Simplified Chinese support became chao with 2.2.6

[Longbow]

Hi vBulletin team

just upgraded my 2.2.5 to 2.2.6, some Simplified Chinese characters are displayed incorrect now. They turned out to be another character, which made some sentenses unreadable.

any hint would be highly appreciated.

[Scott MacVicar]

This error has occured with japanese as well, is the meta encoding correct in the headers?

If so then open /admin/functions.php

and remove

        $bbcode=str_replace("{", "{", $bbcode); // stop people posting replacements in their posts 

[Longbow]

Originally posted by PPN
This error has occured with japanese as well, is the meta encoding correct in the headers?
If so then open /admin/functions.php
and remove

thanks for Ur reply, i've commented this line, but the problem still exist.

the meta encoding is correct, only some words appeared to be wrong.

other method i could try? thanks a lot.

[Scott MacVicar]

What characters are causing the problem?

Could you post a url to the affected post.

[Longbow]

Originally posted by PPN
What characters are causing the problem?

Could you post a url to the affected post.

http://forum.cloudworld.net/showthread.php?s=&threadid=17929

the first word of the thread title appeared to be wrong - i'm not sure if U have Chinese support or not.

http://forum.cloudworld.net/showthread.php?s=&postid=213426#post213426

in this post, the first character in first line, and the second character in second line, are both wrong.

this problem did appear in preivous version as well, very rarely, and once i edit them, they'll be fine.

but this time, it happens constantly to certain words.

thanks for help.

[Scott MacVicar]

I have chinese support installed :P
Ill have to have a good look at it when i get back in about 4 hours. cause i now have an exam to go to :O

[Longbow]

Originally posted by PPN
I have chinese support installed :P
Ill have to have a good look at it when i get back in about 4 hours. cause i now have an exam to go to :O

ha, exam. good luck.

btw, i'm sure there's something to do with the PHP part, cause when i directly modify the db, those words are correct.

[Longbow]

any update? thanks!

[mystara]

same thing is happening to my board which uses traditional chinese. it's not the "& # 123;" problem, i have corrected that after 2.25.

the problem is still with the "functions.php" file as when i uploaded the 2.25 version of it back, the problems are solved.

[Longbow]

thanks. i uploaded my old function.php and the problem did disappear!

i presume there is no hidden danger or some sort mixed them up?

thanks!

[Mike Sullivan]

Ok, I believe I've discovered the cause of this problem. These two lines in functions.php:

$text = str_replace(chr(173), '_', $text); 
$text = str_replace(chr(160), '_', $text);

In a single-byte (ASCII) language, those are blank spaces (but different than just a space). In a double-byte language, those are just going to be part of one letter.

So if you're using a double-byte language, remove those two lines and it should work.

[mystara]

thanks for the fix.

[shanx24]

does anyone have a japanese or chinese interface to the VB forums? i don't mean data, i mean the interface. i tried searching these forums but didnt come up with anything.

thanks in advance!

[Longbow]

Originally posted by Ed Sullivan
Ok, I believe I've discovered the cause of this problem. These two lines in functions.php:

$text = str_replace(chr(173), '_', $text); 
$text = str_replace(chr(160), '_', $text);

In a single-byte (ASCII) language, those are blank spaces (but different than just a space). In a double-byte language, those are just going to be part of one letter.

So if you're using a double-byte language, remove those two lines and it should work.

will U consider to provide a double-byte version? i upgraded to 2.2.7 RC today, but later found out i have to modify this again.

also, a "search function" which support double-byte characters would be fantastic.

i'd only to suggest this approach, as currently i can't ask for more from vBulletin. thanks.

[ccd1]

I think using Unicode instead of BIG5, GB or JIS would eliminate this problem, am I right?