pytest is the preferred way to write and run unit tests these days and
it has a cleaner interface - so lets switch to it. The tokenizer tests
especially are much easier to read/understand.

This was mostly done with find/replace regexes and then cleaned up
manually.

Fixes #221.

Step one of refactoring - making SmartList into its own
package, with each class having its own file.  No code
changes were made.

Note that SmartList and ListProxy import each other,
so had to import SmartList as a full package name
rather than use from ... import ... construct.

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

- Add a _SliceNormalizerMixIn to properly handle slices.
- Use floor division when applying key.step.
- Implement sort() without 'cmp' parameter.
- Fix bytes(list) behavior.
- Children of _ListProxies are now _ListProxies, not regular lists.

- Skip CTokenizer tests if CTokenizer is not available.
- TestStringMixin: Don't make assumptions about default encoding.
- Add urllib stuff to mwparserfromhell.compat.
- Fix compat.py's line endings.
- gen.next() -> next(gen)
- assert*Equals() -> assert*Equal()

Add from __future__ import unicode_literals to a few files.