ShellStream.Expect does not remove all matched data when multi-byte character encoding is used

[JonathanFortier-eaton]

When we receive multi-bytes characters from a SSH server (for example, 'é' encoded in UTF-8 giving bytes [0xC3, 0xA9]), the ShellStream.Expect methods remove less bytes from the _incoming byte queue than what was really matched, so that the next read will return some of the data that should have been removed by Expect.

The bug seems to be when the functions call "_incoming.Dequeue()": they assume that the number of bytes to remove is the same as the number of character matched, which is incorrect for multi-byte encoding.

Simple fix could be to calculate the number of byte used by the matched string (_encoding.GetByteCount(...)) like what ShellStream.ReadLine does. However, this can still return a value different of what was actually read (see https://stackoverflow.com/q/9740553).

A more complete fix could be to use "_encoding.GetDecoder().Convert(...)" (see https://msdn.microsoft.com/en-us/library/h6w985hz(v=vs.110).aspx) to calculate the number of bytes that correspond to the expected number of characters.

What do you think?

[IgorMilavec]

I have created a discussion #926 for all ShellStream related issues.

[jscarle]

@IgorMilavec @WojciechNagorski This was fixed by #1315.