Le vendredi 02 juillet 2010 à 15:48 -0400, Máirín Duffy a écrit :
accents seem to work.
If I write 'Máirín' in MgOpen Modata, the á and í print out, but not in
the MgOpen Modata letterforms. Instead the font system falls back to
DejaVu (I believe, could be another fallback font but looks like DejaVu
to me).
You can easily test the language coverage of a font file using the
command:
FC_DEBUG=256 fc-query <filename>
It will print all sorts of information, including a list of ISO 639
language codes followed by the number of associated missing codepoints
in the font file. For example:
aa(10) ab(90) af(17) ak(21) am(264) an(14) ar(36) as(64) ast(14) av(67)
ay(8) { 00c4 00cf 00d1 00dc 00e4 00ef 00f1 00fc }
az-az(14) az-ir(40) ba(82) be(68) ber-dz(18) ber-ma(32) bg(60)
…
de(7) { 00c4 00d6 00dc 00df 00e4 00f6 00fc }
dv(49) dz(95) ee(47) el(0) en(20) eo(12) es(14) et(12)
…
If the number is zero (el(0)) the font file contains all the bits needed
to write the associated language (Greek, not surprizing since Modata is
a Greek font), if it is non-zero it is missing this number of
codepoints, if it is non-zero and less than ten fc-query will print the
list of missing codepoints after the language number :
ay(8) { 00c4 00cf 00d1 00dc 00e4 00ef 00f1 00fc }
Of course just because a font includes support for lots of codepoints
does not mean they are all well designed. So it's just a minimal test.
Also don't test just the Normal font, the coverage of Bold/Italic/etc
can ans is often different.
As for Modata, ga(28) means it sucks for Gaelic (Irish)
Regards,
--
Nicolas Mailhot