I know this has come up before, but I'd really like to see some basic support for Unicode/wide character encodings in boost.
While this is a really complicated issue (which I confess I understand very little about), I think a good start would be to only worry about conversions between various encodings -- I know there are a few facets lying around in the files section at yahoo doing just this. I am definitely a rank amateur in this area (which is why I'd like to see something in boost which does the hard work for me :) ), so I can't offer any substantive comment on what such a library should have; however, as a programmer I find that I often want to be able to read in strings in encoding X and spit them out in encoding Y. I believe Dinkumware sells a set of facets doing this for popular encodings (http://www.dinkumware.com/libDCorX.html), but I feel that this is an important enough problem that there should be an equivalent library in boost -- I think a lot of free projects would benefit from it. As far as I can see, this is a far simpler problem than full-blown support for Unicode (with string comparisons/equality and the like) -- in the general case, I suspect that providing a friendly wrapper over a library such as ICU (http://oss.software.ibm.com/developerworks/opensource/icu/project/) would be a good start... I am more than willing to help someone else (hopefully someone with more domain experience) with the work, but I feel I don't know anywhere near enough in this area to embark on the project by myself. I really don't see significant difficulty in doing the conversion facets, but I know that every time anything to do with wide characters comes up on the newsgroups I get lost in all the little problems people raise; so it may just be my inexperience with the area. Any takers? Damien _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
