Request both formats at once + Open Library 403 and LazyLibrarian fixes

README.md

@@ -11,7 +11,8 @@ Libreseerr is a book request management application for [Readarr](https://readar
 ## Features
 
 - Search for books by title, author, or ISBN
-- Request ebooks or audiobooks from separate Readarr, Bookshelf, or LazyLibrarian instances
+- Request ebooks, audiobooks, or both at once from separate Readarr, Bookshelf, or LazyLibrarian instances
+- Optional Hardcover integration for richer search/discovery metadata (falls back to Open Library)
 - Track download progress with real-time status updates
 - View quality profiles and root folders from your configured server
 - Manage and remove requests from a dedicated requests page
@@ -102,14 +103,16 @@ Select the **Server Software** dropdown to choose between Readarr, Bookshelf, or
 
 Click **Test Connection** to verify each server is reachable, then **Save**.
 
+Optionally, add a **Hardcover** API token in **Settings** to use [Hardcover](https://hardcover.app/) as the search/discovery metadata source. When no token is set, Libreseerr uses the public Open Library API by default.
+
 ![Libreseerr Settings Page](screenshots/settings.png)
 
 ## Usage
 
 1. Log in with your admin credentials (default: `admin` / `admin`).
 2. Go to the **Discover** page and search for a book by title, author, or ISBN.
 3. Click a book card to open the download dialog.
-4. Select **ebook** or **audiobook**, choose a quality profile and root folder, then click **Download**.
+4. Select **ebook**, **audiobook**, or both. Each selected format gets its own quality profile and root folder (formats whose server isn't configured are disabled). Click **Download** to submit — requesting both creates one request per format.
 5. Switch to the **Requests** page to monitor progress.
 6. Click **Refresh Status** to poll your server for the latest download status.
 
@@ -187,6 +190,20 @@ python app.py
 
 The development server starts on `http://0.0.0.0:5000` with debug mode enabled.
 
+### Testing and Linting
+
+Install the dev dependencies, then run the linter and test suite:
+
+```bash
+pip install -r requirements-dev.txt
+ruff check .   # lint
+pytest         # test suite
+```
+
+Both run in CI (`.github/workflows/ci.yml`) on every push and pull request,
+alongside a job that builds the Docker image and verifies the container serves.
+Run `ruff check .` and `pytest` locally before opening a PR — both gate the build.
+
 ## License
 
 See [LICENSE](LICENSE) for details.

TODO.md

@@ -0,0 +1,35 @@
+# TODO
+
+## Bindery backend support (upstream issue #12)
+
+**What:** Users can connect Libreseerr to a Bindery server (Service Type:
+Readarr) and pass the connection test, but every book request fails with
+`HTTP 400 {"error":"isbn or asin parameter required"}`.
+
+**Why it happens:** Bindery (`vavallee/bindery`) is *not* a Readarr API clone.
+It shares the `/api/v1` prefix and `X-Api-Key` header but deliberately splits
+lookup into different endpoints:
+
+| Purpose         | Bindery                          | Readarr (what Libreseerr sends)        |
+|-----------------|----------------------------------|----------------------------------------|
+| Free-text search| `GET /api/v1/search/book?q=…`    | `GET /api/v1/book/lookup?term=…`       |
+| ISBN/ASIN lookup| `GET /api/v1/book/lookup?isbn=…` | `GET /api/v1/book/lookup?term=isbn:…`  |
+| Author search   | `GET /api/v1/search/author?q=…`  | `GET /api/v1/author/lookup?term=…`     |
+
+Bindery only advertises arr-compatibility for `/api/queue`; the rest is its own
+contract. So `ReadarrClient.search_books` and `.lookup_by_isbn` both 400.
+
+**Decision:** This is a *new backend*, not a Readarr bugfix. Add a dedicated
+`BinderyClient` (like `bookshelf.py` / `lazylibrarian.py`) and a `server_software`
+option, rather than special-casing `ReadarrClient`.
+
+**Before building:**
+- Verify the write path against a live Bindery instance. Only `/api/queue` is
+  confirmed arr-shaped; the `POST /book` + author-ensure + `command BookSearch`
+  flow in `readarr.py:196-291` is unverified and may differ.
+- Confirm `get_queue` / `get_book_status` / `get_history` shapes on Bindery.
+
+**Already tried:** None — scoped only. Worth doing because Readarr was archived
+June 2025 (dead metadata backend) and Bindery is the main migration target.
+
+Refs: https://github.com/vavallee/bindery , docs/API.md in that repo.

app.py

@@ -30,6 +30,14 @@
 from lazylibrarian import LazyLibrarianClient
 from readarr import ReadarrClient
 
+# Open Library / Internet Archive filters requests whose User-Agent starts with
+# "python-requests" (the requests library default), returning 403 on every API
+# call. Send an identifying UA so search/discover work on a fresh install. This
+# runs at import time, before any HTTP request is issued.
+# https://github.com/zamnzim/Libreseerr/issues/10
+_USER_AGENT = "Libreseerr/1.0 (+https://github.com/zamnzim/Libreseerr)"
+http_requests.utils.default_user_agent = lambda *a, **kw: _USER_AGENT
+
 app = Flask(__name__)
 
 # Honor X-Forwarded-* headers from a reverse proxy (haproxy, nginx, traefik, etc.)
@@ -1057,22 +1065,14 @@ def get_root_folders(server_type):
 
 # ─── Download / Request API ───
 
-@app.route("/api/request", methods=["POST"])
-@login_required
-def create_request():
-    data = request.json
-    server_type = data.get("server_type")
-    book_data = data.get("book")
-    quality_profile_id = data.get("quality_profile_id")
-    root_folder = data.get("root_folder")
-
-    if not all([server_type, book_data, quality_profile_id, root_folder]):
-        return jsonify({"error": "Missing required fields"}), 400
 
-    client = get_client(server_type)
-    if not client:
-        return jsonify({"error": f"{server_type} server not configured"}), 400
+def _create_single_request(server_type, book_data, quality_profile_id, root_folder):
+    """Build and submit one request entry for a single server slot.
 
+    Returns the request_entry dict (status 'processing' on success, 'error' on
+    failure). Does NOT insert into requests_history or persist — the caller does
+    that for all entries together under `lock`.
+    """
     title = book_data.get("title", "Unknown")
     authors = book_data.get("authors", [])
     author_name = authors[0] if authors else "Unknown"
@@ -1094,52 +1094,80 @@ def create_request():
     }
 
     try:
-        # First, try to find the book in Readarr via ISBN lookup
+        client = get_client(server_type)
+        if not client:
+            raise ValueError(f"{server_type} server not configured")
+
         readarr_books = []
         if isbn:
             readarr_books = client.lookup_by_isbn(isbn)
         if not readarr_books:
             readarr_books = client.search_books(f"{title} {author_name}")
 
         if readarr_books:
-            # Use the full Readarr lookup result — it has the correct
-            # editions, images, links, etc. that Readarr expects.
-            # We only override the author if Readarr returned empty data.
             readarr_book = readarr_books[0]
             if not readarr_book.get("author", {}).get("authorName"):
-                readarr_book["author"] = {
-                    "authorName": author_name,
-                    "foreignAuthorId": "",
-                }
+                readarr_book["author"] = {"authorName": author_name, "foreignAuthorId": ""}
             app.logger.info(
-                "Readarr match for '%s': title='%s', author=%s",
+                "Backend match for '%s': title='%s', author=%s",
                 title, readarr_book.get("title"), json.dumps(readarr_book.get("author", {})),
             )
-            request_entry["status"] = "processing"
         else:
-            # Fallback: build data from Open Library
             readarr_book = {
                 "title": title,
-                "author": {
-                    "authorName": author_name,
-                    "foreignAuthorId": "",
-                },
+                "author": {"authorName": author_name, "foreignAuthorId": ""},
                 "foreignBookId": isbn or book_data.get("id", ""),
             }
-            app.logger.info("No Readarr match, using Open Library fallback for '%s' by '%s'", title, author_name)
-            request_entry["status"] = "processing"
+            app.logger.info("No backend match, using Open Library fallback for '%s' by '%s'", title, author_name)
 
+        request_entry["status"] = "processing"
         result = client.add_book(readarr_book, quality_profile_id, root_folder)
         request_entry["readarr_book_id"] = result.get("id")
     except Exception as e:
         request_entry["status"] = "error"
         request_entry["error"] = str(e)
 
+    return request_entry
+
+
+@app.route("/api/request", methods=["POST"])
+@login_required
+def create_request():
+    data = request.json or {}
+    book_data = data.get("book")
+    targets = data.get("targets")
+
+    if not book_data or not isinstance(targets, list) or not targets:
+        return jsonify({"error": "Request must include 'book' and a non-empty 'targets' list"}), 400
+
+    for t in targets:
+        if not isinstance(t, dict):
+            return jsonify({"error": "Each target must be an object"}), 400
+        st = t.get("server_type")
+        if st not in ("ebook", "audiobook"):
+            return jsonify({"error": "target server_type must be 'ebook' or 'audiobook'"}), 400
+        if not t.get("quality_profile_id") or not t.get("root_folder"):
+            return jsonify({"error": f"{st} target missing quality_profile_id or root_folder"}), 400
+        if not get_client(st):
+            return jsonify({"error": f"{st} server not configured"}), 400
+
+    entries = [
+        _create_single_request(
+            t["server_type"], book_data, t["quality_profile_id"], t["root_folder"]
+        )
+        for t in targets
+    ]
+
     with lock:
-        requests_history.insert(0, request_entry)
+        used_ids = {r["id"] for r in requests_history}
+        for entry in entries:
+            while entry["id"] in used_ids:
+                entry["id"] += 1  # avoid same-millisecond / existing-history id collisions
+            used_ids.add(entry["id"])
+            requests_history.insert(0, entry)
         save_requests()
 
-    return jsonify(request_entry)
+    return jsonify(entries)
 
 
 @app.route("/api/requests", methods=["GET"])