Unicode support differences in Oracle jdbc driver

I have replaced ojdbc14.jar with ojdbc5-11.2.0.2.0.jar in my Jav开发者_如何学JAVAa project because I need to move the support to Java 1.5 but since then all the unicode characters only allow half the length specified.

I use java.sql.DatabaseMetaData to get the table columns and their column size and validate their sizes. And I halved the size when the data types are unicode. It worked before but after I changed the driver the logic broke.

For example, a column was specified as NVARCHAR(200) before I can enter 200 characters but now I can only enter 100 characters.

My Oracle database is using "AL16UTF16" as its character set. Are there any differences in terms of unicode support between ojdbc14 and ojdbc5?