Support Markdown import

[finkef]

Closes #867.

This PR adds an action for markdown import. It makes use of marked for parsing markdown into tokens and transform them into a supported plaintext format to be used with importText.

Tests are used to mainly cover the convertMarkdownToText function in favor of state mutations to decouple parsing markdown and handling potentially yet unsupported thoughts.

Open topics:

• Determine how to trigger the action from the UI.
• Support currently unsupported attributes, e.g. =blockquote, =code and =image.

[raineorshine]

Thanks, this looks very good. The token processor is well-documented and easy to read, and the test coverage looks great.

One thing is that separate lines need to be parsed as separate thoughts. There is no difference between a paragraph, newline, or list item in em. The only exception should be blockquotes and code blocks, where newlines are allowed, as we discussed, and will be supported in the future. Otherwise, they should be parsed as separate thoughts. I've pushed a couple simple tests to demonstrate. Another test should probably be added for nested multiple lines.

As for the UI, currently all types of supported content (mainly text and html) are imported via paste or by dragging and dropping a file into the editor. Both trigger importText. Ideally the user does not need a separate mechanism to import markdown. Can we autodetect markdown to determine whether their text needs to be passed through convertMarkdownToText?

[finkef]

Removed collapse pass and instead handling scope in initial pass. All tests passing with minimum ::/=scope application, pretty happy with this refactor.

[raineorshine]

Looking good!

I did find one case that's not working as expected. There is a ::/=scope false positive. I pushed a test in 218a63b.

[finkef]

Fixed all comments and failing tests apart from the open TS target lib discussion.