Perform character set conversion after item formatting step instead of before it. This should remedy the potential issue of formatting not taking UTF8 multibyte into account, as our formatting unfortunately does not support multibyte encoding. This way the data should stay in ISO-8859-1 format just up to outputting it.