Right now we assume UTF-8 for everything. Files uploaded through the browser typically don't have text encoding specified at all, so we need to do sniffing in that case (or provide an explicit override). Then we need to do conversion when generating the full-text content for meta-files.
Right now we assume UTF-8 for everything. Files uploaded through the browser typically don't have text encoding specified at all, so we need to do sniffing in that case (or provide an explicit override). Then we need to do conversion when generating the full-text content for meta-files.