Better width functions for the current arpeggio print functions.

Better width functions for the current arpeggio print functions.

[MikeSol, 2011-05-11 02:10:31]

I had added a Rietveld issue about this, but the example on my Rietveld issue
was bunk so I pulled it.  The current width function in arpeggio.cc does not
give the correct width for chord brackets or chord slurs, nor does it give the
correct width for arpeggios with arrows (I think).  This should fix it w/o
causing problems w/ the print function triggering a vertical alignment.

It is kinda code-dupy, so if anyone has suggestions for how to roll this into
one unified function, I'll take it!

[Carl, 2011-05-11 13:51:28]

Seems reasonable to me.

I couldn't think of any way to generalize the internal call.

Carl

[mike_apollinemike.com, 2011-05-11 14:18:50]

On May 11, 2011, at 9:51 AM, Carl.D.Sorensen@gmail.com wrote:

> Seems reasonable to me.
> 
> I couldn't think of any way to generalize the internal call.
> 
> Carl
> 
> 
> http://codereview.appspot.com/4517051/


Thanks!

There is a note in arpeggio.cc saying that width cannot be gleaned from the
print function because it triggers a vertical alignment when arpeggios are
cross-staff.  By turning the function into an internal function and calling it
form outside functions, I'm assuming that this avoids triggering the vertical
alignment, but I'm not sure this is the case.  Does anyone know
how/where/when/why the vertical alignment is triggered?

Cheers,
MS

[Neil Puttock, 2011-05-11 16:01:36]

On 11 May 2011 15:18, mike@apollinemike.com <mike@apollinemike.com> wrote:

> There is a note in arpeggio.cc saying that width cannot be gleaned from the
print function because it triggers a vertical alignment when arpeggios are
cross-staff.  By turning the function into an internal function and calling it
form outside functions, I'm assuming that this avoids triggering the vertical
alignment, but I'm not sure this is the case.  Does anyone know
how/where/when/why the vertical alignment is triggered?

Probably when the print function reads 'positions:
ly:arpeggio::calc-positions requests the common refpoint in the
Y-axis, which will be a VerticalAlignment for cross-staff arpeggios.

Cheers,
Neil

[Neil Puttock, 2011-05-11 16:02:26]

http://codereview.appspot.com/4517051/diff/1/lily/arpeggio.cc
File lily/arpeggio.cc (right):

http://codereview.appspot.com/4517051/diff/1/lily/arpeggio.cc#newcode98
lily/arpeggio.cc:98: MAKE_SCHEME_CALLBACK (Arpeggio, internal_print, 1);
Why are you exporting these internal functions?

[mike_apollinemike.com, 2011-05-11 18:11:32]

On May 11, 2011, at 12:02 PM, n.puttock@gmail.com wrote:

> 
> http://codereview.appspot.com/4517051/diff/1/lily/arpeggio.cc
> File lily/arpeggio.cc (right):
> 
> http://codereview.appspot.com/4517051/diff/1/lily/arpeggio.cc#newcode98
> lily/arpeggio.cc:98: MAKE_SCHEME_CALLBACK (Arpeggio, internal_print, 1);
> Why are you exporting these internal functions?
> 
> http://codereview.appspot.com/4517051/

The internal function idea is now bunk - you're right about the positions thing.

The issue is that, for the chord bracket and chord slur (and Bertrand's eventual
chord brace, which hypothetically varies significantly in its X dimension as it
gets larger), the width of the grob is dependent on knowing the extremal note
head positions, which triggers a vertical alignment.  Any suggestions on how to
get this information without triggering the vertical alignment?

Cheers,
MS

[Neil Puttock, 2011-05-11 18:19:10]

On 11 May 2011 19:11,  <mike@apollinemike.com> wrote:

> The issue is that, for the chord bracket and chord slur (and Bertrand's
eventual chord brace, which hypothetically varies significantly in its X
dimension as it gets larger), the width of the grob is dependent on knowing the
extremal note head positions, which triggers a vertical alignment.

Surely not the chord bracket?  It has a fixed width (line 171 of
arpeggio.cc), though perhaps it should be tweakable.

> Any suggestions on how to get this information without triggering the vertical
alignment?

Sorry, no.

Cheers,
Neil

[mike_apollinemike.com, 2011-05-11 18:35:40]

On May 11, 2011, at 2:19 PM, Neil Puttock wrote:

> On 11 May 2011 19:11,  <mike@apollinemike.com> wrote:
> 
>> The issue is that, for the chord bracket and chord slur (and Bertrand's
eventual chord brace, which hypothetically varies significantly in its X
dimension as it gets larger), the width of the grob is dependent on knowing the
extremal note head positions, which triggers a vertical alignment.
> 
> Surely not the chord bracket?  It has a fixed width (line 171 of
> arpeggio.cc), though perhaps it should be tweakable.
> 

Sorry - yes, you're right.  I amend my comment to chord slurs and braces.

>> Any suggestions on how to get this information without triggering the
vertical alignment?
> 
> Sorry, no.
> 

I'll try to think of something...

~Mike