Search code examples
javautf-8character-encodinglatin1

MySQL latin1 to UTF-8 using Java & Hibernate & JPA

I have a database that is in charset=latin1 and collation latin1_swedish_ci. The user has entered UTF-8 characters in this connection ( Greek characters ) and in PHP he can read everything just fine.

But when I try to read the database using JAVA + JPA + Hibernate ( all Latest versions ) the characters are totally broken.

Note that I have already tried my jdbc string with:

...?useUnicode=true&characterEncoding=latin1&connectionCollation=latin1_swedish_ci
...?useUnicode=true&characterEncoding=UTF-8&connectionCollation=utf8_general_ci
...?characterSetResults=ISO8859_1
...and combinations of those

but still I can't read the characters.

The best I have managed is to use:

byte ptext[] = myString.getBytes(windows-1252); 
String fixed = new String(ptext, UTF_8);

with:

?useUnicode=true&characterEncoding=UTF-8&connectionCollation=utf8_general_ci

But still there are many characters that in eclipse's output are "?" and in log4j output everything is broken.

Any suggestions?

Solution

  • Solution:

    SELECT CONVERT(CONVERT(CONVERT( column_name USING latin1) USING binary) using utf8) FROM...

    But it ties you down to Native SQL. You cannot use JPA Queries. There is no other way. Only MySQL knows how to convert what it converted when the data was entered in the database.