Re: [bug-gnu-libiconv] ISO-2022-JP to UTF-8

[Bruno Haible]

Jeff Diehl wrote:
> Nayuta - This patch does the trick.  Thank you for pointing it out.
> 
> Bruno - Any chance this patch can be incorporated into the next
> revision?

I have added an ISO-2022-JP-MS converter to libiconv. See the patch here:
http://git.savannah.gnu.org/gitweb/?p=libiconv.git;a=commitdiff;h=3a33986e1d2c819dc2b8a84684e067cf177d1815

The converter behaves mostly like Windows codepage 50221.

I did not use the patch from
http://www2d.biglobe.ne.jp/~msyk/software/libiconv-1.13-ja-patch.html
for various reasons:
  - Its iso2022_jpms.h does not handle the JIS X 0212 characters,
    that all of CP50220, CP50221, CP50222 handle.
  - It uses an escape sequence ESC '$' '(' '?' which violates the principles
    of ISO 2022 (even more than what Microsoft did).
  - It kills the mappings of user-defined characters. Ken Lunde's book
    explains why they are important.
  - It adds a number of unneeded encodings (such as EUC-JP-MS)
    and makes unrelated changes to the CP932 and JIS X 0201 converters.
  - It has no copyright assignment submitted to the FSF.

Bruno
-- 
In memoriam Hana Brady <http://en.wikipedia.org/wiki/Hana_Brady>