-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Alexey Pechnikov wrote: > Tcl, Python and other langs have different unicode implementations. The > realizations are more simple than ICU library but millions of applications > are > using these. I'm will glad to see Tcl/Python/etc. unicode implementation in > SQLite.
Speaking for Python, you can trivially hook the Python unicode support into SQLite if you use pysqlite or APSW. However if you think the builtin Python Unicode support is good then you are mistaken :-) See http://www.cmlenz.net/archives/2008/07/the-truth-about-unicode-in-python > And I think we must have different Unicode implementations for UTF8 and UTF16 > by performance reasons. ICU works with UTF16 and isn't good for UTF8 > databases. Tcl has native encoding UTF16 too. Python uses UTF16 or UTF32 internally for Unicode (a compile time option). It isn't unreasonable to expect developers who are making heavy use of ICU like functionality (ie enough for performance to be noticeable) to set the default database encoding to UTF16 which means no encoding conversions while using ICU. There may be other such "fast path" optimizations to improve performance when using ICU. Roger -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) iEYEARECAAYFAknBQpoACgkQmOOfHg372QQmaQCfYxB/5SR9oxwR/0dCyimRpDvV TYsAoMNWoT9xG4+cH3lehMmems+/chK7 =hroj -----END PGP SIGNATURE----- _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
