Re: [Gnustep-cvs] Added missing method [isValidGlyphIndex:].

[Fred Kiefer]

Alexander Malmberg wrote:
> > Commit from FredKiefer (2003-09-29 17:25 EDT)
> > ----------------------
> > Added missing method [isValidGlyphIndex:].
>
> [snip]
>
> This seems a bit odd. glyphs->glyph_length contains the number of
> consecutive built glyphs from the start, so by only checking that field,
> the method will only return yes if the glyph for the given index has
> already been built. Is this really what the method is supposed to do?
>
> Although the documentation doesn't explicitly state that "valid" means
> that it's out-of-range (as opposed to not built), all other methods
> ensure that the demand-building of glyphs isn't visible to the caller.
> In particular, -glyphAtIndex:isValidIndex: is explicitly documented as
> building glyphs if necessary, and I don't see why -isValidGlyphIndex:
> should give different results wrt. validity.

I am also not sure, only a test on a Cocoa system could prove what Apple 
really does. But I assume, that for all methods that do perform glyph 
generation they explicitly state, that they do it. So if this statement 
is missing, I wont expect glyph generation. Than the only possible check 
is of course against the highest created glyph index, which is what I 
did implement.

Cheers
Fred