Automatically extract links between topics

[chartgerink]

Topics start accumulating cross references that show their interrelatedness. These associations can be used to create a visualization of the topics and how they are interconnected, and what topics relate to each other.

It would be useful to create a script that automatically goes through these to identify them and put them into a machine-readable list. Currently these references occur in all content/*.md files, and are the links that refer to other .md files (so a regex could be \[*\]\(*.md\) to extract these links).

Need to think about this, and discussion/PRs/tips welcome!

[chartgerink]

Okay, started working on this and optimized the regex grep -E "(\w+-?)+.md" $FILE. First commit of this in bd03ab0c83526fcf6b1f97da0dc05752ee33b248 --- this script creates tidy data in links.csv but can only be figured out exactly how to output this when we know how the visualisation is created in #2

./scripts/associated-topics.sh .md content/*.md

To-do

• [x] Update (if necessary) to restructure data output for visualisation
• [x] Add script to run in last echo call

[chartgerink]

Seems like I have to take these steps given the progress mentioned in #2

• [x] collect all topics in content/*.$EXT
• [ ] Prettify topic strings to push (no extension, capitalized)
• [x] push all topics to assets/graphFile.json its nodes object
• [x] make template link object
• [x] For each content/*.$EXT file
  • [x] set source index for topic
  • [x] find all references to other topics in topic
  • [x] identify target index of referenced topic
  • [x] push all links into assets/graphFile.json

[chartgerink]

NodeJS kept stalling on me and making my computer fan at massive RPM. Updated the regex to (\w-?)+.md and that works 👼

[chartgerink]

Added hyperlinking to the relevant files in 8af6badf6fa7f72ed2f5a3f39f6e86f4ca1d2196 (still refers to .md files, should be updated to .md.html in combination with work on generate-pages branch)

[chartgerink]

Completed and on master.