Re: [ft-devel] About the Cmap format 14

[George Williams]

On Mon, 2007-10-08 at 07:46, Werner LEMBERG wrote:
> I've skimmed the specs but didn't look more closely.  Do you
> volunteer?  It seems that we need a special interface for that, right?
It needs a special interface.
   The most obvious reason is that it requires two codepoints (base
codepoint + variant selector) to specify a glyph.
   Also because the format=14 subtable is not stand-alone, it requires
lookups into a unicode charmap. In other words it could not be passed to
FT_Set_Charmap.

   As the spec currently stands there can only be one format 14 charmap
(with platform=0, specific=5) so there is no need to select among
several. So I see no need for an equivalent to 
         FT_Set_Charmap
         FT_Select_Charmap
These should return an error on a format 14 charmap.

   I propose that
         FT_Get_Char_Index
should return 0 (or 0xffffffff?) if called on a format 14 charmap (which
shouldn't be possible in the first place). Instead a new routine
         FT_Get_Char_Variant_Index(FT_Face  face,
                                   FT_ULong charcode,
                                   FT_ULong variantSelector)

   Similarly
         FT_Get_First_Char
         FT_Get_Next_Char
will never be called on a format 14 charmap. I'm not sure what best
reflects the spirit of these routines. Should there be routines which
return all codePoints for a given variationSelector, or routines which
return all variationSelectors for a given codePoint, or perhaps just a
list of all variantSelectors? or all three?
   I assume all of them

         FT_Get_First_Char_Of_Variant(FT_Face  face,
                                   FT_ULong variationSelector,
                                   FT_UInt *agindex)
(returns the first codePoint, for that variationSelector and sets the
GID)
         FT_Get_Next_Char_Of_Variant(FT_Face  face,
                                   FT_ULong variantSelector,
                                   FT_ULong codePoint,
                                   FT_UInt *agindex)
(returns the next codepoint for the variationSelector)
         FT_Get_First_Variant_For_Char(FT_Face  face,
                                       FT_ULong codePoint,
                                       FT_UInt *agindex)
(returns the first variation selector for that codePoint, and sets the
GID)
         FT_Get_Next_Variant_For_Char(FT_Face  face,
                                      FT_ULong codePoint,
                                      FT_ULong variantSelector,
                                      FT_UInt *agindex)
(returns the next variation selector for the codePoint)

         FT_Get_First_Variant_Selector(FT_Face  face)
(returns the first variationSelector)
         FT_Get_Next_Variant_Selector(FT_Face  face,
                                   FT_ULong variantSelector)
(returns the next variationSelector)

General questions: 
   How should it behave if the current cmap is not a unicode cmap?
Should it look for a unicode cmap in the list, or just return error?
   Should I extend the CMap class with these additional functions, or
just make direct calls?