Just a note: html entity is a non-breaking space, so is 0xA0 character code in Unicode. So technically this (HtmlDecode) is working as expected. We need to make sure all non-breaking spaces are converted to regular spaces before importing to DB