Re: Re: Alternative mbstring implementation using ICU

On Fri, Jul 31, 2009 at 2:37 AM, Stanislav Malyshev<stas@zend.com> wrote: > Hi! > >> Aren't there any interests on this? If you think PHP 6 is gonna cover >> all of the functionality that allegedly-cruft mbstring currently >> provides, that is almost wrong :-p > > Could you please explain why PHP6 doesn't provide what mbstring is doing? > I.e, let's go over the functions: > > mb_parse_str - since detecting encoding doesn't work per RFC, what is the > usefulness of this function? Wouldn't PHP 6 do the same with correct > charset? As for this you got the point. > mb_str* - shouldn't you in 6 just convert them to unicode and do all string > operations with Unicode strings? Also, in 5 isn't there some intersection > with grapheme_* functions? mb_strwidth() and mb_strimwidth() are not covered. > mb_output_handler - shouldn't setting the proper encoding in 6 do the same job? > mb_convert_encoding - don't we already have a number of functions that do encoding > conversions? I don't think It can gracefully handle characters that have no corresponding entries in the target character set. I'm even thinking of adding a class interface that is dedicated to encoding conversion with which one can deal with such characters in a user-supplied handler. Regards, Moriyoshi