From: Tomas K. <to...@us...> - 2006-02-25 14:07:55
|
> I noticed that the subject lines in Squirrelmail's message index are much > shorter with Japanese enabled, and sometimes have characters like & and > # > before the ... > > The reason for this seems to be that they are having characters converted > to HTML entities before the substring. This problem is magnified by > spaces and other common characters being translated to HTML entities, > making the text appear to substring to be much longer than it normally > would: > > > <td bgcolor="#ffffff" align="left"><b><a > href="read_body.php?mailbox=INBOX&passed_id=742&startMessage=1" > title="Re:  > [compadultsk8s] Re: Do you in line skate ?  > ;whats the difference compareed...">Re:  > [compadultsk8s] Re: Do you&...</a></b></td> Please stop blaming mbstring. It has only one dangerous feature. function overloading. This feature is not enabled by default and not enabled in your setup. SquirrelMail translations don't stay in English because you have mbstring. System locale issue is specific to gettext extension and not to mbstring. Main Japanese translation code was written before changes in SquirrelMail header decoding. If decoding changes broke Japanese translation, we must fix decoding code and Japanese extra code functions. -- Tomas |