⨳ Some Jekyll Tools ⨳

2024 Apr 10, Wednesday

Seeing as how a Jekyll site is really just piles of marginally-documented markdown files, scripting against that pile has some interesting/useful possibilities.

refresh-sitemap.sh

First off, I wanted to extract the frontmatter from all of the markdown files. This will dump out the file path along with its frontmatter as-is (minus the frontmatter markers) to a file named sitemap.yaml.

This map will bet borked if you are one to add a single like consisting of --- to add horizontal lines in your markdown pages. (If you add some leading spaces to those lines, this should work fine.)

refresh-tag-pages.sh

I try to be consistent and thorough when it comes to tagging, and I think I do a fairly-comprehensive tag normalization on this site. But I wasn’t too confident that (1) I captured all tags and (2) created a tag page for each tag in use. Tag information is in the frontmatter, which means it’s in the sitemap!

I use yq (think jq, but works with .yaml files) to pull a listing of tags being used and create a tag page for each unique tag.

As yq reads the sitemap, the filter iterates over all files .[], then over all tags .tags[], and outputs each tag [.]. That result is then striped of the leading item marker (^- ) and sorted uniquely.

Because tags can contain spaces, we have to tell bash to treat new lines as the list item separator in the for loop. Each distinct tag is then slugified on the permalink line and for the output filename.

With the initial rm wildcard and subsequent file write, this ensures all used tags have a corresponding tag page.