really true; this doesn't seem to be worth the
time, though.
Raising this issue is probably also not worth the time, since it doesn't seem
to make
the script much more complicated and doesn't do any harm at all...
_______________________________________________
Mutopia-discuss mailing list
Mutopia-discuss@mutopiaproject.org
http://lists.bcn.mythic-beasts.com/mailman/listinfo/mutopia-discuss
