O'Reilly EPUB Exporter

Turn any O'Reilly Learning book into a clean, e-ink-optimized EPUB in one click — whether you pay for O'Reilly or read it free through your public library.

Yes, really: many public and university library cards include the complete O'Reilly catalog, served through a library portal — and this extension runs there out of the box. Downloads are strictly for personal use — see the Disclaimer.

Two ways to read O'Reilly, one extension

Most tools assume a personal learning.oreilly.com subscription. But a huge number of libraries license the entire O'Reilly catalog for free with a library card, served through an EZproxy gateway on a different hostname. This extension is built for both — same one-click download, same quality report, same validated EPUB:

	Personal subscription	Library card
Cost to you	O'Reilly subscription	free with your library card
Where you read	learning.oreilly.com	your library's O'Reilly portal (EZproxy)
Works out of the box	yes	yes — Seattle Public Library
Other libraries	—	a two-line edit
One-time setup quirk	none	one Chrome "Safety warning" click-through

Session expiry is even diagnosed per mode — a library user is told to sign back in through the library portal, not to "log in to O'Reilly".

 

Features

Built for e-ink readers

• Code blocks restyled for grayscale: typographic emphasis instead of color-based syntax highlighting
• All images kept at original quality; EPUB 2 NCX fallback so older readers (tested on Boox) handle the TOC
• Output validates clean in epubcheck 5.1.0 — 0 errors

Nothing goes missing silently (see it)

• Quality report after every download: missing chapters, failed images, and stylesheet problems are counted and listed, with one-click re-download
• Pre-packaging integrity gate (OPF ↔ ZIP ↔ spine reconciliation) — a structurally broken EPUB is blocked, never delivered
• Handles both classic XHTML books and O'Reilly's newer HTML-fragment format (epub:type structural semantics preserved); cover always included via an API fallback

A first-class citizen in your Calibre library

• Rich metadata: language, publisher, subjects, description, publication date
• Unicode-safe filenames — CJK and accented titles survive intact
• Runs entirely in your browser against your own session — no backend, no third-party servers

Install (about a minute, no build step)

You need Chrome (or any Chromium browser) and access to O'Reilly Learning — a personal subscription or a library that provides it (see Library access).

1. Download oreilly-epub-exporter-x.y.z.zip from the latest release and unzip it.
2. Open chrome://extensions/ and turn on Developer mode (top-right toggle).
3. Click Load unpacked and select the unzipped oreilly-epub-extension folder.

The extension isn't distributed through the Chrome Web Store, so it installs via Chrome's standard developer-mode flow. There is no build step and no server — what you unzip is the complete, readable source.

Install from source instead

git clone https://github.com/ludengz/oreillybook-epub-exporter.git

Then Load unpacked → select oreilly-epub-extension/.

Usage

1. Open any book — on learning.oreilly.com, or through your library's O'Reilly portal. (Library users: if Chrome shows a red "Safety warning" on your library's link, see the one-time click-through below.)
2. Click the extension icon — the popup shows the detected book title and author(s).
3. Click Download EPUB.
4. The extension fetches all chapters, images, and stylesheets, then packages them into an EPUB.
5. The file downloads automatically when complete, followed by a per-book quality report.

Popup States

Book Detected	Downloading	Quality Report	Report with issues
Title & author, ready to download	Progress with live chapter/image counts	"Everything downloaded"	Exactly what failed, with retry

Library access

Many public and university libraries provide O'Reilly Learning through an EZproxy gateway, which serves the same content from a rewritten hostname — learning.oreilly.com becomes learning-oreilly-com.<proxy-host>. The extension supports this natively.

Didn't know your library card could include the full O'Reilly catalog? Many do — search your library's "databases" or "e-resources" page for "O'Reilly".

Out of the box: Seattle Public Library (learning-oreilly-com.ezproxy.spl.org).

Chrome shows a "Safety warning" the first time

Chrome's lookalike-domain heuristic sees a well-known domain embedded in a subdomain and flags it. The hyphenated shape is exactly what EZproxy must produce to match its wildcard TLS certificate, so the warning is a false positive — but Chrome cannot know that, and the extension cannot run while the warning is on screen.

Click Details → Continue to <host> (unsafe) once. Chrome remembers the choice for that site. This happens even when you follow your library's own link.

Adding another library

Two files, one hostname:

1. oreilly-epub-extension/manifest.json — add https://learning-oreilly-com.<your-proxy-host>/* to host_permissions and web_accessible_resources[0].matches, and both /library/view/* and /library/cover/* variants to content_scripts[0].matches.
2. oreilly-epub-extension/lib/path-utils.js — add the same hostname to PathUtils.ALLOWED_IMAGE_HOSTS.

Then reload the unpacked extension. A test (tests/path-utils.test.js) fails if you update one file and forget the other.

To find your hostname, open a book through your library and read the address bar.

What is not supported

• Proxy-by-port gateways (https://ezproxy.example.org:2443/...) and http-only proxies.
• Rewriting proxies other than EZproxy (WAM, MUSE) may work if you declare their hostname, but are untested.
• OpenAthens / institutional SSO needs no changes: those log you in on the real learning.oreilly.com, which the extension already supports.

The personal-use disclaimer below applies to library access too. Check your library's terms.

How It Works

• Content script runs on learning.oreilly.com and the declared library-proxy hosts, fetching book content via same-origin API calls (session cookies — O'Reilly's or your library's — are included automatically)
• Book metadata (title, authors, language, publisher, subjects, description, publication date, cover URL) is retrieved from the O'Reilly search API and normalized before entering the EPUB
• Session expiry is detected in both modes — including EZproxy's login redirect, which is invisible to a naive fetch — and reported with mode-specific recovery guidance
• Service worker relays progress, updates the badge, and proxies cover/CDN image fetches behind a host allowlist covering both origins
• JSZip packages everything into a valid EPUB 3.0 file

For developers — project structure, testing, design spec

Project Structure

oreilly-epub-extension/
├── manifest.json          # Chrome Extension Manifest V3
├── content.js             # Main orchestration — fetching, parsing, EPUB assembly
├── background.js          # Service worker — message relay, badge, notifications
├── popup.html/js/css      # Extension popup UI
├── lib/
│   ├── jszip.min.js       # EPUB packaging
│   ├── path-utils.js      # Pure helpers: path resolution, filename sanitizing, host allowlist
│   ├── epub-builder.js    # EPUB structure generation (OPF, NCX, TOC) + metadata normalization
│   ├── epub-validator.js  # Pre-package integrity gate (manifest/ZIP/spine reconciliation)
│   ├── fetcher.js         # HTTP fetching with retry, backoff, session-expiry detection
│   └── eink-optimizer.js  # E-ink optimization + XHTML namespace repair
├── styles/
│   └── eink-override.css  # E-ink display overrides injected into EPUB
├── icons/                 # Extension icons
└── tests/                 # Browser-based test runner

Testing

Tests run in a real browser (no Node.js test runner):

# Serve the repository root, then open the runner
python -m http.server 8765
# → http://localhost:8765/oreilly-epub-extension/tests/test-runner.html

The suite covers pure helpers, the download lifecycle (via a chrome API mock), and an integration pass that unpacks generated EPUBs to audit OPF/ZIP consistency — including fixtures that run the full pipeline against a simulated library-proxy origin.

Design

See the full design spec at docs/superpowers/specs/2026-03-15-oreilly-epub-chrome-extension-design.md.

---

If this put your library's O'Reilly catalog on your e-reader, a ⭐ helps other readers find it.

License

MIT. The code is fully open and auditable — what you unzip is what runs.

Disclaimer

This extension requires legitimate access to O'Reilly Learning — a personal subscription or a library that provides it. Downloaded EPUB files are strictly for personal use only — they must not be redistributed, shared, or used for any commercial purpose. Users are responsible for complying with O'Reilly's Terms of Service, your library's terms of use, and applicable copyright laws. This project is not affiliated with or endorsed by O'Reilly Media.