Adds ly:define-event-class

Adds ly:define-event-class



Create ly:define-event-class

[Neil Puttock, 2010-08-28 19:56:10]

Hi Mike,

I don't trust Rietveld not to swallow my comments, so I'm posting a few at a
time just in case it breaks again. :)

Cheers,
Neil

http://codereview.appspot.com/1867050/diff/17001/18001
File input/regression/scheme-text-spanner.ly (right):

http://codereview.appspot.com/1867050/diff/17001/18001#newcode5
input/regression/scheme-text-spanner.ly:5: texidoc = "uses define-event-class to
create a fully functional
A more generic docstring would be better, since you're exercising several
functions useful for implementing spanner engravers.

http://codereview.appspot.com/1867050/diff/17001/18001#newcode61
input/regression/scheme-text-spanner.ly:61: (side-axis . ,Y)
remove this to check `set-axis!' works

http://codereview.appspot.com/1867050/diff/17001/18001#newcode65
input/regression/scheme-text-spanner.ly:65: (Y-offset .
,ly:side-position-interface::y-aligned-side)
remove (same as side-axis)

http://codereview.appspot.com/1867050/diff/17001/18001#newcode111
input/regression/scheme-text-spanner.ly:111: (ly:grob-set-property!
(begin
  (ly:grob-set-property! grob 'side-axis axis)
  (ly:grob-chain-callback ...

[Neil Puttock, 2010-08-28 20:18:58]

http://codereview.appspot.com/1867050/diff/17001/18001
File input/regression/scheme-text-spanner.ly (right):

http://codereview.appspot.com/1867050/diff/17001/18001#newcode201
input/regression/scheme-text-spanner.ly:201: a b c \schemeTextSpannerEnd d |
+ another spanner to check ly:engraver-announce-end-grob works, e.g.,

\override SchemeTextSpanner #'to-barline = ##t
a4\schemeTextSpannerStart b d c |
a1\schemeTextSpannerEnd |

[Neil Puttock, 2010-08-28 20:22:04]

My comments in define-event-classes.scm don't work, so here goes:
+(define-public (define-event-class leaf heritage)
+ (cond
+ ((not (eq? leaf (car heritage)))
+ (ly:warning "All classes must be the last in their matrilineal line."))

(ly:warning (_ "...") for translations

+ ((not (equal? (cdr heritage)
+ (list-head (hashq-ref ancestor-lookup (cadr heritage) '())
+ (length (cdr heritage)))))
+ (ly:warning "All classes must have a well-defined pedigree in the existing
class heirarchy."))

(_ "...")

hierarchy

+ (else (hashq-set! ancestor-lookup
+ leaf
+ (cons leaf
+ (hashq-ref ancestor-lookup
+ (cadr heritage)
+ '()))))))

[MikeSol, 2010-08-28 21:12:25]

http://codereview.appspot.com/1867050/diff/17001/18001
File input/regression/scheme-text-spanner.ly (right):

http://codereview.appspot.com/1867050/diff/17001/18001#newcode5
input/regression/scheme-text-spanner.ly:5: texidoc = "uses define-event-class to
create a fully functional
On 2010/08/28 19:56:11, Neil Puttock wrote:
> A more generic docstring would be better, since you're exercising several
> functions useful for implementing spanner engravers.

Done.

http://codereview.appspot.com/1867050/diff/17001/18001#newcode61
input/regression/scheme-text-spanner.ly:61: (side-axis . ,Y)
On 2010/08/28 19:56:11, Neil Puttock wrote:
> remove this to check `set-axis!' works

Done.

http://codereview.appspot.com/1867050/diff/17001/18001#newcode65
input/regression/scheme-text-spanner.ly:65: (Y-offset .
,ly:side-position-interface::y-aligned-side)
On 2010/08/28 19:56:11, Neil Puttock wrote:
> remove (same as side-axis)

Done.

http://codereview.appspot.com/1867050/diff/17001/18001#newcode111
input/regression/scheme-text-spanner.ly:111: (ly:grob-set-property!
On 2010/08/28 19:56:11, Neil Puttock wrote:
> (begin
>   (ly:grob-set-property! grob 'side-axis axis)
>   (ly:grob-chain-callback ...

Done.

http://codereview.appspot.com/1867050/diff/17001/18001#newcode201
input/regression/scheme-text-spanner.ly:201: a b c \schemeTextSpannerEnd d |
On 2010/08/28 20:18:58, Neil Puttock wrote:
> + another spanner to check ly:engraver-announce-end-grob works, e.g.,
> 
> \override SchemeTextSpanner #'to-barline = ##t
> a4\schemeTextSpannerStart b d c |
> a1\schemeTextSpannerEnd |

Done.

[Graham Percival (old account), 2010-10-02 16:02:37]

As far as I can see, Mike has resolved all the known problems with this patch. 
Are there any other problems, or should we accept it?

[Neil Puttock, 2010-10-03 22:19:49]

On 2010/10/02 16:02:37, Graham Percival wrote:
> As far as I can see, Mike has resolved all the known problems with this patch.

> Are there any other problems, or should we accept it?

I'm working on it now.