1068514 - Remove abandoned gfx-only Japanese and Traditional Chinese encodings

Status: RESOLVED FIXED

(Core :: Internationalization, defect)

Description

[Henri Sivonen (:hsivonen)]

nsUnicodeToHKSCS and nsUnicodeToJISx0201 are unused encode-only encodings. They are most likely leftovers from the pre-FreeType X11 gfx days. Let's remove them as dead code.

Comment 1

[Henri Sivonen (:hsivonen)]

Attachment: Remove dead code

https://tbpl.mozilla.org/?tree=Try&rev=918adb454556

Comment 2

[Henri Sivonen (:hsivonen)]

Oh, and the patch removes some dead-code OS/2-specific data tables for the OS/2 variant of Shift_JIS.

Comment 3

[Masatoshi Kimura [:emk]]

Comment on attachment 8490613 [details] [diff] [review]
Remove dead code

Review of attachment 8490613 [details] [diff] [review]:
-----------------------------------------------------------------

::: intl/uconv/ucvja/japanese.map
@@ -50,5 @@
>  static const uint16_t * const gCP932Index[] = {
>    gCP932IndexShiftJis, gCP932IndexJis0208
>  };
>  
> -static const uint16_t gIBM943IndexShiftJis[] = {

Please modify intl/uconv/tools/jamap.pl and regenerate japanese.map instead of removing the table it by hand.

Comment 4

[Henri Sivonen (:hsivonen)]

Attachment: change jamap.pl and regenerate japanese.map

(In reply to Masatoshi Kimura [:emk] from comment #3)
> Please modify intl/uconv/tools/jamap.pl and regenerate japanese.map instead
> of removing the table it by hand.

I edited jamap.pl to remove the IBM parts. Then I obtained the data files from unicode.org and ran jamap.pl. japanese.map changed more than I expected (the diff is to the manual edit from the previous patch). What's going on?

Comment 5

[Masatoshi Kimura [:emk]]

(In reply to Henri Sivonen (:hsivonen) from comment #4)
> Created attachment 8491282 [details] [diff] [review]
> change jamap.pl and regenerate japanese.map
> 
> (In reply to Masatoshi Kimura [:emk] from comment #3)
> > Please modify intl/uconv/tools/jamap.pl and regenerate japanese.map instead
> > of removing the table it by hand.
> 
> I edited jamap.pl to remove the IBM parts. Then I obtained the data files
> from unicode.org and ran jamap.pl. japanese.map changed more than I expected
> (the diff is to the manual edit from the previous patch). What's going on?

It looks like the change is expected (though I'll double-check it later).

The index 8648 to 9211 was used for the IBM943-specific mapping data. Now IBM943 was killed, so the extra data was removed from the table (it's the reason I asked you to regenerate the table).
gJIS0212Index has index values to the second table, so the most index values are shifted.

Comment 6

[Masatoshi Kimura [:emk]]

Comment on attachment 8491282 [details] [diff] [review]
change jamap.pl and regenerate japanese.map

I ran the patched jamap.pl locally and got exactly the same file.

Comment 7

[Henri Sivonen (:hsivonen)]

Thanks. Landed folded:
https://hg.mozilla.org/integration/mozilla-inbound/rev/52b0ec2f0522

Comment 8

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/52b0ec2f0522