decoding into a bytearray then using python's .decode() method

[willlahr]

No description provided.

[willlahr]

fix for #42

[ccrighton]

This PR handles the #42 Multibyte utf-8 characters are urldecoded incorrectly bug. However, it will throw exceptions on non unicode encoding like %XX and invalid unicode encoding. I think it is worth making a bit more robust so that it doesn't throw exceptions on spurious input. I've attempted to do that with the following branch: https://github.com/ccrighton/phew/tree/utf8-urldecode.

Currently, this branch uses the official unicode replacement character 0xFFFD (? in diamond) in place of invalid utf-8 encoding. Perhaps it could ignore it. Either way it seems better than throwing an exception.

See https://github.com/ccrighton/phew/blob/utf8-urldecode/tests/urldecode_test.py for more detail on behaviour in the event of bad input.

I've also ensured that this code will run on micropython and cpython. PR #43 isn't cpython compatible.

If the consensus that throwing an exception is a better option, I'd still suggest making it cpython compatible.