This function seems to support UTF-8 only from the source code.
Hope to support GBK at least. 
Moreover, the behavior of readMultiByte(length,'utf-8') is not exactly the same as readUTFBytes(length) in flash, the latter can automatically skips BOM. 