John: > 2. UTF-8 is NOT the same as ASCII for values greater than 127. > Similarly, UTF-16 is NOT the same as UCS-2 (the wide "Unicode" chars > used by MS APIs), though it looks the same at low values. UTF-16 is a > multibyte character set, while UCS-2 is always 2 bytes per character. > You have to convert these values. Better to just use the regular UTF8 > versions.
Really, here you touched tangentially the core of my question. Besides all those great theories, at last I have UTF-8 encoded data in a dBase, and the UCS-2 encoded data of the MS Win32 API (w_chars in muy Cpp app). The question is: What is the concrete way to and from that w_chars to the UTF-8 that need SQLite? (my guess is use UTF-8 to the stored data) Can I use wcstombs() or even WideCharToMultiByte()? -this last seem a nightmare to me-. > If you are only now internationalizing your code, and you've been > passing 8bit strings to SQLite, you may already have a problem, since > some of the data stored may be invalid if you attempt to treat it as a > UTF-8 string. You may find that you'll need to dump the data, then > convert from ASCII to UTF-8 and rebuild the database. > Really, preventing future possible problems, my textual data was stored as blob. The idea is rebuild converting from 8 bits ASCII to UTF-8. Then the question: can be done with mbstowcs() prior to using sqlite3_bind_text()? A.J.Millan _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
