Re: self-alignment-X has the opposite effect on Lyric syllables narrower than the note-head

[Aaron Hill]

On 2022-01-20 3:05 am, Simon Albrecht wrote:
> when setting self-alignment-X on a lyric syllable that is narrower
> than its corresponding note-head, it takes the opposite effect of what
> should be expected:
>
> %%%%%%%%%%%%%%%%%%
> \version "2.23.5"
>
> <<
>   { 1 1 }
>   \addlyrics {
>     \tweak self-alignment-X #3 test
>     \tweak self-alignment-X #3 i
>   }
> > >
> %%%%%%%%%%%%%%%%%%

Firstly, alignment values outside the interval [-1, 1] in general behave 
oddly as the result depends on the size of the object in question.  
Larger things do end up moving more.

Secondly, LyricText has parent-alignment-X set to '() which, as far as I 
can tell, results in inheriting self-alignment-X.  Choosing #RIGHT, for 
example, means to have the right side of the text align with the right 
edge of the note.

Combine those quirks together and do you get some very unusual results 
with out-of-range alignments.  A value of (* 3 RIGHT) would mean you are 
aligning to a point much to the right of the note in question.  And the 
same so-called "super-right-aligning" of the text also means its 
reference point is beyond the right bound.  Wider text like "test" has 
the reference point much farther to the right, pushing it farther left 
compared to "i".

Try setting parent-alignment-X to, say, #CENTER (0) to remove a variable 
from the equation.  "test" will still move more than "i", but it should 
hopefully be easier to see that both texts are moving in the same 
direction (relative to the note center).

Perhaps an unspecified parent-alignment-X should take self-alignment-X's 
value but clamped to [LEFT, RIGHT]:

%%%%
\override LyricText.parent-alignment-X =
  #(lambda (grob) (max LEFT (min RIGHT (ly:grob-property grob 
'self-alignment-X))))
%%%%


-- Aaron Hill