ChordNames break synchronization

[aspiers]

The ChordNames context does not have an impact on MIDI rendering in LilyPond, but it does result in multiple note events being rendered as a single text grob (e.g. "Eb-7") whose PDF annotation links back to a location in the .ly source which looks like a normal note e.g. ef1:m9, which the tokenizer currently cannot distinguish from a normal note.

This makes it almost impossible for the synchronization algorithm to work. It's cleaner just to remove chords from the equation. This can easily be done with the MIDI by creating a separate \score which excludes the chords, but they still appear in the PDF.

[aspiers]

I kicked off a thread on lilypond-user about this, and eventually discovered that it's possible to hack the definition of the grob-cause function in LilyPond's output-ps.scm, since

(assoc-get 'name (ly:grob-property grob 'meta))

will return 'ChordName when the grob is a ChordName. This can then be used to avoid generating point-and-click annotations for chord names. It may also be possible to do it by making the tokenizer smarter, but it would probably have to be able to convert chords like ef1:m9 into the full set of notes which would appear in the MIDI, which would require reinventing the LilyPond chord handling wheel.

[aspiers]

Moving this to 0.3 due to the available workaround, i.e. exclude the chords from MIDI generation, and hack output-ps.scm to avoid annotating ChordNames. I'll aim to convert the latter hack into a more elegant solution involving \pointAndClickTypes or similar, and push it back upstream.

[aspiers]

I have an even better solution cooking, thanks to Jan Nieuwenhuizen who I met at FOSDEM and who helped me implement it! It's in the spacetime branch, but not quite there yet.

[aspiers]

For reference, here's the key Scheme code at the heart of the new algorithm.