Re: Completion_heads_engraver doesn't play nicely with lyric alignment

[Alexander Kobel]

Some more investigation:

in Lyric_engraver::process_music, the LyricText's self-alignment-X is 
set according to lyricMelismaAlignment. If I add a separate check for 
completionBusy there, things change:

lyric-engraver.cc:84:

  if (last_text_
      && voice
      && (to_boolean (voice->get_property ("melismaBusy"))
          /* the following is new */
          || to_boolean (voice->get_property ("completionBusy")))
      && !to_boolean (context ()->get_property ("ignoreMelismata")))
    last_text_->set_property ("self-alignment-X",
                              get_property ("lyricMelismaAlignment"));

But:

1.) I don't know why this changes anything at all - completionBusy is in 
the default list of melismaBusyProperties, and I assume that means it 
*should* be acknowledged by default anyways. Cause apparently, the 
single "melismaBusy" captures all other properties from that list, 
including Ties, Slurs, and explicit melisma?

2.) The modified variant is too eager: it also left-aligns non-splitted 
notes if there are more fine-grained events in a different voice:

\version "2.19.82"

\paper { ragged-last = ##f }

\score {
  <<
    { c'4 c'4 c'4 c'4 c'4 c'4 c'2 }
    { c'4 c'4 c'4 c'4 c'2 c'2 }
    \addlyrics { These should all be center-aligned. __ Right? }
  >>
}

\score {
  <<
    { c'4 c'4 c'4 c'4 c'4 c'4 c'2 }
    { c'4 c'4 c'4 c'4 c'2 c'2 }
    \addlyrics { These should all be center-aligned. __ Right? }
  >>
   \layout {
     \context {
       \Voice
       \remove Note_heads_engraver
       \consists Completion_heads_engraver
     }
   }
}

gives the attached output with the above change in lyric-engraver.cc. So 
apparently, completionBusy is set because when encountering the first 
half note, it's not yet clear whether the note *might* be split later 
on, and that decision is only taken on the moment which doesn't have 
anything to do with this voice.
In other words, completionBusy tells whether the 
Completion_heads_engraver is running behind the scenes, not whether it 
actually changes anything.

But in that case, completionBusy should not belong to the 
melismaBusyProperties?

Something's fishy here, AFAICS...


Cheers,
Alex


On 20.05.19 13:58, Alexander Kobel wrote:
> Hi all,
>
> I'd expect the two scores in the following excerpt to be identical, alas 
> they aren't: Notes splitted into tied notes by the 
> Completion_heads_engraver produce melismata as expected, but don't cause 
> the Lyrics to be left-aligned accordingly. (See the attached output.)
>
>
> \version "2.19.82"
>
> \paper { ragged-last = ##f }
>
> \score {
>    { c'4 c'4 c'4 c'4~ c'4 }
>    \addlyrics { This should be left-aligned. __ }
> }
>
> \score {
>    { c'4 c'4 c'4 c'2 }
>    \addlyrics { This should be left-aligned. __ }
>    \layout {
>      \context {
>        \Voice
>        \remove Note_heads_engraver
>        \consists Completion_heads_engraver
>      }
>    }
> }
>
>
> Unfortunately, this seems to imply that there is no easy/clean way to 
> produce a Mensurstriche layout and a "standard" layout with splitted 
> notes from the same source.
> In case this is a known (but, in this case, AFAICS undocumented) flaw, 
> or unlikely to have an easy fix: I'd be happy to hear about an 
> workaround for same-source ditions, if someone has one in mind.
>
>
> Cheers,
> Alex
>
> _______________________________________________
> bug-lilypond mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/bug-lilypond

example2.png
Description: PNG image

smime.p7s
Description: S/MIME Cryptographic Signature