Re: Issue 1110 in lilypond: Wrong octave of repetition chord with \relative and #{ #} syntax

[lilypond]

Comment #30 on issue 1110 by address@hidden: Wrong octave of repetition chord  
with \relative and #{ #} syntax
http://code.google.com/p/lilypond/issues/detail?id=1110

Now since issue 2240 aka 2270 is on the countdown, it makes some sense to  
think about this one again.  We now have the situation that EventChord is a  
satisfactorily reliable indicator of an actual chord.

I see two ways to deal with this issue now.

a) have q work in the parser (as it does now), let \relative completely  
reconsider the relation between q and its preceding chord.  This is  
actually pretty ugly.

b) have q never record any pitches (it won't be fazed by \relative then).   
Instead, the iterator for the repeated chord fills in the pitches.  The  
question is just where it gets them from.  It can't bother the EventChord  
iterator for that since in the case of << { c <c d e> } q >> the  
event-chord-iterator is called later than the corresponding repeated chord  
iterator.  Basically we need to comb through the whole expression that is  
going to be iterated, and fill in the repeated chords.

\relative does its respective work on demand.  How can we avoid unnecessary  
work?  Do it on demand as well?

\repeatChords \relative c' << { c <c d e> } q >>
Then the user would be responsible not to apply \relative after  
\repeatChords...

That would be a very straightforward solution without extra cost if you did  
not actually use the feature.  Normally it would be sufficient to put the  
contents of your \score into a single \repeatChords.  I doubt that this  
will make people happy...