Hello, I'm reporting a problem with decoding the input line and converting arguments to appropriate encoding. I have a database in UNICODE encoding but I'm accessing it in LATIN2 form. Now this works (where 'xxx' are special characters in latin2): psql mail mail=> \encoding latin2 mail=> update folder set description='xxx' where folderid=102; UPDATE 1 mail=> select description from folder where folderid=102; description ------------- xxx (1 row) But this doesn't work: mail=> \encoding latin2 mail=> update folder set description='\354\271\350' where folderid=102; UPDATE 1 mail=> select description from folder where folderid=102; NOTICE: utf_to_local: could not convert UTF-8 (0xecb9e8). Ignored I tried to do: mail=> \encoding unicode mail=> select description from folder where folderid=102; description ------------- xxx (1 row) It seems to me that when the characters are presented to the parser in the form \ddd, it doesn't convert them and saves tham incorrectly in UNICODE encoding. Version: PostgreSQL 7.1.3 on i686-pc-linux-gnu, compiled by GCC 2.95.4 on a testing debian verision Debian testing distro. Best regards Ondrej Palkovsky ---------------------------(end of broadcast)--------------------------- TIP 4: Don't 'kill -9' the postmaster
