Just like the Windows wheels, these allow for Linux users to install
mwparserfromhell and use the faster CTokenizer without needing to
have build tools installed.

Under the hood, this uses pypa manylinux1 docker image to build and
tag the wheels, then publishes them to pypi if a new tag was pushed.

Fixes #170.

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.

Drop Python 2 support

In addition to replacing the manual version check, this will also
instruct pip to download an older version of mwparserfromhell for
users running earlier Python versions rather than just getting something
broken.

Fixes #221.

move smart_list into sub-package/multiple files

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.

ignore InteliJ IDEA files

add missing public items to __all__

Added a few more public classes, and sorted the __all__ value.
The Attribute and Parameter classes are not included as they can
be imported from the `mwparserfromhell.nodes.extras` without
linter complaining.  They could be added to this __all__ too.

fix example: wrap dict.values in list()

In Python 3, `dict.values` returns a dict_values object, which does not support indexing.

Drop support for EOL Python 2.6, 3.2 and 3.3

Add <wbr> to definitions.py

Added to MediaWiki in ff74113bea (T54468).