updated some testcases for "strm << container_format() % c" format. Now, precision of doubles and floats can be manipulated in one line.