[04-08-2005 21:31, Dan Wellisch escreveu]
We just put a SQLlite application in production. It handles the display of ISO 8859-1characters just fine if they appear in the search
results.

Please bear in mind that sqlite3 expects UTF-8 input, not ISO8859-1, so
you need to make sure you convert every string to UTF-8 before calling
the API.

The same is valid for the output. If you display as another encoding
(which you probably do), remember to convert from UTF-8 to the encoding
you want.

If you don't do this, functions like SUBSTR() and every other string
manipulation function will not behave as expected when feed with 8-bit
chars (like ISO-8859-1 accents, the copyright symbol, etc).

Unfortunately, english only developers forget about this, making the
database data not consistent for international users (and even for
themselves if any (c) or other symbol is used).


Regards,
~Nuno Lucas

Reply via email to