Token creation speedup

Before:
$ python -m timeit -s 'from mwparserfromhell.parser.tokens import Text' 'Text(text = "abc")'
1000000 loops, best of 3: 0.91 usec per loop

After:
$ python -m timeit -s 'from mwparserfromhell.parser.tokens import Text' 'Text(text = "abc")'
1000000 loops, best of 3: 0.223 usec per loop

Patch mostly by @spointy.

* Removed unittest2 dependency on Python >2.6.
* Moved discover_tests.py into tests/.
* tokenizer.c: Fixed errors noted by -Wshorten-64-to-32.

setup.py: Raise Exception on unsupported platforms

Requested by Σ.

Reported by Σ.

**Breaking change**; closes #51.

Bump version to 0.4.dev.