[pdftex] 2 questions

[Thierry Bouche]

Reinhard Kotucha a écrit :

> You can run t1disasm on a type1 font to see how é is implemeted
> there and then write a program that retrieves information from the
> afm/tfm/vpl files and create a new character /eacute and then run
> t1asm on that file.  If this works, it should be possible to put this
> algorithm into pdftex.

I think that this trick has been used for some times for CS fonts, but
my question was really different (labelling some TeX construct with a
ToUnicode, this could also make maths searchable...?) 

Yes, it is certainly quite easy to make what fontinst does with a SEAC
in the type 1, e.g., or with y&y's fmp. But my question is about the
feasibility of asking tex to add ToUnicode info for complex things.

Thierry