Re: [frogs] tablature: ties and harmonics (issue1669041)

[Marc Hohl <marc@xxxxxxxxxx>]

n.puttock@xxxxxxxxx schrieb:
> http://codereview.appspot.com/1669041/diff/1/2
> File input/regression/tablature-tie-harmonic.ly (right):
>
> http://codereview.appspot.com/1669041/diff/1/2#newcode3
> input/regression/tablature-tie-harmonic.ly:3: \header{ texidoc =
> "Harmonics in tablature follow the rules for
> \header {
>   texidoc = "Harmonics in tablature follow the rules for
> fret numbers in tie situations."
> }
Done.
> http://codereview.appspot.com/1669041/diff/1/3
> File ly/engraver-init.ly (right):
>
> http://codereview.appspot.com/1669041/diff/1/3#newcode748
> ly/engraver-init.ly:748: %% the HarmonicParenthesesItem engraver has to
> listen to the
> HarmonicParenthesesItem grob
>
> http://codereview.appspot.com/1669041/diff/1/3#newcode750
> ly/engraver-init.ly:750: \override HarmonicParenthesesItem
> #'after-line-breaking =
> #harmonic-parentheses-item::apply-tab-note-head-visibility
> It'd be simpler to use 'transparent (see my comments concerning
> 'whiteout vs 'transparent below)
Ok, done. But at the moment, 'whiteout and 'transparent can be
used independently - if you compile the regression file, you see
white spots. What is the right way to solve this?
> http://codereview.appspot.com/1669041/diff/1/4
> File scm/define-grobs.scm (right):
>
> http://codereview.appspot.com/1669041/diff/1/4#newcode871
> scm/define-grobs.scm:871: (angularity . 1.5)
> These (angularity, white-padding & width) need documenting in
> define-grob-properties.scm and adding to an interface in
> define-grob-interfaces.scm (either a new interface or
> parentheses-interface.)
Ok, done.
> http://codereview.appspot.com/1669041/diff/1/4#newcode872
> scm/define-grobs.scm:872: (half-thickness . 0.05)
> This should come from line-thickness.
Done.
> http://codereview.appspot.com/1669041/diff/1/4#newcode878
> scm/define-grobs.scm:878: (width . 0.5)
> could be more specific
I replaced it by "bracket-width" globally.
> http://codereview.appspot.com/1669041/diff/1/5
> File scm/output-lib.scm (right):
>
> http://codereview.appspot.com/1669041/diff/1/5#newcode505
> scm/output-lib.scm:505: (let* ((lp (draw-angled-bracket-stencil LEFT
> grob))
> indent
Sorry, done. (I moved the code into the helper function and didn't check 
the indentation)
> http://codereview.appspot.com/1669041/diff/1/6
> File scm/tablature.scm (right):
>
> http://codereview.appspot.com/1669041/diff/1/6#newcode146
> scm/tablature.scm:146: (let* ((articulations (ly:event-property
> (event-cause grob) 'articulations))
> let
Yes, of course.
> http://codereview.appspot.com/1669041/diff/1/6#newcode150
> scm/tablature.scm:150: (if (eq? 'harmonic-event (ly:event-property art
> 'class))
> (ly:in-event-class? art 'harmonic-event)
Done. Is this function documented somewhere? I didn't find it.
> http://codereview.appspot.com/1669041/diff/1/6#newcode157
> scm/tablature.scm:157: (half-thickness 0.05) ;; urg, hard-coded values
> Hmm, tricky.  It'll be difficult (rather convoluted) to get from
> TabNoteHead -> HarmonicParenthesesItem.
I don't know whether it is worth the effort (probably no-one wil
touch the harmonic angle brackets), but in general it is not
very elegant to store the values in one place while not being able
to use them globally.
> (not half-thickness though, since this should come from line-thickness
> and ly:output-def-lookup)
Done.
> http://codereview.appspot.com/1669041/diff/1/6#newcode161
> scm/tablature.scm:161: (dirp (ly:stencil-aligned-to
> could do with a more descriptive name
Done.
> http://codereview.appspot.com/1669041/diff/1/6#newcode291
> scm/tablature.scm:291: (display "\ntransparent: ")(display
> tab-note-head-transparent)
> remove debug code
Done.
> http://codereview.appspot.com/1669041/diff/1/6#newcode293
> scm/tablature.scm:293: ;; we simply remove the stencil, so we don't have
> to care about 'whiteout
> I've thought about this for a while, and it's silly that 'whiteout and
> 'transparent aren't mutually exclusive: if something's transparent, its
> whiteout shouldn't be visible.
>
> (see grob.cc)
You are right - can this behavior be changed in grob.cc to be valid
for all grobs in lilypond? If yes, how?

Thank you

Marc

http://codereview.appspot.com/1669041/show


---
----
Join the Frogs!