Skip to content
  • Content »
  • Lab »

Scrivener's compile supports Multimarkdown and Pandoc output.
Obsidian uses Markdown.
MkDocs renders Markdown into a static HTML site.
Pandoc renders Markdown into MediaWiki.

Sounds simple and perfect to build a documentation workflow ...
... but turns out to be a rabbit hole that I'm not sure I want to jump into.

Imagetest

Tuttle or Buttle.jpg

Tuttle or Buttle.jpg

flowchart TB
    Scrivener -- MarkDown --> Obsidian
    Obsidian -- MarkDown --> MkDocs
    MkDocs -- HTML --> GitHub_Pages
    Obsidian -- MarkDown --> Pandoc
    Pandoc -- WikiMarkup --> MediaWiki

Issues

Kartaverse Workflows / Creating Volumetric NeRFs /Equipment Needed to Explore Instant NGP:
The "Git Client" image is 'image35'

Kataverse Workflows / KickAss ShaderZ for Fusion / Natural:
The "kas_DarkBlueIceShard" image is 'image35'

TODO: docconv

https://pkg.go.dev/code.sajari.com/docconv

Misc

To maybe try some other directions: https://stackoverflow.com/questions/41183642/how-to-convert-github-wiki-into-github-pages ... but not so easy

Embed Video:
https://pypi.org/project/python-markdown-oembed/
https://gist.github.com/dmitric/6266559

Auch ein paar nette dabei:
https://chrieke.medium.com/the-best-mkdocs-plugins-and-customizations-fc820eb19759

Grundlagen:
https://www.mkdocs.org/dev-guide/plugins/
https://python-markdown.github.io/extensions/api/

Die kaputten Links:
https://github.com/Jackiexiao/mkdocs-roamlinks-plugin
Lab/Folder Zero/index

Und Bildgrößen müsste ich noch gucken - die aber vielleicht einfach als Extension?!?

Okay, MultiMarkdown still seems to be the best option. With and .opml (Titles only) I can restore the articles' order - format is simple and straight forward. Have to check what exacly happens

https://www.youtube.com/watch?v=yTMo9WkOphE
und hier ist was erklärt, das helfen könnte

https://www.youtube.com/watch?v=xisfLmWGXqA und vielleicht hier

https://medium.com/@TKalippke/all-you-have-to-do-is-type-yourself-494a2c0b2fd6

mit docker verbasteln - als beispiel
https://hub.docker.com/r/silentstorm/pandoc-mkdocs/

Another test is Fusion

\(p(x|y) = \frac{p(y|x)p(x)}{p(y)}\), \(p(x|y) = \frac{p(y|x)p(x)}{p(y)}\).

This is Lua's #!lua if x ~= 42 then -- ... inlined in text.

And some Cmd+Option+Q, Ctrl+Alt+Del keystrokes,

Config/

Scrivener

"Compile for: MultiMarkdown" and then "Duplicate & Edit Format..." the "Basic Pandoc".

Named the new format "Pandoc (nmbr73)" (or whatever you want).

Under 'Styles' add our 'Code Block (Lua)' style.

Check the 'Treat as raw markup' box.

Set 'Prefix/suffix:' to \n```lua\n and \n```\n. Enter the new line by pressing Opt + Enter in the input fields.

Todo

  • https://iandol.github.io/scrivomatic/ , resp. https://github.com/iandol/scrivomatic
  • https://michalwyrwa.org/posts/2019-06-08-pandocscripts/

Document content in Scrivener is stored in RTF (Rich Text Format).

open Sources/Kernfusion.scriv/Files/Data/59587C22-40EF-4FD1-8674-595E763625CB/content.rtf

Pandoc comes already with a reader for RDF. To show the internal (native) pandoc representation (abstract syntax tree):

pandoc -s -t native 'Sources/Kernfusion.scriv/Files/Data/59587C22-40EF-4FD1-8674-595E763625CB/content.rtf'

Binder and styles are defined in XML. All the things are then tied together with UUIDs ... so if nothing else works, I could read that raw data.

Todo

  • http://lit-n-lat.blogspot.com/2010/02/coding-xml-formats-in-cocoa_05.html#links
  • https://blog.jayway.com/2020/01/23/attributed-strings-and-rich-text-files/

It would make total sense to apply more semantic markup in Scrivener. I managed to define 'Styles' (for example for "Code Block (Lua)", "Code Block (Python)") and to create my own 'Format' (for Multimarkdown and Pandoc) to apply the respective Markdown formatting. But this process is pretty tedious and very limited in the resulting markup we can achieve with it.

Internal links, 'Title' Style, and many more gets lost when exporting from Scrivener to (Pandoc) Markdown

MkDocs

MkDocs generates static HTML sites out of Markdown sources and a single configuration YAML file.

MkDocs uses Python-Markdown, which is almost completely compliant with the original Markdown reference implementation. But MkDocs includes support for extending the Markdown syntax with Python-Markdown extensions.

Todo

  • Could be an option to creating other output formats from an MkDocs source using pandoc: mkdocs-combine

Currently the image paths are sometimes broken!?!

  • https://stackoverflow.com/questions/71074662/configuring-image-filepaths-correctly-in-mkdocs

Pandoc

  • Guter Artikel: https://ulriklyngs.com/post/2019/02/20/how-to-use-pandoc-filters-for-advanced-customisation-of-your-r-markdown-documents/
  • Auch umfangreich: https://jmablog.com/post/pandoc-filters/

Pandoc uses by default its own Pandocs Markdown flavor.

Scrivener includes MultiMarkdown support.

Check for interoperability options for Pandoc and Obsidian

Pandoc could in particular be the tool of choice to make convert the Obsidian Markdown files into WikiMarkup, MkDocs Narkdown, etc.

Obsidian

Markdown: https://www.markdownguide.org/tools/obsidian/

Todo

  • https://obsidian-publisher.netlify.app
  • https://github.com/ObsidianPublisher/obsidian-mkdocs-publisher-template
  • https://github.com/ObsidianPublisher/obsidian-mkdocs-publisher-python
  • https://github.com/ObsidianPublisher/obsidian-github-publisher
  • https://github.com/jobindj/obsidian-publish-mkdocs
  • https://github.com/jobindjohn/obsidian-publish-mkdocs

Create notes / descriptions / hierarchy for the folders in Obsidian:

  • https://github.com/xpgo/obsidian-folder-note-plugin
  • https://github.com/akosbalasko/zoottelkeeper-obsidian-plugin

Index files
- https://github.com/aidenlx/alx-folder-note
- https://forum.obsidian.md/t/folder-link-default-load-readme-md-file-just-like-github/41335

Scrivener_hierarchy01.png

Scrivener_hierarchy02.png