Backing up UTF-8 content using mysqldump

[Tal Rotbart]

Hi!

I'm using MySQL 4.1.15 on Debian GNU/Linux, and am using mysqldump to create a backup for the database. I am using mysqldump and not hot-copy on purpose as I want the backup to be in modifyable plain-text.

My problem is that my schema contains UTF-8 characters. UTF-8 works just fine while working with the DB from Java, phpMyAdmin but not with mysqldump.

The table's default encoding and colation is latin-1 but I choose specific columns to be in the utf8-bin collation and encoding.

The output from mysqldump seems to be ok:
A. The table creation includes the UTF-8 encoding and colation on the apropriate column.
B. The table data contains the unicode characters (double bytes) in the apropriate column data.

However when imported back in, each UTF-8 is shown as its composite bytes.

I tried adding the connection encoding to equal UTF-8 at the beginning of the dump but this didn't help.

Any clue?

Thanks,
Tal Rotbart
iContact Media Ltd.