Problems with unicode

[Brian McAnulty]

Hi,

I'm having problems getting the unicode support to work, JDK 1.4.2_06, Connector/J 3.18, MySQL 4.1.10 on Win2K.

Basically, I need to store unicode characters in a table that exists in a non unicode database, ie: the default database set is 'latin1'.

I can't change the default database set as I need the ability to pull other MyISAM tables into this database now and then, and you can't do this if the current DB is 'utf8' and the DB you're pulling the .myd/.myi/.frm files are older (in this case 3.23.38 MyISAM tables). MySQL does a neat trick here and reduces the width of text columns to a 3rd of their original width. Useful.

(If I set the default character set to 'utf8' everything works fine - I just can't!)

So...I need to leave the default character set as latin1 on the server.

I've tried a combination of the following JDBC connection params:

useUnicode=true
characterEncoding=utf8
characterSetResults=utf8

And none of these seem to do the trick.

Well, I say "they don't seem to do the trick" but perhaps they are - its quite difficult to tell. If I use a MySQL client (DOS) and just do the usual SELECT x FROM mytable I get ?????? even though my my.cnf file has the 'default-character-set=utf8' in the client section (not in the mysqld section). So, I would assume that the characters aren't being stored as unicode - correct?

I noticed, stepping through the code, that there is another param (I couldn't find documentation on this) - useOldUTF8Behavior.

So, anyone got any ideas? It can't be that difficult, surely?!

Cheers
Brian