Feedmark

Version 0.17. Subject to change in backwards-incompatible ways without notice.

Feedmark is a format for embedding structured data in Markdown files in a way which is both human-readable and machine-extractable. The structured data is intended to be "curational" in nature. Articles in Chrysoberyl, Some Games of Note, and other collections of Cat's Eye Technologies' are written in Feedmark.

Informally, the format says that every h3-level heading in the Markdown file gives the title of an entity, and may be followed immediately by the entity's "plaque", which is a bullet list where every item is in the form "field-name: field-value".

Why Feedmark?

In the same way that a Markdown file is still a readable text file, which is nice, a Feedmark file is still a readable Markdown file, which is still a readable text file, which is nice.

While some structured data formats like YAML are fairly easy to read, many web-based file viewers (such as GitHub, GitLab, Forgejo, klaus, etc.) will automatically format Markdown as HTML, making it that much nicer.

Example Feedmark documents can be found in the eg/ directory, and in the projects mentioned above.

Implementation

This repository contains a Python 3 program, feedmark, which is a reference implementation of a processor for the Feedmark format.

To use it, you can clone this repository and run it as bin/feedmark from the directory of your clone, or you can put the bin directory on your executable search path, and run it as feedmark anywhere.

Or you can install it using pip:

pip install Feedmark==0.17

(Depending on your needs, you may wish to establish a virtual environment first. How to do this is outside the scope of this document.)

feedmark is currently able to do the following things:

Parse and check Feedmark documents

Loading documents will always check that they are minimally well-formed.

feedmark eg/*.md

You can also check documents against a Feedmark schema, which is simply another Feedmark document, one in which each entry describes a property that entries should have.

feedmark eg/*Sightings*.md --check-against=eg/schema/Llama\ sighting.md

Convert Feedmark documents to various formats

The original use case of this tool was to generate an Atom (née RSS) feed of entries in a document:

feedmark "eg/Recent Llama Sightings.md" --output-atom=feed.xml
python3 -m http.server 7000 &
python3 -m webbrowser http://localhost:7000/feed.xml

It can now also output entries as JSON, indexed by entry, or by property, or by publication date:

feedmark --output-json eg/*.md
feedmark --by-property eg/*.md
feedmark --by-publication-date eg/*.md

...or as Markdown or HTML (using the toc extension, which generates link anchors on headings compatible with the ones generated by GitHub and which are now in widespread use).

feedmark --output-markdown eg/*.md
feedmark --output-html eg/*.md

Rewrite documents in-place

Feedmark documents can be rewritten in-place; they will be parsed as Feedmark, and then written out as Markdown, to the same filename that were read in as input. Note! This is a destructive operation, and it isrecommended that the files being processed this way are kept under version control such as git, as this will easily allow the changes to be reverted if necessary.

feedmark --rewrite-markdown eg/*.md

Interlink documents

Markdown supports "reference-style" links, which are not inline with the text.

feedmark can rewrite reference-style links that match the name of an entry in a previously-created "refdex", so that they can be kept current and point to the canonical document in which the entry exists, since it may exist in multiple, or be moved over time.

feedmark eg/*.md --output-refdex >refdex.json
feedmark --input-refdex=refdex.json --rewrite-markdown eg/*.md

See also

Feedmark is a subset of Markdown, which is something it has in common with Falderal, however it has decidedly different goals.

See TODO.md for planned features and HISTORY.md for a record of features added in past versions.