274102 - Pashto, Urdu and Sindhi should use Arabic font pref

Status: RESOLVED FIXED

(Core :: Layout: Text and Fonts, defect)

Description

[Henri Sivonen (:hsivonen)]

(Pointed out by Andreas Prilop in ciwah.)

Steps to reproduce:
1) Set a distinctive font for Arabic.
2) Load http://www.unics.uni-hannover.de/nhtcapri/temp/lang-attribute.htm

Actual results:
Text annotated as ps or ur is not rendered like the text annotated as ar or fa.

Expected results:
Expected ar, fa, ps and ur to use the same font.

Comment 1

[Henri Sivonen (:hsivonen)]

Attachment: Map ps, ur and sd to ar

Trivial fix.

Comment 2

[Henri Sivonen (:hsivonen)]

Comment on attachment 168432 [details] [diff] [review]
Map ps, ur and sd to ar

Requesting r=.

Comment 3

[Simon Montagu :smontagu]

Comment on attachment 168432 [details] [diff] [review]
Map ps, ur and sd to ar

Some day we might want finer grained subdivisions, since Urdu typically uses a
different style of fonts from Arabic (I don't know what Pashto and Sindhi use),
but meanwhile r=smontagu.

Comment 4

[Henri Sivonen (:hsivonen)]

Comment on attachment 168432 [details] [diff] [review]
Map ps, ur and sd to ar

requesting sr=

Comment 5

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Comment on attachment 168432 [details] [diff] [review]
Map ps, ur and sd to ar

sr=dbaron, although I'm wondering if you want to do something about the
commented-out lines for Kashmiri and Kurdish.  It seems like they should
probably either be uncommented or the comment should be changed (I don't know
which).

Comment 6

[Henri Sivonen (:hsivonen)]

A. Prilop: Ping? What's your take on dbaron's comment on Kashmiri and Kurdish? A
quick glance at omniglot.com suggests that Kashmiri should use the Arabic font
pref but contemporary Kurdish should not.

Comment 7

[Simon Montagu :smontagu]

Personally I don't think we should set a default lang group at all for languages
like Kashmiri and Kurdish (and also Punjabi) which are written in different
scripts by different groups. Perhaps we should change the comments to be like
the comment for Punjabi:
#ks=Devanagari (India) and Arabic (Pakistan)
#ku=Arabic; also Latin (Turkey) and Cyrillic (former USSR)

Comment 8

[Henri Sivonen (:hsivonen)]

Yeah, I suspected even the Kashmiri language is politically problematic in that way.

I don't have CVS access. smontagu, could you check this is, please? Or should I
attach another patch with different comments and go through another round of r/sr?

Comment 9

[Simon Montagu :smontagu]

Patch checked in as is. I'll revise a lot of the comments in the file anyway
when I fix bug 192636.

Comment 10

[Andreas Prilop]

If a language can be written in two scripts or is not recognized
at all by Mozilla, then the typeface should be chosen according to
the charset parameter.

For example:

   charset=ISO-8859-5   lang=mn

is currently (Mozilla 1.7) displayed in the Western typeface because
the language Mongolian is not recognized. IMHO, it should be displayed
in the Cyrillic typeface because of "ISO-8895-5".

Another example is Serbian, which _need not_ to be written in the
Cyrillic script. In fact, Serbian is often written in Latin letters
on the Internet. Even books published in Belgrade in Serbian language
may use the Latin script.

 http://www.google.com/intl/sr/   http://www.google.co.yu/
have only recently switched to Cyrillic.