// the take
A CLI that scrapes the web serial Worm, its sequel Ward, and the prequel Glow-worm from their respective sites and compiles them into a clean EPUB. It's for people who want to read these specific free web serials in an ebook reader instead of a browser.
The text-fixup pipeline (lib/convert-worker.js plus per-book .subs files) is doing real work: normalizing inconsistent chapter titles, fixing typos and punctuation across hundreds of chapters scraped from a wordpress blog. The chapter-title scheme is unusually well thought out — it offers numeric, original, or character-name title modes specifically to manage spoilers in the interludes, which shows someone who actually cares about the reading experience rather than just dumping scraped HTML into XHTML templates. It ships CI (test.yml, publish.yml) and an actual test suite (test/cli.js), and npm-shrinkwrap.json pins the dependency tree for reproducible installs.
It's a scraper against undocumented third-party markup (a WordPress blog and a bespoke site) with no visible resilience story — if either source changes its HTML structure, scraping silently breaks until someone notices and patches book-data/*.js. Each title is hardcoded in its own book-data file, so this isn't a general web-serial-to-EPUB tool; adding a new serial means writing code, not configuration. There's no mention of request throttling or caching between runs, so re-running the tool re-fetches everything from the source site each time. At 272 stars it's a one-maintainer hobby project for a finished serial, so don't expect fast turnaround if the source sites change layout.
We read the README, the directory tree and a key source file in October 2026. Nobody pays to be listed. How we judge →