(we are switching to a new theme, don't worry if things look different for a while!)

Site Navigation

Home
Search
All Pages
Recent Changes

Plans

Massive Wiki Roadmap
Pier2Pier
Massive Wiki Builder redesign

For Testing

wiki link test page
Mistletoe parser test page
Mistletoe & the ampersand story

For Fun

The Walrus

2023-01-20 MWB Program flow

This note outlines the steps taken to build a Massive Wiki static website. The code outlined is mwb.py, version 2.1.0.

The mwb.py main function steps:

• initialize: parse command line arguments

• store configuration

• get Jinja2 env

• set up lunr filename and sitepath indices

• render the wiki:

  • remove and recreate output directory

  • generate dict of filenames and wikipaths to support wikilinks (first loop through all files)

  • copy wiki to output; render .md files as HTML

    • second loop through files
      • parse Markdown files (frontmatter and markdown body)
      • collect git commit messages and times
      • append to all_pages list

  • build lunr search index (if specified)

  • set up search web page

  • build all_pages web page

---

cf. MWB-dataflowV3.png

---

Pete's notes, 2023-02-17

Paths for all non-hidden Markdown files

• absolute file path of source file
  • /Users/peterkaminski/Documents/GitHub/developer-massive-wiki/Wiki Theory and Culture/Link As You Think.md
• absolute file path or directory for output of .md, .html, .json
  • /Users/peterkaminski/Documents/GitHub/developer-massive-wiki/.massivewikibuilder/output/Wiki_Theory_and_Culture/Link_As_You_Think.md
  • /Users/peterkaminski/Documents/GitHub/developer-massive-wiki/.massivewikibuilder/output/Wiki_Theory_and_Culture/Link_As_You_Think.html
  • /Users/peterkaminski/Documents/GitHub/developer-massive-wiki/.massivewikibuilder/output/Wiki_Theory_and_Culture/Link_As_You_Think.json
• absolute web path (with optional rootdir) (with non-safe r'([ _?\#]+)' characters replaced with _)
  • /rootdir/Wiki_Theory_and_Culture/Link_As_You_Think.html

Files

All non-hidden files of any type:

• are copied from wiki to output

All non-hidden Markdown files:

• name of file is used as a key in the list of wikilink targets
• are entered into the search index
• file is parsed and
  • a .html version is saved to the location that will be next to the .md file
  • a .json extract of the YAML frontmatter is saved to the location that will be next to the .md file

Wikilinks

If we accumulate all links seen during Markdown parsing, after all Markdown files are parsed, we can compare that to the list of wikilink targets, and determine:

• unlinked files (could write an HTML report file)
• links to non-existent files (each of which could be created as a stub HTML file)

These lists could also be output as log messages.

MassiveWikiBuilder V3-pseudocode

---