Handling of special characters in getchar

[pkch]

When the user presses arrow up, the click.getchar() returns b'\xe0'. This behavior is probably good, but I guess it should be documented (along with other special characters), since it's not a valid utf-8 value even though the terminal encoding is utf-8. (The naive click.getchar().decode('utf-8') would crash on this example with UnicodeDecodeError: 'utf-8' codec can't decode byte 0xe0 in position 0: unexpected end of data.)

[fdavis]

hey @pkch is this still an issue for you? I don't seem to get this running on master (different python3 invocation because I seem to get a str type back in py3)

(.clickve) [~/git/click]$ python2 -c 'import click;print(click.getchar().decode("utf-8").__repr__())'
u'\x1b[A'
(.clickve) [~/git/click]$ python -c 'import click;print(bytes(click.getchar(),"utf-8").__repr__())'
b'\x1b[A'