Subject: Re: [boost] [Boost.Locale] New Release
From: Artyom (artyomtnk_at_[hidden])
Date: 2010-03-13 15:19:54
> From: Mathias Gaunard
>
> That's a very interesting wrapper of ICU.
> It might even kill off my Unicode library.
>
I don't think so, I've been watching after this library and it was very
interested in it.
Note, there is quite a big difference:
Boost.Locale provides **localization** centric API. I do not bother too
much with character properties, I even do not provide UTF-8/16/32
code-point iterator because you are rarely need to go down to such low
level. You rather need higher level abstraction: characters, words etc.
Most of algorithms provided in Boost.Locale are actually locale dependent
with very few exceptions:Unicode normalization, case folding and
character set conversion just because they extremely important.
More then that I fully support non-Unicode character sets (however I
do not recommend them).
So I do think that Unicode library that deals with lower level
is important and they would rather complete each other then
replace it.
Artyom