> > > In practice, right now we just look up a charmap with either > > TT_ADOBE_ID_STANDARD or TT_ADOBE_ID_CUSTOM encoding, which > > essentially works, but having a way to distinguish the actual > > font-provided charmap from the FreeType-synthesized one seems > > cleaner. > > Up to now nobody has asked for such a feature. Can you give a short > outline what you would like FreeType to provide? >
Given that Matplotlib will need to support not-too-recent versions of FreeType at least for a while, I think just clarifying the documentation may be sufficient, e.g. replacing (in the docs for FT_Encoding) "By default, FreeType enables a Unicode charmap and tags it with FT_ENCODING_UNICODE when it is either provided or can be generated from PostScript glyph name dictionaries in the font file." by something like "When loading a font, FreeType will make a Unicode charmap active if possible (either when the font provides such a charmap, or if FreeType can synthesize one from PostScript glyph name dictionaries; in either case, the charmap is tagged with FT_ENCODING_UNICODE). If such a charmap is synthesized, it will be placed at the first position of the charmap array." (Assuming this behavior can be relied on.) (I guess in the current wording, "by default" refers to the enabling of the charmap, (which can be changed) but until your reply I thought it referred to the synthesis (which cannot actually be changed). The new wording hopefully makes it clearer that synthesis is not togglable and improves one's chances to identify the synthesized charmap.) Alternatively, perhaps the synthesis of the charmap could be completely toggled off using a new params entry in FT_Open_Args, but perhaps that's a bit overkill or too specialized to be worth an API addition. Regards, Antony _______________________________________________ Freetype mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/freetype
