| msg19221 - (view) |
Author: Jason R. Mastaler (jasonrm) |
Date: 2003-12-01 22:05 |
| As discussed last week on the email-sig list, the attached patch adds support for the CJKCodecs package as an alternative to the {Chinese,Japanese,Korean}Codecs packages. CJKCodecs 1.0.2 and above should work with this patch. This is advantageous because the Chinese and KoreanCodecs packages are no longer supported,maintained or available for download. This patch does not break compatibility with {Chinese,Japanese,Korean}Codecs, so they can still be used if desired. Lastly, this patch fixes a small typo that broke GB2312. |
|
|
| msg19222 - (view) |
Author: Barry A. Warsaw (barry) *  |
Date: 2003-12-29 14:49 |
| Logged In: YES user_id=12800 Correct me if I'm wrong, but won't the attached patch work better? It simply removes the entries from CODEC_MAP that are already provided by cjkcodecs.aliases (and japanese.aliases and korean.aliases). See Charset.py.diff |
|
|
| msg19223 - (view) |
Author: Barry A. Warsaw (barry) * |
Date: 2003-12-29 14:52 |
| Logged In: YES user_id=12800 Oops, with the typo fix for gb2312. |
|
|
| msg19224 - (view) |
Author: Barry A. Warsaw (barry) * |
Date: 2003-12-29 15:19 |
| Logged In: YES user_id=12800 One more rev of Charset.py.diff |
|
|
| msg19225 - (view) |
Author: Barry A. Warsaw (barry) * |
Date: 2003-12-30 04:33 |
| Logged In: YES user_id=12800 I'm not even sure this patch is correct, since it breaks the test suite. The problem is that self.output_codec ends up being different with the patch than without it (in Charset.__init__()). For example: Python 2.3.3 (#1, Dec 19 2003, 11:33:00) [GCC 3.2.2 20030222 (Red Hat Linux 3.2.2-5)] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> from email.Charset import Charset >>> c = Charset('euc-jp') >>> c.output_codec 'japanese.iso-2022-jp' >>> But now with Charset.py.diff applied: ... >>> c.output_codec 'euc-jp' We need to figure out what the right thing to do here is. |
|
|
| msg19226 - (view) |
Author: Jason R. Mastaler (jasonrm) |
Date: 2003-12-30 05:59 |
| Logged In: YES user_id=85984 comments regarding Barry's Charset.py.diff: You shouldn't mention KoreanCodecs and ChineseCodecs in the comments as alternatives to CJKCodecs. Both are no longer maintained, or even available for download. Both have been completely replaced by CJKCodecs. Only JapaneseCodecs remains as a substitute package. |
|
|
| msg19227 - (view) |
Author: Barry A. Warsaw (barry) * |
Date: 2003-12-30 15:16 |
| Logged In: YES user_id=12800 Latest version of the patch, with updated comments as per Jason's followup, and including Tokio Kikuchi's fix for the test suite regression. |
|
|
| msg19228 - (view) |
Author: Barry A. Warsaw (barry) * |
Date: 2003-12-30 16:52 |
| Logged In: YES user_id=12800 Applied to Python trunk (2.4). This will be applied to Python 2.3 and closed when that branch's freeze is lifted. |
|
|
| msg19229 - (view) |
Author: Barry A. Warsaw (barry) * |
Date: 2004-05-10 15:04 |
| Logged In: YES user_id=12800 Closing this without applying it to 2.3. The general consensus was not to mess with this in a bug fix release, so it will have to wait until Python 2.4/email3. |
|
|