python: Implement markdown token scanner

[elchupanebrej]

🤔 What's the problem you're trying to solve?

I develop https://pypi.org/project/pytest-bdd-ng/ and try to use as many official tools of gherkin as possible. To have the possibility to execute *.md files I need an appropriate token scanner

[temyers]

Already looking at this here: #64

[mattwynne]

cucumber/gherkin-streams#56 is a blocker on this

[mattwynne]

Closing in favour of #64