Changing logic for converting markdown to json and back

[andrewshell]

Here's what I felt should be changed:

1. In stringify there was a typo where you referenced opmlToJs and the variable was all lowercase.
2. At the top of the markdown file, there can be page-level attributes that should be stored as head attributes
3. Attributes under a headline have tabs for the indenting, but then two spaces so they align with the text of the headline.

• You were not removing the spaces so an attribute like collapsed:: true was turned into
• OPML of _ collapsed="true" which is not valid

I've created a small project that takes a sample LogSeq markdown file and converts it to JSON, OPML, and a new markdown file and compares it with the original. https://github.com/andrewshell/test-opml-md

[scripting]

@andrewshell --

I started a thread for discussion of interop between Drummer and LogSeq et al. Whatever changes you feel are necessary must be discussed there, not here.

scripting/drummerRFC#4 (comment)

[scripting]

@andrewshell -- I looked into the first item in the list, and it's unclear to me how that code ever worked, but it does work, I use that routine all the time. Very weird.

Anyway I hope you do join that other thread. We're going to get this right, it's going to be iterative, and slow, and done with enough consideration.

[scripting]

OK -- I did break it, it was a stupid error, of course, but it's now fixed. That's for catching that. ;-)