Short UTF-8 Sequences will not be caught if InputStream does not support Length property

[R-maan]

Due to the limitation of StreamReader.Read() where it returns -1 if no more characters are available - regardless whether there are any sequence left or not, we are using Stream.Length property to check for short sequences of UTF-8 inputs.

But not all the implementations of Stream will support Length.
If an implementation does not support Length property, short utf-8 sequences will not be caught.