Skip to main content
metabook-py

metabook-py

PR Matrix Release Python License: MIT

📚 X-ray any book's structure — chapters, paragraphs, sentences, clauses, words — as clean JSON. REST + MCP, Project Gutenberg search or your own EPUB. No book text ever leaves the API.

API docs · MCP server · Desktop example · Workflow

Metabook desktop example searching Pride and Prejudice and exploring its structural schema

Overview

Metabook is a Book Structure API: give it a title, an ISBN, a Gutenberg ID, or an EPUB file, and it returns the book's structural schema — what kind of book it is and how it's shaped — without ever returning the text itself.

Under the hood it locates the book via Gutendex (or walks your uploaded EPUB's spine), downloads and cleans the text, detects the structural schema, and counts everything at every level.

title / ISBN / Gutenberg ID ──► Gutendex ──► fetch + clean ──┐
                                                             ├──► detect schema ──► counts-only JSON
your .epub ──► Vercel Blob ──► package doc + spine XHTML ────┘

✨ Features

  • 🔍 Fuzzy search — find books by title or author via Gutendex, or go straight to an ISBN / Gutenberg ID
  • 🧬 Schema detection — classifies each book as scripture, sectioned_book, standard_book, essay_collection, or flat, with a confidence rating
  • 🔢 Counts, not content — chapters, paragraphs, sentences, clauses, and words per node; a word node is just its index. No book text is ever included in a response
  • 📤 Bring your own EPUBPOST /api/books/upload stores the file in Vercel Blob and analyses it with the same pipeline
  • 🔌 Two interfaces, one service layer — a FastAPI REST API and a FastMCP server share the same core
  • 🖥️ Native desktop client — a GPUI example app with an animated structure tree and a synced JSON code editor

Detail levels

The detail parameter controls how deep the returned tree nests beneath each paragraph:

detail Nested nodes
paragraph paragraphs only (counts for the rest)
sentence + sentence nodes
clause + clause nodes
word + word nodes (index + position only)

🚀 Quick start

make setup              # install dependencies (uv sync)
make dev                # run the API with auto-reload on :8000
make test               # run the test suite
make docker-up          # or run it via docker compose

Then open the interactive docs at http://127.0.0.1:8000/api/docs, or try:

curl "http://127.0.0.1:8000/api/books/structure?title=Pride+and+Prejudice&detail=sentence"

EPUB uploads

Uploads need a Vercel Blob read-write token in the environment (or .env / .env.local):

export BLOB_READ_WRITE_TOKEN=vercel_blob_rw_...

If the project is linked to Vercel, vercel env pull writes the token to .env.local (gitignored), which the app loads automatically — values there override .env.

🌐 REST API

Method Endpoint Purpose
GET /api/books/structure Analyse by title, isbn, or gutenberg_id (+ detail)
GET /api/books/structure/schemas List the supported structural schemas
POST /api/books/upload Upload an EPUB and analyse it
GET /health Liveness + cache stats

Interactive OpenAPI docs live at /api/docs.

🤖 MCP server

A FastMCP server is mounted at /mcp, so agents can use the same service layer through three tools:

Tool Description
search_book_structure Analyse a book by title, ISBN, or Gutenberg ID
upload_book_epub Analyse an EPUB passed as base64 or a URL
list_supported_schemas Enumerate the structural schemas

🖥️ Desktop example (GPUI)

example/ is a native macOS client built with GPUI and gpui-component: fuzzy search or EPUB upload, an animated lazily-materialised structure tree, and a read-only JSON code editor (tree-sitter highlighting, folding) that scrolls to and highlights whichever node you select in the tree. Light and dark themes, spring animations, routed form transitions.

Metabook desktop example — structure tree synced with the JSON schema editor
More screenshots
Search results Dark mode
Search matches Dark mode

Run it (API first, then the app):

make dev
cd example && cargo run

See example/README.md for the Metabook.app bundle (with app icon) and METABOOK_API configuration.

🌳 Branching and release

CI/release pipeline scaffolded from exegia/corpora-py: same branch model, GitHub Actions workflows, composite actions, and make-driven release automation.

See .github/WORKFLOW.md for the full branch model, versioning rules, and workflow reference. Quick summary:

<type>/<slug> --PR--> dev --(daily/manual)--> next --cut--> release/vX.Y.Z --draft PR--> main
                 (deleted on merge)         (preview)                       (deleted on release)

Common commands

make help                # list all targets
make ci                  # everything CI runs on a PR (lint + test)
make pack                # build the publishable wheel
make rulesets-diff       # rulesets GitHub currently has
make rulesets-apply      # push .github/rulesets/*.json

Bootstrap

There are no dev / next branches until the first wrapup. Run the Release workflow manually (Actions → Release → Run workflow), or locally:

make bootstrap-lanes

📄 License

MIT © Emmanuel De Freitas

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

metabook_py-2.0.0.tar.gz (2.4 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

metabook_py-2.0.0-py3-none-any.whl (33.4 kB view details)

Uploaded Python 3

File details

Details for the file metabook_py-2.0.0.tar.gz.

File metadata

  • Download URL: metabook_py-2.0.0.tar.gz
  • Upload date:
  • Size: 2.4 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: uv/0.12.7 {"installer":{"name":"uv","version":"0.12.7","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}

File hashes

Hashes for metabook_py-2.0.0.tar.gz
Algorithm Hash digest
SHA256 88ae1ec31184af3fbcfe6630c431d426f590c3496fb0f98708ffe27a3d7c857b
MD5 bca2a3bd13fe7d201f6cdbd105d9c0b3
BLAKE2b-256 19e0350ababa432c313a1bf8154d962ab8291d710671a4df29ea2b02172aa919

See more details on using hashes here.

File details

Details for the file metabook_py-2.0.0-py3-none-any.whl.

File metadata

  • Download URL: metabook_py-2.0.0-py3-none-any.whl
  • Upload date:
  • Size: 33.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: uv/0.12.7 {"installer":{"name":"uv","version":"0.12.7","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}

File hashes

Hashes for metabook_py-2.0.0-py3-none-any.whl
Algorithm Hash digest
SHA256 cd6bdc5fdd7a96a9b9d1bce28043d985a35aacad3f33a79bd9a1c63054ce62d1
MD5 d7b10bac6d74fc9aeb9af5b2294e0b59
BLAKE2b-256 1a7074c810fa71232be2c943cf8d5e3a8c669838df74332e82fadba8b8663023

See more details on using hashes here.

Release history Release notifications | RSS feed

3.0.0

2 files

This release

2.0.0 This release

2 files

1.0.1

2 files

1.0.0

1 file

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page