Re: snippet fails with 2.23.

[Thomas Morley]

Am So., 15. Mai 2022 um 21:44 Uhr schrieb Thomas Morley
<thomasmorley65@gmail.com>:
>
> Am So., 15. Mai 2022 um 14:11 Uhr schrieb Jean Abou Samra 
> <jean@abou-samra.fr>:
> >
> >
> >
> > Le 15/05/2022 à 14:01, Thomas Morley a écrit :
> > > Hi,
> > >
> > > the new doc-tagged LSR-snippet 'tambourine-example.ly'
> > > https://lsr.di.unimi.it/LSR/Item?id=1070
> > >
> > > %%%%%%
> > > \paper { tagline = ##f }
> > >
> > > #(define mydrums '((tambourine default #t 0)))
> > >
> > > \new DrumStaff \with { instrumentName = #"Tambourine" }
> > >
> > > \drummode {
> > >    \set DrumStaff.drumStyleTable = #(alist->hash-table mydrums)
> > >    \override Staff.StaffSymbol.line-positions = #'( 0 )
> > >    \override Staff.BarLine.bar-extent = #'(-1.5 . 1.5)
> > >
> > >    \time 6/8
> > >    tamb8. 16 8 8 8 8 |
> > >    tamb4. 8 8 8 |
> > >    % the trick with the scaled duration and the shorter rest
> > >    % is neccessary for the correct ending of the trill-span!
> > >    tamb2.*5/6 \startTrillSpan s8 \stopTrillSpan |
> > > }
> > > %%%%
> > >
> > > works fine with 2.22. but fails with master even after a run of
> > > convert-ly, returning:
> > >
> > > fatal error: unrecognised percussion sign: "#t"
> > >
> > >    tamb8. 16 8 8 8 8 |
> > >
> > > I don't get what's wrong.
> > > Hints?
> > >
> > > Thanks,
> > >    Harm
> >
> >
> > It works if you replace '((tambourine default #t 0)) with
> > '((tambourine default #f 0)). That's in line with the documentation:
> >
> > """
> > Existing notations may be redefined as an association list
> > where each entry has to be comprised of four items:
> > a name, the note head style (or @code{default}), an
> > articulation sign if needed (or @code{#f} if not), and
> > the note head's position on the staff.  That list must then
> > be converted into a Scheme hash table, using the
> > @code{alist->hash-table} function.
> > """
> >
> > This apparently changed with
> >
> > commit 61cd3bc1f3254b430bf04acd587c4082253602d4
> > Author: Lukas-Fabian Moser <lfm@gmx.de>
> > Date:   Mon Dec 27 01:25:43 2021 +0100
> >
> >      Make articulation-type a symbol? instead of a string?
> >
> > which contains
> >
> > diff --git a/lily/drum-note-engraver.cc b/lily/drum-note-engraver.cc
> > index 7e6e3cef7a..d4078a2e3b 100644
> > --- a/lily/drum-note-engraver.cc
> > +++ b/lily/drum-note-engraver.cc
> > @@ -93,12 +93,12 @@ Drum_notes_engraver::process_music ()
> >             if (scm_is_symbol (style))
> >               set_property (note, "style", style);
> >
> > -          if (scm_is_string (script))
> > +          if (scm_is_true (script))
> >               {
> >                 // Error out if script doesn't exist
> >                 if (scm_is_false (ly_assoc (script, get_property
> > (context (), "scriptDefinitions"))))
> >                   ev->origin ()->error (_f ("unrecognised percussion
> > sign: \"%s\"",
> > -                                          ly_scm2string (script)));
> > +                                          ly_scm_write_string (script)));
> >
> >                 Item *p = make_item ("Script", ev->self_scm ());
> >                 make_script_from_event (p, context (), script,
> >
> >
> >
> > Namely, instead of checking (string? the-second-element),
> > it just checks the-second-element, i.e. it searches the
> > script definition if the script is anything but #f.
> >
> > Jean
> >
>
> Hi Jean,
>
> thanks for the background.
> There were some other lsr-snippets with a #t-setting there, which had
> escaped my attention, bombing out 'make doc' after lsr-import.
>
> Next attempt now...
>
> Thanks,
>   Harm

Btw, I found some other snippets worth fixing, currently returning
programming errors or warnings:

incipit.ly  "programming error: Loose column does not have right side
to attach to."
using-marklines-in-a-frenched-score.ly a plethora of "programming
error: cyclic dependency:..."
using-tags-to-produce-mensural-and-modern-music-from-the-same-source.ly
"Loose column ..."

how-to-print-two-rehearsal-marks-above-and-below-the-same-barline-method-2.ly
and
vertical-aligned-staffgroups-without-connecting-systemstartbar.ly
return warnings for conficting rehearsal marks.

Worth fixing ofcourse, though I decided not to postpone finishing
lsr-update see !1360

And there's clip-systems.ly ...
It fails always with:
GNU LilyPond 2.23.9 (running Guile 2.2)
Processing `clip-systems.ly'
Parsing...
Interpreting music...
Preprocessing graphical objects...
Finding the ideal number of pages...
Fitting music on 1 page...
Drawing systems...
Converting to `clip-systems-1.pdf'...
Layout output to `clip-systems-1-from-2.0.1-to-4.0.1-clip.eps'...
Converting to `clip-systems-1-from-2.0.1-to-4.0.1-clip.pdf'...
Layout output to `clip-systems-1-from-0.0.1-to-4.0.1-clip.eps'...
Converting to `clip-systems-1-from-0.0.1-to-4.0.1-clip.pdf'...
Layout output to `clip-systems-1-from-0.0.1-to-6.0.1-clip.eps'...
Converting to `clip-systems-1-from-0.0.1-to-6.0.1-clip.pdf'...
Layout output to `clip-systems-1-from-0.0.1-to-6.0.1-clip-1.eps'...
Converting to `clip-systems-1-from-0.0.1-to-6.0.1-clip-1.pdf'...
Interpreting music...
Preprocessing graphical objects...
Interpreting music...
Preprocessing graphical objects...
fatal error: cannot find file: `clip-systems-1-from-2.0.1-to-4.0.1-clip.eps'
[...]

But it survives 'make doc'
I don't understand ...

Cheers,
  Harm