Re: character by name on the gnome canvas, custom fontencoding

[Werner LEMBERG <wl gnu org>]

> My memory is that character maps in Type1 fonts are limited to 256
> characters.

Type 1 fonts don't have character maps at all.  If a font contains
e.g. `/Encoding StandardEncoding' this only means that Adobe Standard
Encoding is a good first guess to access glyphs in the font.  It
provides a default, nothing more.  An application has to scan either
the font directly or the accompanying AFM file to find out the real
number of glyphs and its glyph names.  Within PostScript you have to
build new /Encoding vectors to access glyphs not currently accessible,
and such encoding vectors are limited to 256 glyphs.

> But in any case, if the characters have the standard unicode names,
> I think FreeType can find them even without an explicit character
> map, so in some sense the way to provide a character map for a Type1
> font is to give things the right names.

The right solution IMHO is that pango gets all glyph names in the font
and constructs a proper cmap itself as explained in another mail.


    Werner