Have a look at the test programs 40UnicodeRoundTrip.t and 41Unicode.t of DBD::ODBC, and especially at the dumpstr() function in the helper module UChelp.pm. Most likely the strings you pass to DBI are not Unicode strings, but byte sequences representing UTF-8, which are interpreted as legacy character strings.
Alexander
In reply to Re: Oddities in UTF-8 and DBD::Pg?
by afoken
in thread Oddities in UTF-8 and DBD::Pg?
by Anonymous Monk
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |