Re: [Freetype] Certain characters of pbf fonts don't render?

[Jan-Åke Larsson]

Werner LEMBERG wrote:
> FreeType can only provide information the font offers.  utmr8a.pfb
> says that it uses /StandardEncoding by default; this is what FreeType
> correctly reports.  What you want is a mechanism to add a charmap to a
> font. 

First of all, my comment was more directed to Joris; using
FT_ENCODING_ADOBE_CUSTOM may not always work.

Once I found out how to do it, it is not a big problem. I'm not sure
there is a real need for this in FT2.

Of course, you are right that any PostScript code might be used to
reencode the font. Fortunately, in TeX the encodings follow a specific
format [1] which I can easily parse. Once that is done it is also easy
to

    FT_Get_Name_Index( face, enc_charname[c] )  

> Hmm, we need a kind of FT_Define_Charmap to create a
> user-defined charmap which can be attached to a font.  Any
> suggestions how such an interface could look like? 

_Ideally_ one would like to load the encoding files [1]:

  struct FT_CharMapRec_* FT_Load_Charmap(FT_String* encodingfile);

This would be a simple API call but probably very restricted to the
TeX case. Possibly, one would do this per-face, i.e., add a face as
argument to the above. If you do find this useful for more people than
me and actually want to implement it, you would probably
(also/instead) want to have

  struct FT_CharMapRec_* FT_Init_Charmap(void);
  FT_Error FT_Done_Charmap(FT_CharMapRec_* charmap);
  FT_Error FT_Set_Char_Index(FT_CharMapRec_* charmap, FT_UInt charindex,
                             FT_UInt glyphindex);
  FT_Error FT_Set_Char_Name(FT_CharMapRec_* charmap, FT_String* charname,
                            FT_UInt glyphindex);

These also possibly per-face.
 
> And is there a volunteer who wants to implement that (provided that
> we all agree that such a feature is useful)? 

I would if I had the time. Unfortunately I don't. On the other hand, I
already have the functionality I need (and I would not miss this
feature overly much).

In any case I also have to parse some files to map TeX font name into
{ PostScript font name, extend factor, slant factor, encodingname },
and to find the font file and the encoding file. Adding a few lines of
code for the encoding parsing is no big deal. But (to those of you
rendering TeX output with FT2:) one should be aware that this is
needed.

Thanks,
/Jan-Åke





[1] Examples:
---------
/StandardEncoding [
/.notdef /.notdef /.notdef /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /.notdef /.notdef /.notdef
/.notdef /.notdef /.notdef /.notdef /space /exclam /quotedbl
/numbersign /dollar /percent /ampersand /quoteright /parenleft
/parenright /asterisk /plus /comma /hyphen /period /slash
<snip>
----------
/TeXBase1Encoding [
% 0x00
  /.notdef /dotaccent /fi /fl
  /fraction /hungarumlaut /Lslash /lslash
  /ogonek /ring /.notdef
  /breve /minus /.notdef 
  /Zcaron /zcaron 
% 0x10
 /caron /dotlessi 
 /dotlessj /ff /ffi /ffl 
 /.notdef /.notdef /.notdef /.notdef
 /.notdef /.notdef /.notdef /.notdef
 /grave /quotesingle 
% 0x20 (ASCII begins)
 /space /exclam /quotedbl /numbersign
 /dollar /percent /ampersand /quoteright
 /parenleft /parenright /asterisk /plus /comma /hyphen /period /slash
<snip>