Re: Issue 2149 in lilypond: Patch: Creates non-negative-integer? predicate.

[lilypond]

Comment #21 on issue 2149 by address@hidden: Patch: Creates  
non-negative-integer? predicate.
http://code.google.com/p/lilypond/issues/detail?id=2149

I have turned to the Guile manual for guidance about what would fit into  
the scheme of names, and we have basically "length" as a textual  
description for things like vector sizes or "index" for addressing them.

Guile actually has no problems creating a vector of length 0.  "length" in  
LilyPond is likely not a good idea since it suggests metric lengths, and  
also does not obviously include 0.  index? is not all that optimal since  
using it for the size rather than the index of an array is a bit strained  
(it's basically the index beyond the array).  However, it seems less  
strained than using "count?" as an index type since one does not really  
_count_ the values processed so far.

And I think it will usually correspond rather closely to what we would be  
using it for.

Barring protests in the next few days, I'll check in a version using index?  
instead.  I think that it should, for most of our purposes, be more natural  
than non-negative-integer?.  For a similar reason, I had created a  
predicate fraction? rather than non-negative-integer-pair?.  It does not  
capture all possible uses of this predicate, but for the real existing  
uses, being totally exact would be much more cumbersome.