Add core.pandoc

[bandithedoge]

Very WIP and very temporary solution for reliably exporting Neorg documents into other formats via Pandoc. It will be deprecated once https://github.com/Simre1/neorg-haskell-parser is stable. You can test it right now, but it requires a pandoc binary to be available in your $PATH. This unfortunately introduces a rather large Haskell dependency, therefore it does not completely replace core.export yet (although it will be deprecated by this module in the future).

This module works by walking the AST of an norg document using tree_map_rec and transforming it into Pandoc's native JSON representation with vim.json.encode, making it very efficient¹. This is only an exporter; it is not a proper parser and does not allow for converting other formats to norg.

If you're eager to try it out right now, add the module to your config, open an norg file and run :Neorg pandoc json. Don't panic if it errors out with node types as messages - those are just unimplemented and will not be included in the output. This will output JSON to json.json in your current working directory which can be read by Pandoc, for example: pandoc json.json -o test.md. Keep in mind that this interface probably won't survive very long.

Implementation status (copied from https://github.com/Simre1/neorg-haskell-parser#implementation-status)

• Attached Modifiers: 🚧
• Intersecting modifiers: ❌
• Detached Modifiers: ❌
  • Heading: 🚧
  • List: ❌
    • Unordered lists: ❌
    • Ordered lists: ❌
    • TODO lists: ❌
    • Other list types: ❌
  • Quotes: ❌
  • Markers: ❌
  • Insertions: ❌
• Definitions: ❌
• Tags: 🚧
  • Carryover tags: ❌
  • @document.meta: 🚧
  • @comment: ❌
  • @toc: ❌
  • @ordered: ❌
  • @color: ❌
  • @name: ❌
  • @table: ❌
  • @embed: ❌
  • @code: ❌
  • @math: ❌
  • @image: ❌
  • Custom: ❌
• Trailing modifiers: ❌
• Escaping: ❌
• Horizontal line: ❌
• Links: ❌
• Anchors: ❌

Legend

• Implemented: ✅
• Work in Progress: 🚧
• Not implemented: ❌

Footnotes

1. Source: just trust me bro. There is currently no way to properly benchmark it since neither this module nor @Simre1's parser fully implement the spec. For now just believe me when I say it's fast enough for you not to notice it. ↩

[raj-magesh]

Sorry if this isn't the best place to discuss this, but I was wondering if this could be turned into a custom reader/writer for Pandoc since the Haskell parser project appears to be temporarily stalled. I'm interested in helping work on this!

[vhyrro]

Since there are WIP implementations of true pandoc parsers I think it's safe to close this issue as "superseded by x". Currently the following parser has worked for almost any document up to level 3 for me! The official pandoc parser made in haskell will probably follow.