Re: [Boost-users] [regex] ECMAScript support?

From: Armel Asselin (asselin.armel_at_[hidden])
Date: 2007-05-05 08:15:00

"John Maddock" <john_at_[hidden]> a écrit dans le message de news:
016901c78cde$b1fa7cc0$fe500252_at_fuji...
> Armel Asselin wrote:
>> So my questions:
>> - is boost.regex totally compliant with ECMAScript RE? or at least
>> more than
>> Perl engine? [1.33.1 vs. 1.34 ?]
>
> I believe so, if there's anything missing from 1.34 let me know.
>
>> - did I understand well that I can use iterators and have both UTF8
>> (char*)
>> and wchar_t searches in the same executable build? (thanks to their
>> template nature?)
>
> Iterators yes, no problem.  As for character types you can use:
>
> boost::regex with narrow characters, but not really with multibyte
> sequences.
>
> boost::wregex with wchar_t sequences.
>
> boost::u32regex with any of UTF8, UTF16 or UTF32 sequences (they all get
> converted to 32-bit code points on the fly), but you will need ICU to
> provide the Unicode character database.  See
> http://www.boost.org/libs/regex/doc/icu_strings.html for the docs.
>
> HTH, John.
ok so i'll try the 1.34, any idea of the size of the binary (.lib/.a) for
that feature (just the necessary part of ICU+boost::regex?). i'd be really 
happy if it could take a few
dozen KB :-)

Armel