feedsmith
Fast, all‑in‑one JavaScript feed parser and generator for RSS, Atom, RDF, and JSON Feed, with support for popular namespaces and OPML.
feedscout
Advanced feed autodiscovery for JavaScript. Collect feeds, hubs, and blogrolls from any webpage using multiple discovery methods.
feedcanon
Find the canonical URL for any web feed by comparing actual content. Turn messy feed URLs into their cleanest form.
feedmatchAlpha
Classify and deduplicate feed items into inserts and updates.
feedsweepAlpha
Tidy up the HTML content in web feeds. Fix feed-specific quirks so content displays in its best possible form.
- Where did these come from?
- I built them while working on Feedstand, a feed reader. I wanted to take feed processing seriously: read the specs closely, test against real-world feeds, and handle every edge case. They exist as standalone packages because the underlying problems are general. Open source and available for anyone else building in this space.
- What does ”millions of feeds“ actually mean?
- I keep a local archive of over 2 million real-world feeds — about 300 GB of RSS, Atom, RDF, JSON Feed — and scan it whenever I'm working on a parser or a normalizer. It surfaces edge cases which would be hard to identify on my own. While specifications are useful as a starting point, a corpus of this size shows how feeds truly appear.
- Why so many small packages instead of one big library?
- These problems are too different to share much code. They share a design philosophy, but each package stands alone. Install only what you need.
- Can I use these in a browser?
- Yes. Every package ships as ESM with TypeScript types, and runs in Node, Bun, Deno, and modern browsers.
- Are they open source?
- All five are MIT licensed. Source, issues, and releases live on GitHub.