Merge: nitunit: use the Markdown parser from `lib/markdown`

Merge: nitunit: use the Markdown parser from `lib/markdown`

Since the introduction of `nitmd`, `nitunit` was still using the old `docdown` parser.
This PR do the migration.

There will be a lot of unrecognized tests since the old paerser allowed between four and five spaces before a nitunit, the new one enforce the rule of 5 (couting the space after the `#` comment token).

Before:
~~~nit
#    assert true
~~~
Was recognized as nitunit, but if you ignore the `# ` string as prefix of each comment, there is only 3 spaces before the nitunit.

After:
~~~nit
#     assert true
~~~
We need 4 spaces before the nitunit plus the `# ` prefix, this gives us 5 spaces.

Note, this is already the behavior followed by Nitdoc.

Should fix #1331

Pull-Request: #1332
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Reviewed-by: Jean Privat <jean@pryen.org>