Re: Re: PHP's Zend multibyte support

Hello Alan, that is basically my impression. Although I did not look into who might be using what in particular. Either way the feature is fully dropped in HEAD. marcus Monday, March 24, 2008, 4:12:54 PM, you wrote: > Reading through this, It looks like this is aimed at Big5/shiftJS/? > maybe korean?? > I know from HK and probably TW, who use Big5 (traditional chinese > encoding), that it's pretty much been phased out with the introduction > of unicode (basically AFAIR Windows 95/98/ME/NT chinese editions used > these encodings, latter versions of windows have used UTF8). This tended > to be an issue that old browsers on these platforms tended to need Big5 > encoded web pages, so alot of stuff was written at Big5. > Basically I don't think these days (in the last 3 years at least) anyone > develops sites that target Big5, (we all use UTF8) and if they do, it's > probably only done as a last stage iconv conversion. > So is there a current need for this? - or is it just a legacy issue, > which could be dropped anyway for PHP6? > Regards > Alan > Johannes Schlüter wrote: >> Hi, >> >> On Sun, 2008-03-23 at 15:26 +0100, Hannes Magnusson wrote: >> >> >>>> You can provide a --SKIPIF-- section to detect MB support >>>> (http://qa.php.net/write-test.php ). >>>> --SKIPIF-- >>>> <?php >>>> if (!in_array("detect_unicode", array_keys(ini_get_all()))) { >>>> die "skip Requires --enable-zend-multibyte option"; >>>> >>> WTF? Where did that INI entry come from? :) >>> >>> It isn't in php.ini-dist and the only docs I can find is >>> (http://no.php.net/manual/en/ini.php): >>> detect_unicode "1" PHP_INI_ALL Available since PHP 5.1.0. >>> Removed in PHP 6.0.0. >>> >> >> According to my research when the MB issues popped up that setting was >> added as a hack to help the scanner to have less confusion with it's >> encoding detection when using __HALT_COMPILER(); There's an internals >> discussion, but i don't have it at hand right now. >> >> >>> Is it really PHP_INI_ALL? That doesn't make much sense to me. I >>> thought the multibyte scanning was compile time? >>> (Or is it like the current short_tags proposal; >>> ini_set("detect_unicode"); include "..."; ?) >>> >> >> Yes, it's for the latter, I guess. >> >> johannes >> >> >> Best regards, Marcus