I suspect we've found a bug in cmsms (I mean Russian forum users).
* Disconnected (Connection reset by peer).
* zaytsev already in use. Retrying with zaytsev_...
* Now talking on #cms
* Topic for #cms is: Meeting attendees: Please join the cmsms-dev mailing list || Official Channel for CMS Made Simple ||
http://cmsmadesimple.org || Current Version: 1.2.3 || Don't spam, use the pastebin!
http://cmsmadesimple.org/pastebin. Password: simple
* Topic for #cms set by Ted at Tue Jan 29 22:29:47 2008
* #cms :[freenode-info] channel trolls and no channel staff around to help? please check with freenode support:
http://freenode.net/faq.shtml#gettinghelp
* efix (n=
frank@fxchains.fttp.xmission.com) has joined #cms
* efix (n=
frank@fxchains.fttp.xmission.com) has left #cms
It turns out that you use a log of preg_replace here and there, and you usually omit the "u" modifier which is required by PCRE if the string is in UTF-8.
That does not cause any harm for latin-based languages, however preg_replace when used with unicode strings without the u modifier tends to currupt some national characters like "К" in Russian.
http://forum.cmsmadesimple.org/index.ph ... 903.0.html - link to the discussion
Title: CMS Made Simple Forum: (at forum.cmsmadesimple.org)
calguy1000 got the email thanks
* sportman is having girl issues again
zaytsev_: here and there? quick grep finds 555 lines :/
sportman: join the club
The funny thing is that it depends on the locale settings
And PCRE & PHP versions
why I'm not suprised
* sportman decided that his girl friend dries him so crazy
* sportman needs to figure out how to break up with her, but friday is her birthday
:/
Sometimes, if the correct locale is set it works. Sometimes it does not work even with the correct locale settings (ru_RU.UTF-8). That's why the problem was unnoticed for a long time.
we have a saying "älä sano pahasti, sano hyvästi" rough translation "dont say anything bad, just say goodbye"
works better in finnish
(on Windows, it does not work with UTF-8 without the /u modifier at all because there the locale stuff does not work the same way)
sportman, I'm fed up with girls... welcome to the club
windows doesnt work
sportman, I like Perl and Python scripts. They are more predictable and less ressource hungry.
lol zaytsev
i like that one
tsw, anyway I think it's a problem and it should be taken care of.
* fransman has quit ("Leaving.")
In short: if the current encoding is UTF-8, one must use /u and it will work no matter which version of PCRE is being used and what's the locale settings. If it's anything else, one shouldn't use /u at all.
By the way, it seems like ereg_replace doesn't have such problems at all.
tsw what are u fighting with your girl over again?
So a possible solution is to replace it with ereg's but I am not sure about the performance impact. And maybe sometimes some PCRE-specific features were used.
Maybe I should file a bug or something. It's really important for everybody not using latin-based alphabet.
The guys who discovered this are waiting for devs to reply. But I'm quite sure you wont be replying them in Russian so welll I'm just trying to be helpful.
