Re: Bracketized key signature: \key c \major causes a core dump

[David Kastrup]

Rutger Hofman <address@hidden> writes:

> Good morning list,
>
> I use the snippet http://lsr.di.unimi.it/LSR/Snippet?id=647 to get a
> key signature in brackets. This works fine for \key other than c
> \major, but if I specify c \major I get a core dump:
>
> Preprocessing graphical objects...lilypond:
> /home/gub/NewGub/gub/target/linux-64/src/lilypond-git.sv.gnu.org--lilypond.git-release-unstable/lily/skyline.cc:94:
> Building::Building(Real, Real, Real, Real): Assertion `start_height ==
> end_height' failed.
> Aborted (core dumped)
>
> Rutger Hofman
> Amsterdam
>
> \version "2.19.35"
>
> {
>     % http://lsr.di.unimi.it/LSR/Snippet?id=647
>     \override Staff.KeySignature.stencil = #(lambda (grob)
> (bracketify-stencil (ly:key-signature-interface::print grob) Y 0.1 0.2
> 0.1))
>     \key c \major
>     c''
> }

Well, apparently the code/snippet is not all that well-prepared for what
amounts to an empty key signature.  However, the triggered problem is
deep in the bowels of LilyPond and is a complaint about inconsistent
internal data.  This error should not be triggerable from user code.

While the user code and/or the code underlying bracketify-stencil might
warrant catering for this particular case, the internals must still not
be confused in this manner.  So this will likely warrant fixes in more
than one place even though the symptom might be curable by just changing
the definition of bracketify-stencil.

-- 
David Kastrup