[luatex] luatex - font encoding for type 1 fonts

[Reinhard Kotucha]

On 2016-06-26 at 17:56:59 +0200, Pali Rohár wrote:

 > On Sunday 26 June 2016 17:41:44 Hans Hagen wrote:
 > > On 6/26/2016 5:16 PM, Pali Rohár wrote:
 > > > Apparently, some font styles of CSFonts have different mapping of
 > > > glyph names as basic roman style csr. Looks like /Encoding 256
 > > > array is correct in all PFB files (only problem with Knuth's
 > > > dollar-sterling joke).
 > > 
 > > that order is still unrelated to the real order in the glyph table
 > > (in thebinary section)
 > 
 > Yes, order in that binary section (where are defined curves for glyphs) 
 > is unrelated. But mapping from TFM index to glyph name is in that 
 > /Encoding 256 array.

Not necessarily!  If the TFM file requires another encoding, you
definitely have to re-encode the font.  In this case the built-in
/Encoding isn't used.

A common case to have several encodings for one font is that a font
can have more than 256 glyphs.  In order to access all of them you
need a .tfm and a .enc file for each subset.

However, the csfonts don't need to be re-encoded and actually are not.
Consult the file csfonts.map.

  $ less `kpsewhich csfonts.map`

Regards,
  Reinhard

-- 
------------------------------------------------------------------
Reinhard Kotucha                            Phone: +49-511-3373112
Marschnerstr. 25
D-30167 Hannover                    mailto:reinhard.kotucha at web.de
------------------------------------------------------------------