At my company, we’re using Ibatis to do operations on Oracle databases. As most of our software is designed to be international, we keep our XML files in UTF-8 encoding. Recently we discovered that Ibatis had some trouble parsing the XML files when we were using diacritics in them. As it turns out, Ibatis 2.2 actually ignores the “UTF-8” setting in the XML file header altogether.
This was actually reported as an issue at apache’s issue tracker, and fixed in Ibatis release 2.3 and upward. In the meantime, if you can not swich to a new release because of tight deadlines and no time for regression tests, you can set the file.encoding property to UTF-8, because then Ibatis will parse the XML in the correct encoding.