This repository has been archived on 2026-07-21. You can view files and clone it. You cannot open issues or pull requests or push a commit.
Files
cloud-reader/docs/backend/010-open-library-metadata.md
T
crisleo94 6b4c0c43f8 feat: uv config other feats
- add uv configuration for the backend
- update frontend to make auth work
- add new auth endpoints
- add bookmars feat
- add reader feat
2026-06-03 22:06:01 -05:00

118 lines
3.6 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# 010 — Open Library metadata on import
**Status:** Implemented
**Created:** 2026-06-03
## Objective
After a user uploads an EPUB/PDF (`POST /api/books/ebooks/`), enrich the `EBook` with metadata and a cover from [Open Library](https://openlibrary.org/developers/api), using only **title** and **author** from the upload form. Prefer **Spanish** editions when available; fall back to English/any.
Upload must **never fail** if Open Library is down or no match is found.
## Trigger
- **Automatic:** `EBookUploadSerializer.create()` calls `enrich_ebook_metadata(ebook)` after file save.
- **Manual:** `POST /api/books/ebooks/{id}/enrich-metadata/` re-runs enrichment (owner only).
## Open Library usage
### Search
`GET https://openlibrary.org/search.json`
| Param | Value |
|-------|--------|
| `title` | User-provided title |
| `author` | User-provided author |
| `lang` | `es` (primary) or `en` (fallback) |
| `limit` | `5` |
| `fields` | `key,title,author_name,cover_i,first_publish_year,subject,language,edition_key,number_of_pages_median,publisher` |
Primary pass also uses query filter `language:spa`. Fallback omits language filter.
### Covers
`GET https://covers.openlibrary.org/b/id/{cover_i}-L.jpg` — downloaded and stored on `EBook.cover_image`.
## Match scoring
| Score | Behavior |
|-------|----------|
| ≥ 0.8 | Apply OL title/author + cover + full metadata |
| 0.6 0.8 | Metadata + cover only; keep user title/author |
| < 0.6 | `match_status: not_found`; no field changes except metadata stub |
Author overlap + title similarity (normalized strings, `difflib.SequenceMatcher`). Prefer hits with `cover_i`.
## Data model
No migration. Uses existing fields on `EBook`:
- `metadata_json` — full enrichment payload (see below)
- `cover_image` — downloaded cover file
- `title` / `author` — updated when match score ≥ 0.8
### `metadata_json` shape
```json
{
"source": "openlibrary",
"matched_at": "2026-06-03T12:00:00+00:00",
"match_language": "es",
"match_score": 0.92,
"match_status": "matched",
"user_input": { "title": "...", "author": "..." },
"openlibrary": {
"work_key": "/works/OL...",
"edition_key": "...",
"title": "...",
"authors": ["..."],
"cover_id": 12345,
"cover_url": "https://covers.openlibrary.org/b/id/12345-L.jpg",
"first_publish_year": 1605,
"subjects": ["..."],
"languages": ["spa"],
"publishers": ["..."],
"number_of_pages_median": 320
}
}
```
## API changes
### Upload response (unchanged path)
`POST /api/books/ebooks/` — response may include populated `cover_image` and updated `title`/`author` after sync enrichment.
### Detail
`GET /api/books/ebooks/{id}/` — adds read-only `metadata` (alias of `metadata_json`).
### Manual refresh
`POST /api/books/ebooks/{id}/enrich-metadata/` — returns updated `EBookDetailSerializer` payload.
## Configuration
| Env var | Default | Description |
|---------|---------|-------------|
| `OPENLIBRARY_ENABLED` | `true` | Kill switch |
| `OPENLIBRARY_PREFERRED_LANG` | `es` | Primary `lang` param |
| `OPENLIBRARY_FALLBACK_LANG` | `en` | Fallback `lang` param |
| `OPENLIBRARY_TIMEOUT_SECONDS` | `5` | HTTP timeout |
| `OPENLIBRARY_USER_AGENT` | `CloudReader/1.0` | User-Agent header |
## Code layout
```
backend/apps/books/services/
├── openlibrary.py # Search, scoring, cover download
└── metadata.py # enrich_ebook_metadata orchestrator
```
## Verification
1. Upload with title `Don Quijote`, author `Cervantes` → cover + Spanish-friendly metadata.
2. Upload with nonsense title/author → 201, no cover, `match_status: not_found`.
3. `POST .../enrich-metadata/` on existing ebook refreshes metadata.