> Instead, complete-grob-entry (or whatever it was
> called)
> needs to be changed in order to work non-destructively. It does not
> make sense to do this separately from the original code.
OK, I won't bother yet with add-grob-definition here, but I wonder: if I
use copy-tree on a literal expression, is it alright to modify the
resulting list with assoc-set! and the like? In other words, does
copying the list create another literal that I should again refrain from
tampering with?
I ask because the following attempt at avoidance looks like overkill to
me:
#(define (add-grob-definition grob-name grob-entry)
(let* ((meta (assoc 'meta grob-entry))
(meta-entry (cdr meta))
(class (assoc-get 'class meta-entry))
(ifaces (assoc 'interfaces meta-entry))
(ifaces-entry (cdr ifaces))
(ifaces-entry (append (case class
((Item) '(item-interface))
((Spanner) '(spanner-interface))
((Paper_column) '((item-interface
paper-column-interface)))
((System) '((system-interface
spanner-interface)))
(else '(unknown-interface)))
ifaces-entry))
(ifaces-entry (uniq-list (sort ifaces-entry symbol<?)))
(ifaces-entry (cons 'grob-interface ifaces-entry))
(ifaces (cons (car ifaces) ifaces-entry))
(meta-entry (acons 'name grob-name meta-entry))
(meta-entry
(map (lambda (x)
(if (eq? (car x) 'interfaces) ifaces x))
meta-entry))
(meta (cons (car meta) meta-entry))
(grob-entry
(map (lambda (x)
(if (eq? (car x) 'meta) meta x))
grob-entry)))
(set-object-property! grob-name 'translation-type? list?)
(set-object-property! grob-name 'is-grob? #t)
(set! all-grob-descriptions
(cons (cons grob-name grob-entry)
all-grob-descriptions))))
Thanks for your patience!
David
https://codereview.appspot.com/11614044/
_______________________________________________
lilypond-devel mailing list
lilypond-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/lilypond-devel
