Created attachment 537790 [details]
patch to fix urdu orth file

Description of problem:
We have recently packaged Urdu fonts from
http://www.crulp.org/software/localization.htm in Fedora. These fonts are
actively used for creating content in Urdu language from 5+ years.

fontconfig do not recognize them as complete font for urdu language due to
missing characters

U+0629 ARABIC LETTER TEH MARBUTA
U+0647 ARABIC LETTER HEH

I have got following response from upstream

" Alternate codes are recommended for Urdu
by Unicode standard itself (see the code-wise text explanation after
the code chart at http://www.unicode.org/charts/PDF/U0600.pdf for
U+06C3 as an alternate to U+0629 and U+06BE (or U+06C1; depends on
which Heh one is referring to) as an alternate to U+0647)."

Page 5 In Arabic code U+0600 Unicode chart 6.0 gives
U+06C3  "ARABIC LETTER TEH MARBUTA GOAL, Urdu"
U+06BE "ARABIC LETTER HEH DOACHASHMEE, Urdu"

So,
U+06C3 is recommended by Unicode instead of U+0629
U+06BE is recommended by Unicode instead of U+0647


Version-Release number of selected component (if applicable):
fontconfig-2.8.0-4.fc16.x86_64

How reproducible:
everytime

Steps to Reproduce:
1. yum install nafees-*
2. fc-query /usr/share/fonts/nafees-naskh/Nafees\ Naskh\ v2.01.ttf
3. it gives lang: (s)
  
Actual results:
lang: (s)

Expected results:
lang: ur(s)

Additional info:
attached patch