From: Tomas K. <to...@us...> - 2005-09-08 08:31:05
|
Japanese translation was written to make it work in Japanese language environment, because generic SquirrelMail functions did not support options that were needed for translation. Current SquirrelMail support is improving and developers are solving some issues in a way that does not depend on translation. Implementation depends on php mbstring extension and character set selected by translation. Are you running standard squirrelmail 1.4.4 install or modified one? Can you upgrade to 1.4.5? Are Chinese encodings supported by your php mbstring extension? > My php support mbstring and iconv, but not support > recode. (php said recode is not available on Windows > platforms) > > Was Japanese translation has similar problem before > and has re-written the functions? > > Thanks for your help! > > Angela > >> rfc2822 "2.1.1. Line Length Limits" >> http://www.faqs.org/rfcs/rfc2822.html >> -------------- >> There are two limits that this standard places on > the number of >> characters in a line. Each line of characters MUST > be no more than >> 998 characters, and SHOULD be no more than 78 > characters, excluding >> the CRLF. >> -------------- >> >> >> SquirrelMail can be set to use B encoding in > messages headers when message >> is written in multibyte character set. Current > 1.5.1cvs implementation >> requires mbstring support in php and multibyte > character set supported by >> mbstring extension. Big5 charset is supported since > php 4.3.0, if packager >> compiles php mbstring extension with Chinese > support. Implementation of B >> encoding uses three functions unavailable in 1.4.x > and modifies standard >> header encoding function. >> >> Other option - write custom handling functions for > XTRA_CODE extension in >> $languages. See Japanese translation for example. > You must write all >> functions in 1.4.x or only required functions in > 1.5.1cvs. >> >> Chinese translation must solve other issues too. > Some changes require php >> installation with full mbstring and recode or iconv > support. I think >> Windows php packages are limited and don't have > required extensions. If >> php install does not have recode or iconv support, > you must use basic php >> functions for character set decoding and conversion. > These basic functions >> are slow in big5 and gb2312 character sets. Imagine > how similar functions >> process all gb18030 characters or handle > iso-2022-cn. It is possible that >> gb18030 can be optimized. I haven't tried to find > any mathematical >> patterns in character set mapping table. >> >> I might try porting 1.5.1cvs code to 1.4.x and > provide patch that fixes >> some issues, but I need to know if your php > installation can handle all >> modifications. >> >> -- >> Tomas >> > > _______________________________________ > 想即時收到新 email 通知? > 下載 Yahoo! Messenger http://messenger.yahoo.com.hk |