Hi !
I am using hibernate 3.0.3 to read from a mysql5.0 UTF8 charset database.
I inserted (by phpmyadmin) some russian words in different fields, and then I read them with my application (hibernate+myfaces).
I read cyrillic fonts correctly, but if I write new words and I save them via Hibernate, they are not saved correctly.
I read a lot of ???????? from the database.

How does hibernate manage collation and charsets ?
Should I set any property in the hibernate.cfg.xml ?

Thank you for any help,
Nicola

Hibernate doe's not use charsets, it reads and writes "java.lang.String". Just use correct charset in HTML and see driver documentation to set correct encoding.

Hi !
Could you be more specific ?
HTML charset is correctly set, do I have to set the encoding just before passing to hibernate the string to be updated ?
Bye
Nic

Web server and JDBC driver use encodings to conver bytes to characters and characters to bytes for communication with web server and database.
1) Use HTTP header or HTTP meta tags to set encoding for browser.
2) Configure web server to convert strings to the same encoding as browser expects (1 and 2 must use the same encoding, it can be UTF-8)
3) Configure driver or database to use the same encoding use for communication protocol (it can be UTF-8 too)

Configuration is a web server and database specific stuff, it is common to configure encoding in web applications context or global configuration, drivers are configured using JDBC URL parameter and database storage format can be specified at database create time. See your web server and mySQL documentation for details (parameter names), it can be a version specific stuff too. It is not related to hibernte, it has nothing to do with encodings.

Hi !
I have solved adding this 2 line in hibernate.cfg.xml:



Thank you for your advices,
Nicola

Hm, I tried adding the two lines to my hibernate.cfg.xml.

The database is UTF-8 and the correct charset are stored in there, but when I try to retrieve the data in a JUnit test and write it out, it is still wrong. Any other ideas?