feat: fetch the next page of books before the reader hits the bottom

The prefetch zone is a viewport deep rather than 200px, and the observer
re-observes after a page lands so a chain of loads is not cut short.

.env.prod-example

@@ -5,6 +5,15 @@ CHITAI_TOKEN_SECRET=secret
 CHITAI_DEFAULT_LIBRARY_NAME=Books
 CHITAI_DEFAULT_LIBRARY_PATH="libraries/books"
 
+# Duplicate detection when importing files (optional).
+# Scope: "library" compares against the library being uploaded to, "global" against
+# every library, "off" disables the check.
+CHITAI_DUPLICATE_SCOPE=library
+
+# Where the consume watcher parks files it refused as duplicates. Keep it outside
+# CHITAI_CONSUME_PATH, or the watcher picks them straight back up.
+CHITAI_DUPLICATE_PATH="duplicates"
+
 # You probably should not change these
 CHITAI_API_URL="http://backend:8000"
 CHITAI_API_DEBUG=false

.gitattributes

@@ -1,2 +1,5 @@
 # Mark pdfjs as vendored code
 frontend/static/pdfjs/** linguist-vendored
+
+# Mark foliate-js as vendored code
+frontend/src/lib/vendor/** linguist-vendored

.gitignore

@@ -1,3 +1,4 @@
 .env
 .postgres/
 .venv
+tmp/

AGENTS.md

@@ -10,9 +10,11 @@ a KOSync-compatible endpoint.
 ## Layout
 
 | Path                       | What                                                                                     |
-| --- | --- |
+| -------------------------- | ---------------------------------------------------------------------------------------- |
 | `backend/`                 | Litestar REST API + PostgreSQL. See `backend/AGENTS.md`.                                 |
 | `frontend/`                | SvelteKit SSR web app. See `frontend/AGENTS.md`.                                         |
+| `frontend/src/lib/vendor/` | Vendored `foliate-js` (the EPUB engine), copied by `frontend/scripts/vendor-foliate.sh`. |
+| `frontend/static/pdfjs/`   | Vendored pdf.js viewer, used by the PDF reader in an iframe.                             |
 | `docker-compose.yml`       | Production stack: `db` (postgres:17), `backend`, `frontend`.                             |
 | `docs/screenshots/`        | Images used by `README.md`.                                                              |
 | `shell.nix`                | Root dev shell; composes the two sub-shells.                                             |
@@ -71,6 +73,9 @@ pytest tests/                                        # needs Docker (pytest-data
 ruff format src/
 alchemy --config chitai.database.config.config make-migrations
 alchemy --config chitai.database.config.config upgrade
+
+# Import a Calibre library. Copies files; --dry-run reports without writing.
+litestar --app-dir src/chitai/ calibre-import <path> --library <slug>
 ```
 
 Frontend (from `frontend/`):

TODO.md

@@ -0,0 +1,339 @@
+# TODO
+
+Known issues and deferred work. Agent-facing notes belong in the `AGENTS.md` files;
+this is for things that are broken or missing and not yet scheduled.
+
+## Backend
+
+### Identifier extraction drops most ISBNs
+
+`backend/src/chitai/services/metadata_extractor.py` — `EpubExtractor._extract_identifiers`
+
+The EPUB path validates `DC:identifier` values verbatim:
+
+```python
+for id in epub.get_metadata("DC", "identifier"):
+    if is_valid_isbn(id[0]):
+        ...
+```
+
+`is_valid_isbn` branches on `len(isbn)` being exactly 10 or 13, so anything carrying
+formatting fails. Two consequences:
+
+- **Hyphenated ISBNs are silently dropped.** `978-0-486-28211-4` is 17 characters, so it
+  never reaches the checksum. Most EPUBs write ISBNs hyphenated, so the majority are lost.
+  The PDF path already does this correctly — `_extract_isbns` calls
+  `match.replace("-", "")` before validating.
+- **`urn:isbn:` prefixes are dropped** for the same reason. This is a common EPUB form.
+
+Fix: normalise before validating — strip a leading `urn:isbn:`, then remove everything
+that isn't `0-9` or `X`. Reuse the PDF path's approach rather than duplicating it.
+
+Note this only runs at upload, so fixing it changes nothing for books already imported.
+A backfill would need to re-read the files on disk.
+
+### Non-ISBN identifiers are discarded
+
+Same function. Anything that isn't a valid ISBN is thrown away, including values EPUBs
+routinely carry: `urn:uuid:…`, `calibre:…`, Google Books volume IDs and ASINs.
+
+The `Identifier` model is already generic (`name` + `value`, unique per book), so storing
+them needs no schema change — only the extractor decides what survives. The frontend
+already renders ISBN, ASIN and DOI as links and shows unknown types as plain values, so
+anything stored will display sensibly.
+
+Worth adding at the same time:
+
+- A DOI regex (`10.\d{4,9}/\S+`) alongside the ISBN scan in `PdfExtractor._extract_isbns`
+  — academic PDFs carry one and it is the most useful identifier they have.
+- Deriving ISBN-10 from ISBN-13 when only the latter is present. It is a pure checksum
+  conversion and doubles the chance of an external lookup matching.
+
+### Any authenticated user can delete any library or book
+
+`backend/src/chitai/database/models/library.py`, `backend/src/chitai/database/models/user.py`
+
+There is no authorization tier. `Library` has no owner column, and `User` carries only
+`email` and `password` — no role, no `is_active`. So every authenticated account can create
+and delete libraries, and delete books along with their files on disk. Per-user scoping
+exists only for reading progress and bookshelves, which `provide_book_service` restricts
+correctly.
+
+For a single-household deployment that may well be acceptable. The point is that it is
+emergent rather than chosen. The cheapest meaningful step is an `is_admin` flag gating
+library deletion and `delete_books` — a model change plus a migration.
+
+### Basic auth answers a malformed header with a 500
+
+`backend/src/chitai/middleware/basic_auth.py` — line 22
+
+```python
+username, password = b64decode(auth_header.split("Basic ")[1]).decode().split(":")
+```
+
+Nothing guards the parse. A `Bearer` token raises `IndexError`, non-base64 raises
+`binascii.Error`, and a credential with no colon raises `ValueError` — as does a password
+that *contains* one, since there is no `maxsplit=1`. Every case surfaces as a 500 on an
+unauthenticated endpoint. All of them should be 401.
+
+### An unknown KOSync API key returns 404
+
+`backend/src/chitai/middleware/kosync_auth.py` — line 32
+
+`KosyncDeviceService.get_by_api_key` uses `get_one`, which raises `NotFoundError`, but the
+middleware catches only `PermissionDeniedException`. The global handler in
+`exceptions/handlers.py` then renders it as a 404, so a device presenting a bad key is told
+the route does not exist rather than that it is unauthorized. The same file still carries a
+leftover `print(exc)`.
+
+Worth doing at the same time: `KosyncDeviceService._generate_api_key` uses
+`secrets.token_hex(8)`. 64 bits is thin for a long-lived bearer credential where 32 bytes
+is the convention.
+
+### The multi-book download cannot be driven from a test
+
+`backend/src/chitai/services/book.py` — `BookService.get_files`
+
+`/books/download` is the only handler returning a Litestar `Stream`, and it cannot be
+exercised through `AsyncTestClient`. The request itself succeeds, then fixture teardown
+hangs: the test transport never sends the `http.disconnect` that the streaming response
+waits on, so the app's lifespan shutdown never completes. Coverage therefore sits at the
+service level, on `get_files` directly.
+
+Unresolved whether the endpoint also stalls behind a real ASGI server, where that
+disconnect does arrive. Worth one manual check against `litestar run` before relying on it.
+
+### The production image runs the development server
+
+`backend/Dockerfile` — the final `CMD`
+
+```
+CMD ["litestar", "--app-dir", "chitai", "run", "--host", "0.0.0.0", "--port", "8000"]
+```
+
+`litestar run` is the CLI development runner. Production should invoke uvicorn or granian
+directly, with a worker count.
+
+### Nothing gates formatting, linting or types
+
+`ruff format --check src/` reports 27 of 61 files unformatted, and `ruff check src/` finds
+114 errors — 100 of them unused imports, the rest bare `except`, unused variables and
+`== True` comparisons. `ruff check --fix` clears 51 automatically.
+
+There is no `[tool.ruff]` section in `pyproject.toml`, so only ruff's default `E4/E7/E9/F`
+rules run, and no type checker is configured at all despite `# type: ignore` comments in
+the tree. Individually these are trivial; collectively they say nothing runs on commit.
+
+## Frontend
+
+### Scripted EPUBs run against the app origin
+
+**This is a regression from the foliate-js migration, not a pre-existing gap.**
+
+The old epub.js reader never passed `allowScriptedContent`. epub.js defaults it to
+`false`, which sets `iframe.sandbox = "allow-same-origin"` — no `allow-scripts` — so
+script inside a book never ran. The vendored foliate-js sets, unconditionally:
+
+```js
+// paginator.js — and the same in fixed-layout.js
+// `allow-scripts` is needed for events because of WebKit bug
+this.#iframe.setAttribute("sandbox", "allow-same-origin allow-scripts");
+```
+
+`allow-same-origin` together with `allow-scripts` is the combination that makes the
+sandbox attribute do nothing. Sections are served as same-origin `blob:` URLs, so script
+in a book can reach `/api/*` with the session cookie attached. foliate's README says as
+much and tells you to use a CSP instead; we have not added one.
+
+This is not theoretical. Audiobookshelf shipped the same combination and got
+**CVE-2024-35236** — scripted EPUB plus an unrestricted upload gave remote code
+execution; fixed in 2.10.0 by making scripted content a per-library opt-in, off by
+default. Kavita (CVE-2024-39307) and Jellyfin (fixed 10.9.8) are variations on it.
+Write-up: <https://gebir.ge/blog/every-trick-in-the-book/>.
+
+**An app-wide CSP is the wrong shape.** `kit.csp` with `script-src: ['self']` also blocks
+`mode-watcher`'s inline `setInitialMode`, which sets the dark class before first paint —
+SvelteKit only nonces the bootstrap script it injects itself, so every page load would
+flash the light theme. Pinning a hash of a third-party inline script breaks silently on
+upgrade.
+
+**Grimmory solves it properly**, and it runs foliate-js too. Rather than handing foliate
+the whole file, it serves each EPUB entry from its own endpoint and puts the strict
+policy on that response:
+
+```java
+// EpubReaderController.java
+response.setHeader("Content-Security-Policy", "script-src 'none'");
+```
+
+The app shell keeps its own, more permissive policy. That works because each section is
+then a real same-origin document with its own header, rather than a `blob:` — and a
+`blob:` inherits the CSP of the document that created it, which is exactly why a header
+on `/api/books/download/…` would achieve nothing today.
+
+Two ways forward:
+
+1. **Cheap.** Patch the vendored `sandbox` attribute to drop `allow-scripts`, restoring
+   what epub.js gave us. Cost is the WebKit bug the upstream comment cites: events inside
+   the iframe get swallowed, which would likely break touch/swipe paging and possibly the
+   in-iframe keyboard handling in `foliate-view.svelte`. Needs testing before trusting.
+2. **Right.** Follow grimmory: serve individual EPUB entries from the backend with
+   `script-src 'none'` on each response, and drive foliate through its loader hooks
+   instead of a whole-file blob. This is **net-new capability on both sides**, not a
+   rewiring of something that exists — see below.
+
+Option 2 also fixes the memory cost below, which is why it is worth more than it looks.
+
+#### What option 2 actually involves
+
+Today the browser fetches the whole `.epub` from `download/{book_id}/{file_id}`, which
+returns a Litestar `File` and knows nothing about the archive's contents. foliate then
+opens the zip **in the browser** (`makeZipLoader` in `view.js`) and turns every chapter,
+image and stylesheet into a `blob:` URL via `Loader.createURL` in `epub.js`. A `blob:`
+carries no headers of its own — it inherits the CSP of the document that created it —
+which is why there is nowhere to attach a policy except the app shell.
+
+foliate's parser never touches the zip directly. `EPUB` is constructed with a loader:
+
+```js
+// view.js — makeZipLoader is one implementation; makeDirectoryLoader below is another
+return { entries, loadText, loadBlob, getSize };
+```
+
+`name` is the **zip entry path**, because `makeZipLoader` keys its map on
+`entry.filename` — so `OEBPS/Text/chapter01.xhtml`, `OEBPS/Images/cover.jpg`. The parser
+resolves hrefs from the OPF manifest into those names and asks the loader for them,
+without caring where the bytes come from. A third implementation that fetches over HTTP
+is the same shape.
+
+**Backend.** An endpoint taking a path inside the archive, e.g.
+`GET books/{book_id}/files/{file_id}/entry/{path:path}`, returning a `Stream` over
+`zipfile.ZipFile.open(name)` so an entry never lands in memory whole, with the content
+type from the manifest and `Content-Security-Policy: script-src 'none'` on the response.
+
+Two things to get right:
+
+- **`path` is caller-supplied.** Resolve it against the archive's `namelist()` and reject
+  anything absent, rather than trusting the string — `../` traversal is the hazard.
+- **`getSize` is synchronous** in foliate's loader contract, and it feeds `SectionProgress`,
+  which produces the reading percentage. So the endpoint needs a companion that returns
+  entry names and sizes up front — one extra call at open — because sizes cannot be
+  discovered per request.
+
+Opening the zip per request costs a central-directory read each time. Probably fine for
+chapter-sized reads, worth measuring rather than assuming.
+
+Note the backend already reads inside EPUBs — `metadata_extractor.py` uses ebooklib at
+ingest for title, authors, identifiers and the cover. What is missing is serving an
+arbitrary entry by path, not the ability to open the archive.
+
+**Frontend.** `foliate-view.svelte` stops calling `view.open(file)` and builds an `EPUB`
+around a loader backed by that endpoint. The whole-file fetch in `epub-reader.svelte`
+goes away with it.
+
+### The proxy buffers whole files and drops range headers
+
+Two separate problems that both live in `frontend/src/routes/api/[...path]/+server.ts`.
+
+**Buffering.** Litestar already streams: `ASGIFileResponse` reads in 1 MB chunks
+(`response/file.py`), so the backend never holds a file whole. The proxy then undoes it
+with `await response.arrayBuffer()`, which does not resolve until the last byte arrives —
+so the whole file sits in the node process, per concurrent reader, and the browser gets
+nothing until it completes. Passing `response.body` straight through restores the stream
+and is a small change.
+
+This is now **responses only**. The request side was fixed for the Calibre archive upload,
+which cannot be held in memory: POST and PATCH pass `request.body` through with
+`duplex: 'half'` (`bodyOf` in the same file). The response side is the same shape of fix.
+
+**Range.** The proxy forwards only `Content-Type`, `Content-Disposition` and
+`Content-Length`. It never sends the client's `Range` upstream, and would drop
+`Accept-Ranges` and `Content-Range` coming back — a 206 without `Content-Range` is
+broken. So range support cannot work until the proxy is fixed, whatever the backend does.
+
+**Litestar has no range support of its own.** In 2.21.1 the only mention of 206 in the
+whole package is the `HTTP_206_PARTIAL_CONTENT` constant; there is no `Accept-Ranges` or
+`Content-Range` handling anywhere. This has to be written.
+
+#### What pdf.js actually needs
+
+It decides from the **initial 200 response**, not from anything on a 206.
+`validateRangeRequestCapabilities` in `frontend/static/pdfjs/build/pdf.mjs`:
+
+```js
+if (responseHeaders.get("Accept-Ranges") !== "bytes") {
+  return returnValues; // allowRangeRequests stays false
+}
+```
+
+It also needs a parseable `Content-Length`, `Content-Encoding: identity`, and a length
+greater than twice `rangeChunkSize`. Miss any of those and it downloads the whole file
+however good the range support is.
+
+So the single highest-value header is **`Accept-Ranges: bytes` on the ordinary 200** —
+that is what makes pdf.js switch to fetching progressively at all.
+
+#### Approach
+
+Put it on the existing `get_file` handler in `controllers/book.py`, which already resolves
+`book_id`/`file_id` through the service with library scoping and auth:
+
+- No `Range` → `Stream` the file with `Accept-Ranges: bytes` and `Content-Length`.
+- `Range` present → parse, seek, `Stream` with 206 and `Content-Range`.
+- Proxy: forward `Range` up; pass `response.body` through; forward `Accept-Ranges`,
+  `Content-Range` and the status back.
+
+There are `RangeRequestMiddleware` snippets circulating for Litestar that wrap
+`create_static_files_router`. They are the wrong shape here — book files are served by an
+authenticated handler resolving database ids, not by a directory mapping, and using one
+would mean exposing disk paths as URLs and re-solving ownership checks that already
+exist. The common version also only sets `Accept-Ranges` on the 206, so it would not
+switch pdf.js over, and it derives its path with `str.lstrip(prefix)`, which strips a
+character set rather than a prefix — `/static/castle.pdf` becomes `le.pdf`. Worth reading
+its `parse_range_header` for the parsing rules and writing the rest fresh.
+
+### Remove the epub.js locations-cache purge
+
+`frontend/src/lib/reader/legacy-cache.ts` — `purgeLegacyLocationCache`
+
+The epub.js reader cached generated locations in `localStorage` under
+`${bookId}-locations`, a few hundred KB of JSON per long book. foliate computes
+progress from section byte sizes at open time, so nothing writes those keys any
+more, but existing browsers still hold them — and a reader near the 5–10 MB
+origin quota would make the new reader-settings write throw `QuotaExceededError`.
+
+The reader clears them once per browser, behind a `chitai:locations-purged` flag.
+Delete the module, its call in `epub-reader.svelte` and the flag once deployments
+have had a release or two to run it — after roughly 2026-12.
+
+### Cover dimensions are unknown until load
+
+`book-cover.svelte` renders covers at a fixed height with natural width so nothing is
+cropped or distorted. Because the intrinsic size is not known ahead of time, the text
+beside the cover settles once when the image loads. `min-width` bounds the movement but
+does not remove it.
+
+Removing it properly means storing cover dimensions at ingest and emitting them as
+`width`/`height` attributes so the browser reserves exact space. That is a model change
+plus a migration.
+
+### No series navigation
+
+`Book` has `series` and `series_position`, and the detail page shows both, but there is no
+way to reach the other volumes. The books list endpoint filters by author, publisher, tag,
+shelf and progress — a `series` filter would need adding in
+`backend/src/chitai/services/filters/book.py` and wiring through
+`services/dependencies.py`, following the existing `AuthorFilter` pattern.
+
+### Book pages are sparse for most books
+
+EPUB metadata is thin: most imports arrive with a title, an author and nothing else. Two
+independent directions, neither started:
+
+- **Use what exists.** "More by this author" (the list endpoint already accepts
+  `authors=`), and exposing `Book.created_at` and `BookProgress.updated_at` — both are in
+  the database, neither is on `BookRead` / `BookProgressRead`.
+- **Fetch from outside.** Open Library or Google Books lookup by ISBN for descriptions and
+  covers. This is what actually fixes the sparseness, but it needs outbound requests, rate
+  limiting, a manual-vs-automatic decision, and a rule for not clobbering hand-edited
+  metadata.

backend/AGENTS.md

@@ -88,6 +88,14 @@ The backend owns files on disk, not just rows:
 - **Layout** — `services/filesystem_library.py` (`BookPathGenerator`) renders a Jinja2 template
   against book metadata to decide where a book lives under the library's `root_path`
   (default: `author/series/position - title/`).
+- **One directory per book, never shared.** The generated path is a pure function of the metadata,
+  so two books with the same author and title produce the same one — two editions, or an
+  `allow_duplicates` copy. `BookService._reserve_book_path` moves the later one to `title (2)`
+  before anything is written, and `update_book` reserves the same way so a rename cannot move a
+  book in on top of another. This matters because `book.path` is what deletes, moves and file
+  lookups act on: books sharing a directory means one overwrites the other's files, and deleting
+  either takes both. `_unused_path` does the same job for filenames within a directory.
+  A book that already has a `path` keeps it — `add_files` must follow the book, not the template.
 - **Metadata extraction** — `services/metadata_extractor.py` reads EPUB (ebooklib) and PDF
   (pypdfium2) files; extracted values fill only *empty* fields on the incoming payload.
 - **Covers** — converted to WebP with a UUID filename under `settings.book_cover_path`, served by a
@@ -100,6 +108,235 @@ The backend owns files on disk, not just rows:
   if it differs, moves the directory contents and prunes empty parents. Keep that in mind before
   changing metadata handling.
 
+### Content types come from the extension, and null means null
+
+Every ingest path names a file's format with `guess_content_type` (`services/utils.py`), never with
+`mimetypes.guess_type` directly and never with what the client said. Python's built-in map answers
+`None` for `.mobi`, `.azw`, `.prc`, `.fb2`, `.fbz`, `.lit`, `.lrf` and `.cb7` — most of what a
+library imported from elsewhere carries — so `EBOOK_CONTENT_TYPES` fills those in. A browser's
+`application/octet-stream` is discarded rather than used as a fallback: it is the client saying it
+does not know, and storing it is indistinguishable from having determined a format.
+
+When nothing can name the extension the column **stays null**. That is the honest answer, and only
+one consumer cannot take it: OPDS `Link.type` is a required string, so
+`services/opds/opds.py` substitutes `application/octet-stream` at that boundary. Litestar's
+`ASGIFileResponse` already does its own fallback, so `get_file` can pass a null straight through.
+`FileMetadataRead.content_type` is nullable for the same reason — it was once required, which
+turned a stored null into a 500 on a book that was otherwise fine.
+
+## Duplicate detection
+
+Every ingest path screens incoming files against what is already stored, keyed on
+**`(hash, size)`** — never the hash alone, because it samples 12 KiB (see below) and
+EPUBs from one toolchain often share their first window. `FileMetadata.hash` carries a
+plain, deliberately **non-unique** index: a collision must not be able to fail an import,
+and older databases may already hold duplicates.
+
+Scope comes from `CHITAI_DUPLICATE_SCOPE` (`library`, the default | `global` | `off`).
+
+The policy differs by how deliberate the import is:
+
+| Path | Behaviour |
+| --- | --- |
+| `create_many_from_files` (browser bulk) | Skip per file, skip a whole group whose files are all known, report everything skipped in `ImportResult.duplicates`. Re-dropping a folder to pick up what is new is the case this serves. |
+| `create_book` (single, with metadata) | All-or-nothing: raises `DuplicateFilesError`, which `controllers/book.py` renders as a **409** carrying the refused files in `extra`. |
+| `add_files` | A file the book already carries is a no-op; one stored under another book raises `DuplicateFilesError`. |
+| `create_many_from_existing_files` (consume watcher) | Skips, and **moves the file to `CHITAI_DUPLICATE_PATH/<library slug>/`** — nothing is deleted, and it cannot stay put because `watchfiles` only reports additions. That path must stay outside `consume_path` or the watcher re-imports it and tries to read the directory name as a library slug. |
+| `create_many_from_calibre` (Calibre import) | Skips per file, and skips a whole book whose files are all known. Nothing is moved — the source is somebody else's library. This is what makes a re-run a no-op and an interrupted import resumable by running it again. |
+
+`allow_duplicates=true` overrides all of it, on every endpoint. Keep that working — the
+hash is not proof of identity, so a wrong verdict has to be recoverable, and a scripted
+import needs a way through. The **web UI deliberately does not offer it**: storing the
+same bytes twice splits reading progress and shelf membership across two records that
+can never converge, which is nothing anyone wants on purpose.
+
+Three things to preserve when touching this code:
+
+- **A match only counts while the file is on disk.** `find_duplicate_files` stats each
+  candidate, and `add_files` writes a missing file back into the row that already
+  describes it (`_restore_file`) instead of adding a second row beside it. The hash
+  lives in the database and the file does not, so without this a file deleted behind
+  the app's back would go on refusing its own replacement.
+
+- **Screening runs before anything is written.** `fingerprint_upload` reads the spooled
+  upload and rewinds it; the resulting fingerprints are handed to `_save_book_files`,
+  which skips its own `StreamingHasher` when it already has the answer. Passing them
+  through is what keeps the file from being read twice.
+- **`_screen_for_duplicates` extends the `known` dict as it goes**, so the same bytes
+  submitted twice in one request are caught. Those duplicates report `book_id: None` —
+  there is no row to point at yet.
+
+`POST /books/duplicate-files` answers the same question from fingerprints alone, for
+clients that want to ask before uploading anything.
+
+### Book-level detection — a different question
+
+The file check answers "are these the same bytes?". `find_duplicate_books` answers "is
+this the same book?", which a re-scan, a re-zipped EPUB or another edition cannot be
+asked with a hash. Two signals, either sufficient: a **shared identifier**, or a
+**matching normalized title with at least one shared author**.
+
+It **never blocks**. A metadata match is a guess — a work shares title and author with
+its own translation, its own second edition and its own audiobook — so the book is
+created and the candidates are reported alongside it in
+`ImportResult.possible_duplicates`. File-level dedupe keeps its refuse/skip behaviour;
+that one is near-certain and this one is not. Do not "improve" this into a refusal.
+
+Two rules narrow it, both applied in Python over the small candidate set:
+
+- **A shared author is required for a title match.** Without it every book the
+  extractors gave up on and titled `Unknown` is a duplicate of every other one. A book
+  with no authors can therefore only match on an identifier.
+- **The same series at a different `series_position` disqualifies a match.** A trilogy
+  shares an author and often most of its title; the position is the library saying
+  outright that these are two books.
+
+A book is compared under **several title keys, not one** (`_title_keys`). Ebook files
+are overwhelmingly named `Title - Author.epub`, and wherever nothing inside the file
+overrode that name the author ended up in the title column — so one copy is stored as
+`Building Microservices` and another as `Building Microservices Sam Newman`. Both
+directions are generated, the author stripped off and the author added on, which is why
+the query is `normalized_title.in_(keys)` rather than `==`. This is still exact matching
+on an indexed column: no similarity score, nothing to tune. It does not weaken the
+shared-author requirement, which is a separate condition.
+
+`find_duplicate_book_groups` is the library-wide pass behind
+`GET /books/duplicate-books`, since the import-time check says nothing about a
+collection someone already has. It buckets books by every key they carry and merges the
+buckets with union-find, so A~B by ISBN and B~C by title land in one group. Pairs in
+`duplicate_dismissals` are never merged — a reader disagreeing with one pairing must not
+silently break a group that stands on other evidence.
+
+### Author names have one stored form
+
+`Author.name` is always the canonical form, produced by `format_author_name`. Extractors
+hand over whatever the file said — `Newman, Sam;` from a `DC:creator` list, `Sam Newman`
+from a PDF, `Sam Newman.epub` from a filename — and storing those verbatim is how one
+person becomes four rows in the sidebar, four entries in the author filter, and four
+books that never look like each other.
+
+This is **display** canonicalization, distinct from `normalize_author`, which throws
+away case, accents and spacing to build a comparison key nobody sees. Tidying only
+removes what an extractor added: a trailing separator, a file extension, and the
+`Surname, Given` ordering. It never touches case or accents — `Michał Płachta` and
+`Steve McConnell` are the author's own spelling, not something to correct.
+
+Three places have to agree, and `Author` keeps them together: the `@validates("name")`
+hook, `unique_hash`, and `unique_filter`. `as_unique_async` looks a row up with the
+filter and then constructs with the validator, so if the lookup used the raw name and
+the insert used the tidy one, every variant spelling would miss the existing row and
+then collide with it on the unique index.
+
+A form the rule does not recognise is **left exactly as it was found** — `Dave Thomas,
+Andy Hunt` is two people in one string, and flipping it would invent a third. Leaving a
+mess visible beats rewriting it wrongly.
+
+### The normalized columns are written by validators
+
+`Book.normalized_title`, `Author.normalized_name` and `Identifier.normalized_value` are
+derived from `services/matching.py` and kept current by **SQLAlchemy `@validates` hooks
+on the models**, not by any service. Assigning them directly is always wrong.
+
+This is deliberate and it is invisible at the call sites: `BookService` writes titles
+through at least three paths (`to_model_on_create`, `to_model_on_update`, and the
+`setattr` loop in `_populate_with_unique_relationships`), and a validator is the only
+thing a fourth cannot bypass. The columns carry plain, **non-unique** btree indexes —
+two spellings collapsing onto one value is the entire point.
+
+`services/matching.py` is imported *inside* those validators rather than at module
+scope: reaching it initialises the `chitai.services` package, which imports the
+services, which import the models. Keep the local import.
+
+The validators only fire on write, so a migration that adds one of these columns must
+backfill existing rows through the same helpers — see the `data_upgrades()` hook in
+`2026-08-15_add_book_matching_keys_and_duplicate__4358e7d4743a.py`.
+
+**Changing anything in `services/matching.py` needs a revision that recomputes them.**
+The keys are derived and already written, so a normalization change silently invalidates
+every stored row: a book written under the old rules just stops matching one written
+under the new rules, with nothing to show that anything is wrong. Copy
+`2026-08-15_recompute_book_matching_keys_ed41acf21270.py`, which exists because
+`normalize_title` learned to strip compact edition markers (`2E`, `5e`). It is
+idempotent and safe to re-run.
+
+## Importing from Calibre
+
+Two pieces, deliberately separated:
+
+- **`services/calibre.py`** reads `metadata.db` and the tree beside it. It knows nothing about
+  `Book`, `BookService` or a session, so it is testable without Postgres, and it reports what
+  Calibre wrote rather than what Chitai wants — identifiers come back keyed by `identifiers.type`
+  verbatim. It also unpacks a zipped library (`extract_calibre_archive`).
+- **`BookService.create_many_from_calibre`** does the ingest, next to the two other ingest paths
+  because it needs the same privates they do (`_reserve_book_path`, `_save_cover_image`,
+  `_screen_for_duplicates`). The CLI in `cli.py` is a thin wrapper over it.
+- **`services/calibre_import.py`** is lifecycle only — the job registry behind the endpoints:
+  state, progress, cancellation, and a session of its own.
+
+`docs/calibre-import.md` is the full brief, including the phases not built yet. What matters here:
+
+- **Files are copied, never moved.** `metadata.db` would go on pointing at files that are gone,
+  which quietly ruins a library somebody still uses. `copy_file` streams rather than using
+  `shutil.copy`, which would block the loop for a 40 MB read.
+- **The extractors are not run.** This is the one ingest path that trusts its input: Calibre's
+  catalogue is curated and its filenames are truncated to ~42 characters, so `data.name` locates a
+  file and the database carries the metadata. The title is stored verbatim for the same reason —
+  no edition split out, no subtitle guessed.
+- **The catalogue is copied before it is read**, and the copy is opened read-write. Calibre may be
+  running; opening the live file either sees a torn state or needs to recover a write-ahead log,
+  which read-only access cannot do. `close()` removes the copy in a `finally`, or a failure leaves
+  a catalogue-sized file in the temp directory.
+- **`check_same_thread=False` plus an `asyncio.Lock`.** Every query runs through
+  `asyncio.to_thread`, which hands out whichever worker is free, so the connection outlives the
+  thread that opened it. The lock is what makes that safe. Removing either one reintroduces
+  `SQLite objects created in a thread can only be used in that same thread`, intermittently —
+  the pool often reuses one thread, so it passes until it does not.
+- **One book never costs the run.** A failure is recorded in `CalibreImportResult.failed`, the
+  session is rolled back so the next book can use it, and the files that book had already copied
+  are deleted — an orphaned directory would make the next attempt reserve `title (2)` and look as
+  though it had worked. A cover that PIL cannot open costs the cover, not the book.
+- **`calibre-uuid`, not `uuid`.** `books.uuid` is stable for the life of the row, so it is the
+  durable link back to the source and worth matching on. `uuid` is the name `normalize_identifier`
+  refuses, because an EPUB regenerates one per build.
+
+### The API takes an uploaded archive; the CLI takes a path
+
+**`POST /libraries/{id}/imports/calibre/upload`** is the only way in over HTTP. It takes a zipped
+Calibre library, answers **202** with a job handle, and unpacks into a temp directory the job owns.
+`GET /libraries/imports/{job_id}` is polled; `DELETE` on the same path stops it.
+
+There is deliberately **no endpoint that imports from a server path**. A desktop Calibre install is
+not on the server, and importing from a path the server can already see is a server-side operation
+— which is what `litestar --app-dir src/chitai/ calibre-import <path> --library <slug>` is for,
+including its `--dry-run`. Do not add the path endpoint back without being asked: it was built,
+then removed on purpose.
+
+- **The registry is in memory, so it assumes one worker process.** That holds today (`litestar run`
+  is single-process, and the consume watcher is already an in-process singleton), but the day
+  `TODO.md`'s "production image runs the development server" item is fixed with a worker count, a
+  poll can land on a worker that never heard of the job. `services/calibre_import.py` says so at
+  the top; an `import_jobs` table is the answer when that happens.
+- **The job opens its own session.** The request that started it is long gone and its session
+  closed with it.
+- **Cancelling is not aborting.** A flag is read between books, never during one, so a cancelled
+  import leaves whole books behind and never half of one. `task.cancel()` would abandon a book
+  mid-copy and leave files with no row describing them.
+- **The job deletes its workspace** — the unpacked archive is a second copy of the whole library,
+  and the books worth keeping have been copied into the library proper by the time it ends. Removed
+  even when the run failed, since nothing will come back for it.
+- **Extraction refuses** an entry pointing outside the archive (zip slip), an archive that will not
+  fit on disk, and one with no `metadata.db` within three levels. All three answer 400 before a job
+  exists, rather than as a job that reports FAILED a moment later.
+- **The upload is streamed both sides.** The archive reaches disk in chunks rather than being read
+  whole, and the SvelteKit proxy passes `request.body` through instead of buffering it — see
+  `frontend/AGENTS.md`.
+
+Things Calibre does that will produce wrong data if you forget them are documented at the top of
+`services/calibre.py` — the `0101-01-01` date sentinel, `|` for a comma in an author name, the
+REAL `series_index` that defaults to 1.0 for every book, HTML in `comments`, the views that need
+SQLite functions Calibre registers from Python, and `books_pages_link` being both recent and
+usually empty. `tests/calibre_fixtures.py` builds a library exercising all of them.
+
 ## KOReader hashing
 
 `services/utils.py` reimplements KOReader's partial-MD5 document identifier: 1 KiB samples at

backend/migrations/versions/2026-08-13_add_file_hash_index_e9c2c7e875ae.py

@@ -0,0 +1,80 @@
+"""add file hash index
+
+Revision ID: e9c2c7e875ae
+Revises: 6d72d1bbc0ee
+Create Date: 2026-08-13 14:52:09.341906
+
+"""
+
+import warnings
+from typing import TYPE_CHECKING
+
+import sqlalchemy as sa
+from alembic import op
+from advanced_alchemy.types import EncryptedString, EncryptedText, GUID, ORA_JSONB, DateTimeUTC, StoredObject, PasswordHash, FernetBackend
+from advanced_alchemy.types.encrypted_string import PGCryptoBackend
+from advanced_alchemy.types.password_hash.argon2 import Argon2Hasher
+from advanced_alchemy.types.password_hash.passlib import PasslibHasher
+from advanced_alchemy.types.password_hash.pwdlib import PwdlibHasher
+from sqlalchemy import Text  # noqa: F401
+
+if TYPE_CHECKING:
+    from collections.abc import Sequence
+
+__all__ = ["downgrade", "upgrade", "schema_upgrades", "schema_downgrades", "data_upgrades", "data_downgrades"]
+
+sa.GUID = GUID
+sa.DateTimeUTC = DateTimeUTC
+sa.ORA_JSONB = ORA_JSONB
+sa.EncryptedString = EncryptedString
+sa.EncryptedText = EncryptedText
+sa.StoredObject = StoredObject
+sa.PasswordHash = PasswordHash
+sa.Argon2Hasher = Argon2Hasher
+sa.PasslibHasher = PasslibHasher
+sa.PwdlibHasher = PwdlibHasher
+sa.FernetBackend = FernetBackend
+sa.PGCryptoBackend = PGCryptoBackend
+
+# revision identifiers, used by Alembic.
+revision = 'e9c2c7e875ae'
+down_revision = '6d72d1bbc0ee'
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=UserWarning)
+        with op.get_context().autocommit_block():
+            schema_upgrades()
+            data_upgrades()
+
+def downgrade() -> None:
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=UserWarning)
+        with op.get_context().autocommit_block():
+            data_downgrades()
+            schema_downgrades()
+
+def schema_upgrades() -> None:
+    """schema upgrade migrations go here."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table('file_metadata', schema=None) as batch_op:
+        batch_op.create_index('ix_file_metadata_hash', ['hash'], unique=False)
+
+    # ### end Alembic commands ###
+
+def schema_downgrades() -> None:
+    """schema downgrade migrations go here."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table('file_metadata', schema=None) as batch_op:
+        batch_op.drop_index('ix_file_metadata_hash')
+
+    # ### end Alembic commands ###
+
+def data_upgrades() -> None:
+    """Add any optional data upgrade migrations here!"""
+
+def data_downgrades() -> None:
+    """Add any optional data downgrade migrations here!"""

backend/migrations/versions/2026-08-15_add_book_matching_keys_and_duplicate__4358e7d4743a.py

@@ -0,0 +1,163 @@
+"""add book matching keys and duplicate dismissals
+
+Revision ID: 4358e7d4743a
+Revises: e9c2c7e875ae
+Create Date: 2026-08-15 15:09:02.708914
+
+"""
+
+import warnings
+
+import sqlalchemy as sa
+from alembic import op
+from advanced_alchemy.types import EncryptedString, EncryptedText, GUID, ORA_JSONB, DateTimeUTC, StoredObject, PasswordHash, FernetBackend
+from advanced_alchemy.types.encrypted_string import PGCryptoBackend
+from advanced_alchemy.types.password_hash.argon2 import Argon2Hasher
+from advanced_alchemy.types.password_hash.passlib import PasslibHasher
+from advanced_alchemy.types.password_hash.pwdlib import PwdlibHasher
+from sqlalchemy import Text  # noqa: F401
+
+__all__ = ["downgrade", "upgrade", "schema_upgrades", "schema_downgrades", "data_upgrades", "data_downgrades"]
+
+sa.GUID = GUID
+sa.DateTimeUTC = DateTimeUTC
+sa.ORA_JSONB = ORA_JSONB
+sa.EncryptedString = EncryptedString
+sa.EncryptedText = EncryptedText
+sa.StoredObject = StoredObject
+sa.PasswordHash = PasswordHash
+sa.Argon2Hasher = Argon2Hasher
+sa.PasslibHasher = PasslibHasher
+sa.PwdlibHasher = PwdlibHasher
+sa.FernetBackend = FernetBackend
+sa.PGCryptoBackend = PGCryptoBackend
+
+# revision identifiers, used by Alembic.
+revision = '4358e7d4743a'
+down_revision = 'e9c2c7e875ae'
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=UserWarning)
+        with op.get_context().autocommit_block():
+            schema_upgrades()
+            data_upgrades()
+
+def downgrade() -> None:
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=UserWarning)
+        with op.get_context().autocommit_block():
+            data_downgrades()
+            schema_downgrades()
+
+def schema_upgrades() -> None:
+    """schema upgrade migrations go here."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.create_table('duplicate_dismissals',
+    sa.Column('id', sa.BigInteger().with_variant(sa.Integer(), 'sqlite'), nullable=False),
+    sa.Column('book_a_id', sa.BigInteger().with_variant(sa.Integer(), 'sqlite'), nullable=False),
+    sa.Column('book_b_id', sa.BigInteger().with_variant(sa.Integer(), 'sqlite'), nullable=False),
+    sa.ForeignKeyConstraint(['book_a_id'], ['books.id'], name=op.f('fk_duplicate_dismissals_book_a_id_books'), ondelete='cascade'),
+    sa.ForeignKeyConstraint(['book_b_id'], ['books.id'], name=op.f('fk_duplicate_dismissals_book_b_id_books'), ondelete='cascade'),
+    sa.PrimaryKeyConstraint('id', name=op.f('pk_duplicate_dismissals')),
+    sa.UniqueConstraint('book_a_id', 'book_b_id', name=op.f('uq_duplicate_dismissals_book_a_id'))
+    )
+    with op.batch_alter_table('duplicate_dismissals', schema=None) as batch_op:
+        batch_op.create_index(batch_op.f('ix_duplicate_dismissals_book_a_id'), ['book_a_id'], unique=False)
+        batch_op.create_index(batch_op.f('ix_duplicate_dismissals_book_b_id'), ['book_b_id'], unique=False)
+
+    # `server_default` so the column can be added to a table that already has rows;
+    # `data_upgrades` fills in the real keys immediately afterwards.
+    with op.batch_alter_table('authors', schema=None) as batch_op:
+        batch_op.add_column(sa.Column('normalized_name', sa.String(), nullable=False, server_default=''))
+        batch_op.create_index(batch_op.f('ix_authors_normalized_name'), ['normalized_name'], unique=False)
+
+    with op.batch_alter_table('books', schema=None) as batch_op:
+        batch_op.add_column(sa.Column('normalized_title', sa.String(), nullable=False, server_default=''))
+        batch_op.create_index(batch_op.f('ix_books_normalized_title'), ['normalized_title'], unique=False)
+
+    with op.batch_alter_table('identifiers', schema=None) as batch_op:
+        batch_op.add_column(sa.Column('normalized_value', sa.String(), nullable=True))
+        batch_op.create_index(batch_op.f('ix_identifiers_normalized_value'), ['normalized_value'], unique=False)
+
+    # ### end Alembic commands ###
+
+def schema_downgrades() -> None:
+    """schema downgrade migrations go here."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table('identifiers', schema=None) as batch_op:
+        batch_op.drop_index(batch_op.f('ix_identifiers_normalized_value'))
+        batch_op.drop_column('normalized_value')
+
+    with op.batch_alter_table('books', schema=None) as batch_op:
+        batch_op.drop_index(batch_op.f('ix_books_normalized_title'))
+        batch_op.drop_column('normalized_title')
+
+    with op.batch_alter_table('authors', schema=None) as batch_op:
+        batch_op.drop_index(batch_op.f('ix_authors_normalized_name'))
+        batch_op.drop_column('normalized_name')
+
+    with op.batch_alter_table('duplicate_dismissals', schema=None) as batch_op:
+        batch_op.drop_index(batch_op.f('ix_duplicate_dismissals_book_b_id'))
+        batch_op.drop_index(batch_op.f('ix_duplicate_dismissals_book_a_id'))
+
+    op.drop_table('duplicate_dismissals')
+    # ### end Alembic commands ###
+
+def data_upgrades() -> None:
+    """
+    Fill the matching keys in for rows that already exist.
+
+    The validators on the models only fire when something is written, so without this
+    every book imported before today is invisible to duplicate detection. Run through
+    the same helpers the validators use, so a backfilled row and a freshly written one
+    are guaranteed to agree.
+    """
+    from chitai.services.matching import (
+        normalize_author,
+        normalize_identifier,
+        normalize_title,
+    )
+
+    connection = op.get_bind()
+
+    books = connection.execute(sa.text("SELECT id, title FROM books")).fetchall()
+    _apply(
+        connection,
+        "UPDATE books SET normalized_title = :key WHERE id = :id",
+        [{"id": id, "key": normalize_title(title)} for id, title in books],
+    )
+
+    authors = connection.execute(sa.text("SELECT id, name FROM authors")).fetchall()
+    _apply(
+        connection,
+        "UPDATE authors SET normalized_name = :key WHERE id = :id",
+        [{"id": id, "key": normalize_author(name)} for id, name in authors],
+    )
+
+    identifiers = connection.execute(
+        sa.text("SELECT id, name, value FROM identifiers")
+    ).fetchall()
+    _apply(
+        connection,
+        "UPDATE identifiers SET normalized_value = :key WHERE id = :id",
+        [
+            {"id": id, "key": normalize_identifier(name, value)}
+            for id, name, value in identifiers
+        ],
+    )
+
+
+def _apply(connection, statement: str, parameters: list[dict]) -> None:
+    """Run one update per row, in batches, skipping the work when there are none."""
+    batch_size = 1000
+
+    for start in range(0, len(parameters), batch_size):
+        connection.execute(sa.text(statement), parameters[start : start + batch_size])
+
+
+def data_downgrades() -> None:
+    """Add any optional data downgrade migrations here!"""

backend/migrations/versions/2026-08-15_canonicalize_author_names_49a9e85a0ffc.py

@@ -0,0 +1,130 @@
+"""canonicalize author names
+
+Revision ID: 49a9e85a0ffc
+Revises: ed41acf21270
+Create Date: 2026-08-15 15:59:47.331545
+
+"""
+
+import warnings
+
+import sqlalchemy as sa
+from alembic import op
+from advanced_alchemy.types import EncryptedString, EncryptedText, GUID, ORA_JSONB, DateTimeUTC, StoredObject, PasswordHash, FernetBackend
+from advanced_alchemy.types.encrypted_string import PGCryptoBackend
+from advanced_alchemy.types.password_hash.argon2 import Argon2Hasher
+from advanced_alchemy.types.password_hash.passlib import PasslibHasher
+from advanced_alchemy.types.password_hash.pwdlib import PwdlibHasher
+from sqlalchemy import Text  # noqa: F401
+
+__all__ = ["downgrade", "upgrade", "schema_upgrades", "schema_downgrades", "data_upgrades", "data_downgrades"]
+
+sa.GUID = GUID
+sa.DateTimeUTC = DateTimeUTC
+sa.ORA_JSONB = ORA_JSONB
+sa.EncryptedString = EncryptedString
+sa.EncryptedText = EncryptedText
+sa.StoredObject = StoredObject
+sa.PasswordHash = PasswordHash
+sa.Argon2Hasher = Argon2Hasher
+sa.PasslibHasher = PasslibHasher
+sa.PwdlibHasher = PwdlibHasher
+sa.FernetBackend = FernetBackend
+sa.PGCryptoBackend = PGCryptoBackend
+
+# revision identifiers, used by Alembic.
+revision = '49a9e85a0ffc'
+down_revision = 'ed41acf21270'
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=UserWarning)
+        with op.get_context().autocommit_block():
+            schema_upgrades()
+            data_upgrades()
+
+def downgrade() -> None:
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=UserWarning)
+        with op.get_context().autocommit_block():
+            data_downgrades()
+            schema_downgrades()
+
+def schema_upgrades() -> None:
+    """schema upgrade migrations go here."""
+    pass
+
+def schema_downgrades() -> None:
+    """schema downgrade migrations go here."""
+    pass
+
+def data_upgrades() -> None:
+    """
+    Rewrite every author into the canonical form, merging the rows that collide.
+
+    `Author.name` is only now guaranteed tidy — until this revision extractors wrote
+    whatever the file said, so one person could hold several rows: "Sam Newman" beside
+    "Newman, Sam;" beside "Sam Newman.epub" (the last from a filename whose extension
+    was never stripped). Each showed up as its own author in the sidebar and its own
+    filter, and no amount of fixing the extractors repairs a row already written.
+
+    Rows that canonicalize onto one name are merged into the lowest id, which keeps
+    whichever row the library has been referring to longest. `book.path` is stored, not
+    derived, so renaming an author moves nothing on disk.
+    """
+    from chitai.services.matching import format_author_name, normalize_author
+
+    connection = op.get_bind()
+    authors = connection.execute(sa.text("SELECT id, name FROM authors")).fetchall()
+
+    groups: dict[str, list[int]] = {}
+    for id, name in sorted(authors):
+        # A name with nothing left of it after tidying is left exactly as it was:
+        # merging those together would invent one author out of several unrelated
+        # broken rows, which is worse than leaving the mess visible.
+        if canonical := format_author_name(name):
+            groups.setdefault(canonical, []).append(id)
+
+    for canonical, ids in groups.items():
+        winner, losers = ids[0], ids[1:]
+
+        for loser in losers:
+            # A book credited to both rows would otherwise breach the
+            # (book_id, author_id) unique constraint the moment the link is repointed.
+            connection.execute(
+                sa.text(
+                    "DELETE FROM book_author_links WHERE author_id = :loser AND book_id IN"
+                    " (SELECT book_id FROM book_author_links WHERE author_id = :winner)"
+                ),
+                {"loser": loser, "winner": winner},
+            )
+            connection.execute(
+                sa.text(
+                    "UPDATE book_author_links SET author_id = :winner"
+                    " WHERE author_id = :loser"
+                ),
+                {"loser": loser, "winner": winner},
+            )
+            connection.execute(
+                sa.text("DELETE FROM authors WHERE id = :loser"), {"loser": loser}
+            )
+
+        # Only after the losers are gone, or this collides with the unique index.
+        connection.execute(
+            sa.text(
+                "UPDATE authors SET name = :name, normalized_name = :key WHERE id = :id"
+            ),
+            {"id": winner, "name": canonical, "key": normalize_author(canonical)},
+        )
+
+
+def data_downgrades() -> None:
+    """
+    Nothing to undo.
+
+    The rows a merge removed are gone, and the spellings it replaced were never
+    recorded anywhere else — there is nothing to restore them from.
+    """

backend/migrations/versions/2026-08-15_recompute_book_matching_keys_ed41acf21270.py

@@ -0,0 +1,122 @@
+"""recompute book matching keys
+
+Revision ID: ed41acf21270
+Revises: 4358e7d4743a
+Create Date: 2026-08-15 15:44:28.341020
+
+"""
+
+import warnings
+
+import sqlalchemy as sa
+from alembic import op
+from advanced_alchemy.types import EncryptedString, EncryptedText, GUID, ORA_JSONB, DateTimeUTC, StoredObject, PasswordHash, FernetBackend
+from advanced_alchemy.types.encrypted_string import PGCryptoBackend
+from advanced_alchemy.types.password_hash.argon2 import Argon2Hasher
+from advanced_alchemy.types.password_hash.passlib import PasslibHasher
+from advanced_alchemy.types.password_hash.pwdlib import PwdlibHasher
+from sqlalchemy import Text  # noqa: F401
+
+__all__ = ["downgrade", "upgrade", "schema_upgrades", "schema_downgrades", "data_upgrades", "data_downgrades"]
+
+sa.GUID = GUID
+sa.DateTimeUTC = DateTimeUTC
+sa.ORA_JSONB = ORA_JSONB
+sa.EncryptedString = EncryptedString
+sa.EncryptedText = EncryptedText
+sa.StoredObject = StoredObject
+sa.PasswordHash = PasswordHash
+sa.Argon2Hasher = Argon2Hasher
+sa.PasslibHasher = PasslibHasher
+sa.PwdlibHasher = PwdlibHasher
+sa.FernetBackend = FernetBackend
+sa.PGCryptoBackend = PGCryptoBackend
+
+# revision identifiers, used by Alembic.
+revision = 'ed41acf21270'
+down_revision = '4358e7d4743a'
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=UserWarning)
+        with op.get_context().autocommit_block():
+            schema_upgrades()
+            data_upgrades()
+
+def downgrade() -> None:
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=UserWarning)
+        with op.get_context().autocommit_block():
+            data_downgrades()
+            schema_downgrades()
+
+def schema_upgrades() -> None:
+    """schema upgrade migrations go here."""
+    pass
+
+def schema_downgrades() -> None:
+    """schema downgrade migrations go here."""
+    pass
+
+def data_upgrades() -> None:
+    """
+    Recompute every matching key against the current normalization.
+
+    The keys are derived, so changing a helper in `services/matching.py` silently
+    invalidates every row already written — a book stored under the old rules simply
+    stops matching one stored under the new ones, with nothing to show that anything
+    is wrong. `normalize_title` learned to strip the compact edition markers a cover
+    actually carries ("2E", "5e"), which moved `Building Microservices, 2E` onto the
+    same key as `Building Microservices`.
+
+    Any later change to those helpers wants a revision that looks exactly like this
+    one. It is idempotent and safe to re-run.
+    """
+    from chitai.services.matching import (
+        normalize_author,
+        normalize_identifier,
+        normalize_title,
+    )
+
+    connection = op.get_bind()
+
+    books = connection.execute(sa.text("SELECT id, title FROM books")).fetchall()
+    _apply(
+        connection,
+        "UPDATE books SET normalized_title = :key WHERE id = :id",
+        [{"id": id, "key": normalize_title(title)} for id, title in books],
+    )
+
+    authors = connection.execute(sa.text("SELECT id, name FROM authors")).fetchall()
+    _apply(
+        connection,
+        "UPDATE authors SET normalized_name = :key WHERE id = :id",
+        [{"id": id, "key": normalize_author(name)} for id, name in authors],
+    )
+
+    identifiers = connection.execute(
+        sa.text("SELECT id, name, value FROM identifiers")
+    ).fetchall()
+    _apply(
+        connection,
+        "UPDATE identifiers SET normalized_value = :key WHERE id = :id",
+        [
+            {"id": id, "key": normalize_identifier(name, value)}
+            for id, name, value in identifiers
+        ],
+    )
+
+
+def _apply(connection, statement: str, parameters: list[dict]) -> None:
+    """Run one update per row, in batches, skipping the work when there are none."""
+    batch_size = 1000
+
+    for start in range(0, len(parameters), batch_size):
+        connection.execute(sa.text(statement), parameters[start : start + batch_size])
+
+
+def data_downgrades() -> None:
+    """Add any optional data downgrade migrations here!"""

backend/migrations/versions/2026-08-17_backfill_file_content_types_d2d69065ede3.py

@@ -0,0 +1,114 @@
+"""backfill file content types
+
+Revision ID: d2d69065ede3
+Revises: 49a9e85a0ffc
+Create Date: 2026-08-17 11:37:58.959135
+
+"""
+
+import warnings
+from typing import TYPE_CHECKING
+
+import sqlalchemy as sa
+from alembic import op
+from advanced_alchemy.types import EncryptedString, EncryptedText, GUID, ORA_JSONB, DateTimeUTC, StoredObject, PasswordHash, FernetBackend
+from advanced_alchemy.types.encrypted_string import PGCryptoBackend
+from advanced_alchemy.types.password_hash.argon2 import Argon2Hasher
+from advanced_alchemy.types.password_hash.passlib import PasslibHasher
+from advanced_alchemy.types.password_hash.pwdlib import PwdlibHasher
+from sqlalchemy import Text  # noqa: F401
+
+if TYPE_CHECKING:
+    from collections.abc import Sequence
+
+__all__ = ["downgrade", "upgrade", "schema_upgrades", "schema_downgrades", "data_upgrades", "data_downgrades"]
+
+sa.GUID = GUID
+sa.DateTimeUTC = DateTimeUTC
+sa.ORA_JSONB = ORA_JSONB
+sa.EncryptedString = EncryptedString
+sa.EncryptedText = EncryptedText
+sa.StoredObject = StoredObject
+sa.PasswordHash = PasswordHash
+sa.Argon2Hasher = Argon2Hasher
+sa.PasslibHasher = PasslibHasher
+sa.PwdlibHasher = PwdlibHasher
+sa.FernetBackend = FernetBackend
+sa.PGCryptoBackend = PGCryptoBackend
+
+# revision identifiers, used by Alembic.
+revision = 'd2d69065ede3'
+down_revision = '49a9e85a0ffc'
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=UserWarning)
+        with op.get_context().autocommit_block():
+            schema_upgrades()
+            data_upgrades()
+
+def downgrade() -> None:
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=UserWarning)
+        with op.get_context().autocommit_block():
+            data_downgrades()
+            schema_downgrades()
+
+def schema_upgrades() -> None:
+    """schema upgrade migrations go here."""
+    pass
+
+def schema_downgrades() -> None:
+    """schema downgrade migrations go here."""
+    pass
+
+def data_upgrades() -> None:
+    """
+    Name the format of every file whose content type was never worked out.
+
+    `create_many_from_existing_files` filled the column from `mimetypes.guess_type`,
+    which answers None for `.mobi`, `.azw`, `.fb2` and `.lit` — so a consume-directory
+    import of any of those stored a null, and the OPDS acquisition link a reader app
+    uses to decide what it can open carried nothing.
+
+    Both write paths now go through `guess_content_type`, which this uses too, so the
+    formats in its table get named retroactively. A row it still cannot name is **left
+    null** rather than filled with a placeholder: null is the truth, the column is
+    nullable, and the one consumer that needs a string substitutes one itself.
+    Idempotent: it only looks at rows that carry nothing.
+    """
+    from chitai.services.utils import guess_content_type
+
+    connection = op.get_bind()
+
+    files = connection.execute(
+        sa.text(
+            "SELECT id, path FROM file_metadata "
+            "WHERE content_type IS NULL OR content_type = ''"
+        )
+    ).fetchall()
+
+    parameters = [
+        {"id": id, "content_type": content_type}
+        for id, path in files
+        if (content_type := guess_content_type(path)) is not None
+    ]
+
+    if not parameters:
+        return
+
+    batch_size = 1000
+    for start in range(0, len(parameters), batch_size):
+        connection.execute(
+            sa.text(
+                "UPDATE file_metadata SET content_type = :content_type WHERE id = :id"
+            ),
+            parameters[start : start + batch_size],
+        )
+
+
+def data_downgrades() -> None:
+    """Add any optional data downgrade migrations here!"""

backend/src/chitai/app.py

@@ -16,6 +16,7 @@ from sqlalchemy.ext.asyncio import AsyncSession
 
 
 from chitai import controllers as c
+from chitai.cli import CalibreCLIPlugin
 from chitai.config import settings
 from chitai.database.config import alchemy
 from chitai.database.models.user import User
@@ -133,7 +134,7 @@ def create_app() -> Litestar:
         ],
         exception_handlers=exception_handlers,
         lifespan=[setup_db_connection, setup_directory_watcher],
-        plugins=[alchemy],
+        plugins=[alchemy, CalibreCLIPlugin()],
         on_app_init=[oauth2_auth.on_app_init],
         openapi_config=OpenAPIConfig(
             title="Chitai",

backend/src/chitai/cli.py

@@ -0,0 +1,186 @@
+# src/chitai/cli.py
+
+"""
+Extra commands on the `litestar` CLI.
+
+Registered through `CalibreCLIPlugin` in `app.py`, so they run as
+`litestar --app-dir src/chitai/ calibre-import …` and get the app's own configuration
+without a second way to load it.
+
+The Calibre import lives here as well as behind an endpoint because the case it exists
+for is a one-time migration of a library that may be hundreds of gigabytes. That should
+not depend on a browser tab staying open.
+"""
+
+from __future__ import annotations
+
+import asyncio
+from pathlib import Path
+
+import click
+from click import Group
+from litestar.plugins import CLIPluginProtocol
+
+from chitai.config import settings
+from chitai.database.models import Library
+from chitai.services.book import BookService, CalibreImportProgress, CalibreImportResult
+from chitai.services.calibre import CalibreLibrary, CalibreLibraryError
+from chitai.services.library import LibraryService
+
+
+class CalibreCLIPlugin(CLIPluginProtocol):
+    """Adds `calibre-import` to the Litestar CLI."""
+
+    def on_cli_init(self, cli: Group) -> None:
+        cli.add_command(calibre_import)
+
+
+@click.command(name="calibre-import")
+@click.argument(
+    "source",
+    type=click.Path(exists=True, file_okay=False, path_type=Path),
+)
+@click.option(
+    "--library",
+    "library_slug",
+    required=True,
+    help="Slug of the Chitai library to import into.",
+)
+@click.option(
+    "--allow-duplicates",
+    is_flag=True,
+    help="Import books whose files the library already holds.",
+)
+@click.option(
+    "--dry-run",
+    is_flag=True,
+    help="Read the catalogue and report what it holds, without writing anything.",
+)
+def calibre_import(
+    source: Path, library_slug: str, allow_duplicates: bool, dry_run: bool
+) -> None:
+    """
+    Import a Calibre library from SOURCE, the directory holding its metadata.db.
+
+    Files are copied, never moved: the Calibre library is left exactly as it is, and
+    re-running skips whatever is already stored.
+    """
+    asyncio.run(_import(source, library_slug, allow_duplicates, dry_run))
+
+
+async def _import(
+    source: Path, library_slug: str, allow_duplicates: bool, dry_run: bool
+) -> None:
+    try:
+        library_source = CalibreLibrary(source)
+        await library_source.open()
+    except CalibreLibraryError as exc:
+        raise click.ClickException(str(exc)) from exc
+
+    try:
+        if dry_run:
+            await _report(library_source)
+            return
+
+        async with settings.alchemy_config.get_session() as session:
+            library = await _library(session, library_slug)
+
+            result = await BookService(session=session).create_many_from_calibre(
+                library_source,
+                library,
+                allow_duplicates=allow_duplicates,
+                on_progress=_print_progress,
+            )
+    finally:
+        await library_source.close()
+
+    _print_summary(result)
+
+    if result.failed:
+        raise SystemExit(1)
+
+
+async def _library(session: object, slug: str) -> Library:
+    """Resolve the target library, or explain what the options were."""
+    service = LibraryService(session=session)  # type: ignore[arg-type]
+
+    library = await service.get_one_or_none(Library.slug == slug)
+
+    if library is None:
+        available = ", ".join(sorted(item.slug for item in await service.list()))
+        raise click.ClickException(
+            f"No library with slug '{slug}'. Available: {available or 'none'}"
+        )
+
+    # A read-only library is one pointing at a tree Chitai does not own. Copying books
+    # into it would write into somebody else's directory.
+    if library.read_only:
+        raise click.ClickException(
+            f"Library '{slug}' is read-only, so nothing can be imported into it"
+        )
+
+    return library
+
+
+async def _report(source: CalibreLibrary) -> None:
+    """Describe the catalogue without touching the database."""
+    books = await source.books()
+
+    click.echo(f"{len(books)} book(s) in {source.root}\n")
+
+    for book in books:
+        authors = ", ".join(book.authors) or "unknown author"
+        formats = ", ".join(file.format for file in book.files) or "no files"
+        click.echo(f"  #{book.calibre_id:<6} {book.title}")
+        click.echo(f"  {'':<7} {authors} · {formats}")
+
+    missing = [
+        book
+        for book in books
+        if any(not file.path.is_file() for file in book.files) or not book.files
+    ]
+
+    if missing:
+        click.echo(
+            f"\n{len(missing)} book(s) have files the catalogue lists "
+            "but disk does not:"
+        )
+        for book in missing:
+            click.echo(f"  #{book.calibre_id} {book.title}")
+
+
+def _print_progress(progress: CalibreImportProgress) -> None:
+    marker = {"created": "+", "skipped": "-", "failed": "!"}.get(progress.outcome, " ")
+    detail = f" ({progress.detail})" if progress.detail else ""
+
+    click.echo(
+        f"[{progress.processed:>5}/{progress.total}] {marker} {progress.title}{detail}"
+    )
+
+
+def _print_summary(result: CalibreImportResult) -> None:
+    click.echo(
+        f"\n{len(result.created)} created, {len(result.skipped)} skipped, "
+        f"{len(result.failed)} failed, of {result.total}."
+    )
+
+    if result.duplicate_files:
+        click.echo(
+            f"{len(result.duplicate_files)} individual file(s) were already stored and "
+            "were left out of books that imported otherwise."
+        )
+
+    for failure in result.failed:
+        click.echo(f"  failed  #{failure.calibre_id} {failure.title}: {failure.reason}")
+
+    # Imported all the same — a metadata match is a guess, and the duplicates screen is
+    # where these get decided.
+    for possible in result.possible_duplicates:
+        names = ", ".join(
+            f"{candidate.title} (#{candidate.book_id})"
+            for candidate in possible.candidates
+        )
+        click.echo(
+            f"  possible duplicate  {possible.title} (#{possible.book_id}) "
+            f"may already be in the library as: {names}"
+        )

backend/src/chitai/config.py

@@ -1,9 +1,25 @@
+from enum import StrEnum
+
 from pydantic import Field, PostgresDsn, computed_field
 from pydantic_settings import BaseSettings, SettingsConfigDict
 from advanced_alchemy.extensions.litestar import (
     SQLAlchemyAsyncConfig,
 )
 
+
+class DuplicateScope(StrEnum):
+    """How widely an incoming file is compared against what is already stored."""
+
+    LIBRARY = "library"
+    """Only files in the library being uploaded to count as duplicates."""
+
+    GLOBAL = "global"
+    """A file already held by any library counts as a duplicate."""
+
+    OFF = "off"
+    """No duplicate detection at all."""
+
+
 class Settings(BaseSettings):
     version: str = Field("0.0.1")
     project_name: str = Field("chitai")
@@ -33,6 +49,14 @@ class Settings(BaseSettings):
     # Path to consume directory
     consume_path: str = Field("./consume")
 
+    # Duplicate detection
+    duplicate_scope: DuplicateScope = Field(DuplicateScope.LIBRARY)
+
+    # Where the consume watcher parks files it refused as duplicates. Must sit
+    # outside `consume_path`, or the watcher picks them straight back up and
+    # tries to resolve the directory name as a library slug.
+    duplicate_path: str = Field("./duplicates")
+
     @computed_field
     @property
     def postgres_uri(self) -> PostgresDsn:

backend/src/chitai/controllers/book.py

@@ -12,7 +12,12 @@ from litestar.params import Dependency, Body
 from litestar.enums import RequestEncodingType
 from litestar.response import File, Stream
 from litestar.exceptions import HTTPException
-from litestar.status_codes import HTTP_400_BAD_REQUEST
+from litestar.status_codes import (
+    HTTP_200_OK,
+    HTTP_204_NO_CONTENT,
+    HTTP_400_BAD_REQUEST,
+    HTTP_409_CONFLICT,
+)
 from litestar.datastructures import UploadFile
 from advanced_alchemy.service.pagination import OffsetPagination
 from advanced_alchemy.filters import CollectionFilter
@@ -23,6 +28,24 @@ from chitai.services import dependencies as deps
 from chitai import schemas as s
 from chitai.database import models as m
 from chitai.services import BookService, BookProgressService
+from chitai.services.book import DuplicateFilesError
+
+
+def _duplicate_conflict(exc: DuplicateFilesError) -> HTTPException:
+    """
+    Turn refused files into a 409 the caller can act on.
+
+    The files ride along in `extra` so the client can name them and offer to send them
+    again with `allow_duplicates`, rather than being told only that something clashed.
+    """
+    return HTTPException(
+        status_code=HTTP_409_CONFLICT,
+        detail="These files are already in the library",
+        extra=[
+            s.DuplicateFileRead.model_validate(duplicate).model_dump(mode="json")
+            for duplicate in exc.duplicates
+        ],
+    )
 
 
 class BookController(Controller):
@@ -63,6 +86,7 @@ class BookController(Controller):
         books_service: BookService,
         library: m.Library,
         data: Annotated[s.BookCreate, Body(media_type=RequestEncodingType.MULTI_PART)],
+        allow_duplicates: bool = False,
     ) -> s.BookRead:
         """
         Create a new book with metadata and files.
@@ -73,6 +97,10 @@ class BookController(Controller):
         Path Parameters:
             library_id: The ID of the library the book belongs to.
 
+        Query Parameters:
+            allow_duplicates: If True, store the files even if the library already
+                holds them.
+
         Request Body:
             data: Book creation data including metadata and files.
 
@@ -83,9 +111,17 @@ class BookController(Controller):
         Returns:
             The created book as a BookRead schema.
 
+        Raises:
+            HTTPException: 409 if any of the files is already in the library.
         """
 
-        result = await books_service.create_book(data, library)
+        try:
+            result = await books_service.create_book(
+                data, library, screen_duplicates=not allow_duplicates
+            )
+        except DuplicateFilesError as exc:
+            raise _duplicate_conflict(exc)
+
         book = await books_service.get(result.id)
         return books_service.to_schema(book, schema_type=s.BookRead)
 
@@ -97,13 +133,20 @@ class BookController(Controller):
         data: Annotated[
             s.BooksCreateFromFiles, Body(media_type=RequestEncodingType.MULTI_PART)
         ],
-    ) -> OffsetPagination[s.BookRead]:
+        allow_duplicates: bool = False,
+    ) -> s.BooksUploadResult:
         """
         Create multiple books from uploaded files.
 
         Groups files by directory and creates separate books for each group.
         Metadata is automatically extracted from the files.
 
+        Files the library already holds are skipped rather than refused, and reported
+        back so the caller can say which ones did not make it in and why.
+
+        Query Parameters:
+            allow_duplicates: If True, store every file, even one already held.
+
         Request Body:
             data: Container with list of uploaded files.
 
@@ -112,19 +155,200 @@ class BookController(Controller):
             library: The library the books belong to.
 
         Returns:
-            Paginated list of created books.
+            The books created, and the files skipped as duplicates.
         """
         try:
-            results = await books_service.create_many_from_files(data, library)
+            result = await books_service.create_many_from_files(
+                data, library, allow_duplicates=allow_duplicates
+            )
         except ValueError:
             raise HTTPException(
                 status_code=HTTP_400_BAD_REQUEST, detail="Must upload at least one file"
             )
 
-        books = await books_service.list(
-            CollectionFilter("id", [result.id for result in results])
+        books = (
+            await books_service.list(
+                CollectionFilter("id", [book.id for book in result.books])
             )
-        return books_service.to_schema(books, schema_type=s.BookRead)
+            if result.books
+            else []
+        )
+
+        return s.BooksUploadResult(
+            created=[
+                books_service.to_schema(book, schema_type=s.BookRead) for book in books
+            ],
+            skipped=[
+                s.DuplicateFileRead.model_validate(duplicate)
+                for duplicate in result.duplicates
+            ],
+            possible_duplicates=[
+                s.PossibleDuplicateRead.model_validate(possible)
+                for possible in result.possible_duplicates
+            ],
+        )
+
+    @get(path="duplicate-books")
+    async def list_duplicate_books(
+        self, books_service: BookService, library: m.Library
+    ) -> list[s.DuplicateBookGroupRead]:
+        """
+        Report books already in the library that look like copies of one another.
+
+        The import-time check only ever sees what is arriving, so this is what covers
+        a collection someone already has. Matching is on metadata and therefore a
+        guess: a group is a question for the reader, not a verdict.
+
+        Query Parameters:
+            library_id: The library to review.
+
+        Injected Dependencies:
+            books_service: The book service for database operations.
+            library: The library to review.
+
+        Returns:
+            One entry per group of two or more books. Empty when there is nothing to
+            review, or when duplicate detection is switched off.
+        """
+        groups = await books_service.find_duplicate_book_groups(library)
+
+        return [
+            s.DuplicateBookGroupRead(
+                books=[s.DuplicateBookRead.model_validate(book) for book in group]
+            )
+            for group in groups
+        ]
+
+    @post(path="merge")
+    async def merge_books(
+        self, books_service: BookService, library: m.Library, data: s.BookMerge
+    ) -> s.BookRead:
+        """
+        Fold several books into one and delete the records folded in.
+
+        The survivor keeps its id, so links and bookmarks still resolve. Files, reading
+        progress, shelves, tags and unheld identifiers move onto it; metadata is only
+        changed by what `metadata` names, because choosing between two titles is the
+        reader's judgement rather than this endpoint's.
+
+        Nothing is removed from disk — a wrong merge should cost metadata that can be
+        retyped, not a book.
+
+        Query Parameters:
+            library_id: The library the books belong to.
+
+        Request Body:
+            data: The survivor, the books to fold in, and the resolved metadata.
+
+        Injected Dependencies:
+            books_service: The book service for database operations.
+            library: The library the books belong to.
+
+        Returns:
+            The surviving book.
+
+        Raises:
+            HTTPException: 400 if fewer than two distinct books were named, one is
+                unknown, or they do not all belong to one library.
+        """
+        try:
+            book = await books_service.merge_books(
+                data.survivor_id,
+                data.merged_ids,
+                library,
+                metadata=data.metadata.model_dump(exclude_unset=True)
+                if data.metadata
+                else None,
+            )
+        except ValueError as exc:
+            raise HTTPException(status_code=HTTP_400_BAD_REQUEST, detail=str(exc))
+
+        return books_service.to_schema(book, schema_type=s.BookRead)
+
+    @post(path="duplicate-books/dismissals", status_code=HTTP_204_NO_CONTENT)
+    async def dismiss_duplicate_books(
+        self, books_service: BookService, data: s.DuplicateDismissal
+    ) -> None:
+        """
+        Record that two books are not the same book.
+
+        Without this the review screen proposes the same wrong pair forever, which is
+        how a reader learns to stop looking at it.
+
+        Request Body:
+            data: The two book IDs. Order does not matter.
+
+        Injected Dependencies:
+            books_service: The book service for database operations.
+
+        Raises:
+            HTTPException: 400 if the two IDs are the same or either book is unknown.
+        """
+        try:
+            await books_service.dismiss_duplicates(data.book_a_id, data.book_b_id)
+        except ValueError as exc:
+            raise HTTPException(status_code=HTTP_400_BAD_REQUEST, detail=str(exc))
+
+    @delete(path="duplicate-books/dismissals")
+    async def restore_duplicate_books(
+        self, books_service: BookService, book_a_id: int, book_b_id: int
+    ) -> None:
+        """
+        Undo a dismissal, so the pair is proposed again.
+
+        Query Parameters:
+            book_a_id: One of the two books.
+            book_b_id: The other. Order does not matter.
+
+        Injected Dependencies:
+            books_service: The book service for database operations.
+        """
+        await books_service.restore_duplicates(book_a_id, book_b_id)
+
+    # A question, not a change: 200 rather than the 201 a POST would default to.
+    @post(path="duplicate-files", status_code=HTTP_200_OK)
+    async def check_duplicate_files(
+        self,
+        books_service: BookService,
+        library: m.Library,
+        data: list[s.FileFingerprint],
+    ) -> list[s.DuplicateFileRead]:
+        """
+        Report which of the given files the library already holds.
+
+        Lets a client ask before it uploads anything, which is the difference between
+        re-sending a folder of books and re-sending twelve kilobytes of hashes.
+
+        Query Parameters:
+            library_id: The library to check against.
+
+        Request Body:
+            data: Hash and size for each file, optionally with the name to echo back.
+
+        Injected Dependencies:
+            books_service: The book service for database operations.
+            library: The library to check against.
+
+        Returns:
+            One entry per submitted file that is already stored. Files that are not
+            are absent.
+        """
+        matches = await books_service.find_duplicate_files(
+            ((item.hash, item.size) for item in data), library
+        )
+
+        return [
+            s.DuplicateFileRead(
+                filename=item.filename or match.filename,
+                hash=match.hash,
+                size=match.size,
+                library_id=match.library_id,
+                book_id=match.book_id,
+                book_title=match.book_title,
+            )
+            for item in data
+            if (match := matches.get((item.hash, item.size))) is not None
+        ]
 
     @get(path="/{book_id:int}")
     async def get_book_by_id(
@@ -303,13 +527,20 @@ class BookController(Controller):
         ],
         library: m.Library,
         books_service: BookService,
+        allow_duplicates: bool = False,
     ) -> s.BookRead:
         """
         Add files to an existing book.
 
+        A file the book already carries is ignored, so re-sending one is harmless.
+
         Path Parameters:
             book_id: The ID of the book to modify
 
+        Query Parameters:
+            allow_duplicates: If True, store the files even if the library already
+                holds them.
+
         Request Body:
             files: The files to add to the book
 
@@ -320,9 +551,17 @@ class BookController(Controller):
         Returns:
             The modified book
 
+        Raises:
+            HTTPException: 409 if a file is already stored under a different book.
         """
 
-        await books_service.add_files(book_id, data, library)
+        try:
+            await books_service.add_files(
+                book_id, data, library, allow_duplicates=allow_duplicates
+            )
+        except DuplicateFilesError as exc:
+            raise _duplicate_conflict(exc)
+
         book = await books_service.get(book_id)
         return books_service.to_schema(book, schema_type=s.BookRead)
 

backend/src/chitai/controllers/library.py

@@ -1,12 +1,20 @@
 # src/chitai/controllers/library.py
 
 # Standard library
+import asyncio
+import shutil
+import tempfile
+from pathlib import Path
 from typing import Annotated
 
 # Third-party libraries
+import aiofiles
+from aiofiles import os as aios
 from litestar import Controller, post, get, patch, delete
-from litestar.params import Dependency
+from litestar.enums import RequestEncodingType
+from litestar.params import Body, Dependency
 from litestar.exceptions import HTTPException
+from litestar.status_codes import HTTP_200_OK, HTTP_202_ACCEPTED
 from advanced_alchemy.extensions.litestar.providers import create_service_dependencies
 from advanced_alchemy.service.pagination import OffsetPagination
 from advanced_alchemy.service import FilterTypeT
@@ -14,10 +22,21 @@ from advanced_alchemy.service import FilterTypeT
 # Local imports
 from chitai.database import models as m
 from chitai.services import LibraryService
-from chitai.schemas.library import LibraryCreate, LibraryRead
+from chitai.schemas.library import (
+    CalibreArchiveUpload,
+    CalibreImportRead,
+    LibraryCreate,
+    LibraryRead,
+)
+from chitai.services.calibre import CalibreLibraryError, extract_calibre_archive
+from chitai.services.calibre_import import registry
 from chitai.services.utils import DirectoryDoesNotExist
 
 
+# How much of an uploaded archive is held in memory at a time on its way to disk.
+UPLOAD_CHUNK_SIZE = 262144  # 256 KiB
+
+
 class LibraryController(Controller):
     """Controller for managing library operations."""
 
@@ -74,3 +93,147 @@ class LibraryController(Controller):
         return library_service.to_schema(
             results, total, filters, schema_type=LibraryRead
         )
+
+    @post(
+        path="{library_id:int}/imports/calibre/upload",
+        status_code=HTTP_202_ACCEPTED,
+        request_max_body_size=None,
+    )
+    async def upload_calibre_import(
+        self,
+        library_service: LibraryService,
+        library_id: int,
+        data: Annotated[
+            CalibreArchiveUpload, Body(media_type=RequestEncodingType.MULTI_PART)
+        ],
+    ) -> CalibreImportRead:
+        """
+        Import a zipped Calibre library that was uploaded rather than named on disk.
+
+        For the case where the library is not on the server: zip the Calibre folder and
+        send it. Unpacked into a temp directory the job owns and deletes when it ends —
+        by which time the books worth keeping have been copied into the library proper.
+
+        The server-folder route stays the one for a very large library. This one has to
+        carry the whole archive over HTTP first.
+
+        Path Parameters:
+            library_id: The library to import into.
+
+        Request Body:
+            data: The `.zip` holding the Calibre library.
+
+        Returns:
+            The job, already running.
+
+        Raises:
+            HTTPException: 400 if the archive is not a zip, holds no `metadata.db`,
+                names an entry outside itself, or would not fit on disk; 409 if an
+                import into this library is already running.
+        """
+        library = await self._importable(library_service, library_id)
+
+        if (running := registry.running_for(library_id)) is not None:
+            raise HTTPException(
+                status_code=409,
+                detail="An import into this library is already running",
+                extra={"job_id": running.id},
+            )
+
+        workspace = Path(await asyncio.to_thread(tempfile.mkdtemp))
+
+        try:
+            archive = workspace / "upload.zip"
+            await data.archive.seek(0)
+
+            async with aiofiles.open(archive, "wb") as destination:
+                while chunk := await data.archive.read(UPLOAD_CHUNK_SIZE):
+                    await destination.write(chunk)
+
+            unpacked = workspace / "library"
+            unpacked.mkdir()
+
+            catalogue = await extract_calibre_archive(archive, unpacked)
+
+            # The archive itself is dead weight once unpacked, and the library it
+            # unpacked to can be large.
+            await aios.remove(archive)
+        except CalibreLibraryError as exc:
+            await asyncio.to_thread(shutil.rmtree, workspace, True)
+            raise HTTPException(status_code=400, detail=str(exc))
+        except Exception:
+            await asyncio.to_thread(shutil.rmtree, workspace, True)
+            raise
+
+        job = registry.start(
+            library,
+            catalogue,
+            workspace=workspace,
+            label=data.archive.filename or "uploaded archive",
+            allow_duplicates=data.allow_duplicates,
+        )
+
+        return CalibreImportRead.model_validate(job)
+
+    @get(path="imports/{job_id:str}")
+    async def get_import(self, job_id: str) -> CalibreImportRead:
+        """
+        Report on an import.
+
+        Polled by the client while a run is going. Jobs are held in memory, so this is
+        answered by the process that started it — see `services/calibre_import.py`.
+
+        Path Parameters:
+            job_id: The job to report on.
+
+        Raises:
+            HTTPException: 404 if this process holds no such job.
+        """
+        if (job := registry.get(job_id)) is None:
+            raise HTTPException(status_code=404, detail="No such import")
+
+        return CalibreImportRead.model_validate(job)
+
+    @delete(path="imports/{job_id:str}", status_code=HTTP_200_OK)
+    async def cancel_import(self, job_id: str) -> CalibreImportRead:
+        """
+        Ask an import to stop after the book it is on.
+
+        Deliberately not an abort: a book abandoned mid-copy would leave files on disk
+        with no row describing them. Whatever it has imported stays imported.
+
+        Path Parameters:
+            job_id: The job to stop.
+
+        Raises:
+            HTTPException: 404 if this process holds no such job.
+        """
+        if (job := registry.cancel(job_id)) is None:
+            raise HTTPException(status_code=404, detail="No such import")
+
+        return CalibreImportRead.model_validate(job)
+
+    @staticmethod
+    async def _importable(
+        library_service: LibraryService, library_id: int
+    ) -> m.Library:
+        """
+        The library, if it can be imported into at all.
+
+        Raises:
+            HTTPException: 404 if there is no such library, 400 if it is read-only —
+                a read-only library points at a tree Chitai does not own, so copying
+                books into it would write into somebody else's directory.
+        """
+        library = await library_service.get_one_or_none(m.Library.id == library_id)
+
+        if library is None:
+            raise HTTPException(status_code=404, detail="No such library")
+
+        if library.read_only:
+            raise HTTPException(
+                status_code=400,
+                detail="This library is read-only, so nothing can be imported into it",
+            )
+
+        return library

backend/src/chitai/database/models/__init__.py

@@ -3,6 +3,7 @@ from .book import Book, Identifier, FileMetadata
 from .book_list import BookList, BookListLink
 from .book_progress import BookProgress
 from .book_series import BookSeries
+from .duplicate_dismissal import DuplicateDismissal
 from .kosync_device import KosyncDevice
 from .kosync_progress import KosyncProgress
 from .library import Library

backend/src/chitai/database/models/author.py

@@ -5,6 +5,7 @@ from sqlalchemy import ColumnElement, ForeignKey, UniqueConstraint
 from sqlalchemy.orm import Mapped
 from sqlalchemy.orm import mapped_column
 from sqlalchemy.orm import relationship
+from sqlalchemy.orm import validates
 
 from advanced_alchemy.base import BigIntAuditBase, BigIntBase
 from advanced_alchemy.mixins import UniqueMixin
@@ -16,18 +17,55 @@ if TYPE_CHECKING:
 class Author(BigIntAuditBase, UniqueMixin):
     __tablename__ = "authors"
 
+    # Always the canonical form — see `_canonicalize`. Extractors hand over whatever
+    # the file happened to say: "Newman, Sam;" from a `DC:creator` list, or
+    # "Sam Newman.epub" from a filename. Storing those verbatim is how one person ends
+    # up as several rows in the sidebar.
     name: Mapped[str] = mapped_column(unique=True, index=True)
+
+    # Kept current by `_canonicalize` too — never assign it directly. Not unique: two
+    # spellings that survive canonicalization, "Steve Mcconnell" and "Steve McConnell",
+    # are still one person to a reader, which is what this column exists to express.
+    normalized_name: Mapped[str] = mapped_column(default="", index=True)
+
     description: Mapped[Optional[str]]
 
+    @validates("name")
+    def _canonicalize(self, _key: str, name: str) -> str:
+        """
+        Store the tidied name, and derive the matching key from it.
+
+        A validator so no write can get around it, and `unique_hash` / `unique_filter`
+        below tidy the same way so `as_unique_async` looks the row up under the name it
+        would actually be stored as. All three have to agree: if the lookup used the
+        raw name and the insert used the tidy one, every variant spelling would miss
+        the existing row and then collide with it on the unique index.
+        """
+        # Imported here rather than at module scope: `chitai.services.matching` cannot
+        # be reached without initialising the `chitai.services` package, which imports
+        # the services, which import this module.
+        from chitai.services.matching import format_author_name, normalize_author
+
+        name = format_author_name(name)
+        self.normalized_name = normalize_author(name)
+        return name
+
+    @classmethod
+    def _tidy(cls, name: str) -> str:
+        """The name as `_canonicalize` would store it."""
+        from chitai.services.matching import format_author_name
+
+        return format_author_name(name)
+
     @classmethod
     def unique_hash(cls, name: str) -> Hashable:
         """Generate a unique hash for deduplication."""
-        return name
+        return cls._tidy(name)
 
     @classmethod
     def unique_filter(cls, name: str) -> ColumnElement[bool]:
         """SQL filter for finding existing records."""
-        return cls.name == name
+        return cls.name == cls._tidy(name)
 
     def __repr__(self) -> str:
         return f"Author({self.name!r})"

backend/src/chitai/database/models/book.py

@@ -2,13 +2,10 @@ from datetime import date
 from typing import TYPE_CHECKING, Any, Optional
 
 from sqlalchemy import Index, ForeignKey, UniqueConstraint
-from sqlalchemy.orm import Mapped, mapped_column, relationship
-from sqlalchemy.orm import mapped_column
-from sqlalchemy.orm import relationship
+from sqlalchemy.orm import Mapped, mapped_column, relationship, validates
 from sqlalchemy.ext.orderinglist import ordering_list
 from sqlalchemy.ext.associationproxy import association_proxy
 from sqlalchemy.ext.associationproxy import AssociationProxy
-from sqlalchemy.orm.collections import attribute_keyed_dict
 
 from advanced_alchemy.base import BigIntAuditBase, BigIntBase
 
@@ -44,6 +41,13 @@ class Book(BigIntAuditBase):
     library: Mapped["Library"] = relationship(back_populates="books")
 
     title: Mapped[str]
+
+    # Kept current by `_normalize_title` below — never assign it directly.
+    #
+    # Deliberately not unique: two spellings collapsing onto one value is the whole
+    # point of the column, and a second edition is allowed to exist.
+    normalized_title: Mapped[str] = mapped_column(default="", index=True)
+
     subtitle: Mapped[Optional[str]]
     description: Mapped[Optional[str]]
     published_date: Mapped[Optional[date]]
@@ -111,6 +115,24 @@ class Book(BigIntAuditBase):
     def progress(self) -> Optional["BookProgress"]:
         return self.progress_records[0] if self.progress_records else None
 
+    @validates("title")
+    def _normalize_title(self, _key: str, title: str) -> str:
+        """
+        Derive `normalized_title` from whatever writes the title.
+
+        A validator rather than a service call because `BookService` sets titles from
+        at least three places — `to_model_on_create`, `to_model_on_update` and the
+        `setattr` loop in `_populate_with_unique_relationships` — and a fourth would
+        otherwise leave the key silently stale.
+        """
+        # Imported here rather than at module scope: `chitai.services.matching` cannot
+        # be reached without initialising the `chitai.services` package, which imports
+        # the services, which import this module.
+        from chitai.services.matching import normalize_title
+
+        self.normalized_title = normalize_title(title)
+        return title
+
     def __repr__(self) -> str:
         return f"Book({self.title=!r})"
 
@@ -132,6 +154,24 @@ class Identifier(BigIntBase):
     book_id: Mapped[int] = mapped_column(ForeignKey("books.id", ondelete="cascade"))
     value: Mapped[str]
 
+    # Kept current by `_normalize` below — never assign it directly. Null for an
+    # identifier that cannot carry a match: a per-build UUID, or an ISBN that fails
+    # its own checksum.
+    normalized_value: Mapped[Optional[str]] = mapped_column(index=True)
+
+    @validates("name", "value")
+    def _normalize(self, key: str, value: str) -> str:
+        """Recompute `normalized_value` whenever either half of the pair changes."""
+        from chitai.services.matching import normalize_identifier
+
+        name = value if key == "name" else self.name
+        raw = value if key == "value" else self.value
+
+        self.normalized_value = (
+            normalize_identifier(name, raw) if name and raw else None
+        )
+        return value
+
     def __repr__(self):
         return f"Identifier({self.name!r} : {self.value!r})"
 
@@ -139,6 +179,15 @@ class Identifier(BigIntBase):
 class FileMetadata(BigIntBase):
     __tablename__ = "file_metadata"
 
+    __table_args__ = (
+        # Deliberately not unique. The hash is KOReader's partial MD5, which samples
+        # 12 KiB of the file, so two genuinely different files can collide — and an
+        # existing database may already hold duplicates, which a unique index would
+        # refuse to build over. Duplicate detection pairs it with `size` and treats a
+        # match as advisory, so this only has to make the lookup cheap.
+        Index("ix_file_metadata_hash", "hash"),
+    )
+
     book_id: Mapped[int] = mapped_column(ForeignKey("books.id", ondelete="cascade"))
     book: Mapped[Book] = relationship(back_populates="files")
     hash: Mapped[str]

backend/src/chitai/database/models/duplicate_dismissal.py

@@ -0,0 +1,37 @@
+from sqlalchemy import ForeignKey, UniqueConstraint
+from sqlalchemy.orm import Mapped, mapped_column
+
+from advanced_alchemy.base import BigIntBase
+
+
+class DuplicateDismissal(BigIntBase):
+    """
+    Two books a reader has said are not the same book.
+
+    Title and author matching is probabilistic, so it will keep proposing a second
+    edition, a translation and a sequel that shares its predecessor's name. A review
+    screen with no way to disagree with it nags forever, which is how people learn to
+    ignore a screen.
+
+    The pair is stored ordered — `book_a_id` is always the lower id — so "A and B" and
+    "B and A" are one row and the unique constraint can do its job. Use `pair()` rather
+    than assigning the columns directly.
+    """
+
+    __tablename__ = "duplicate_dismissals"
+    __table_args__ = (UniqueConstraint("book_a_id", "book_b_id"),)
+
+    book_a_id: Mapped[int] = mapped_column(
+        ForeignKey("books.id", ondelete="cascade"), index=True
+    )
+    book_b_id: Mapped[int] = mapped_column(
+        ForeignKey("books.id", ondelete="cascade"), index=True
+    )
+
+    @staticmethod
+    def pair(first: int, second: int) -> tuple[int, int]:
+        """The two book ids in the order this table stores them."""
+        return (first, second) if first <= second else (second, first)
+
+    def __repr__(self) -> str:
+        return f"DuplicateDismissal({self.book_a_id!r}, {self.book_b_id!r})"

backend/src/chitai/schemas/__init__.py

@@ -4,7 +4,15 @@ from .book import (
     BookProgressCreate,
     BookProgressRead,
     BooksCreateFromFiles,
+    BooksUploadResult,
+    BookMerge,
     BookMetadataUpdate,
+    DuplicateBookGroupRead,
+    DuplicateBookRead,
+    DuplicateDismissal,
+    DuplicateFileRead,
+    FileFingerprint,
+    PossibleDuplicateRead,
     FileMetadataRead,
     BookSeriesRead,
 )

backend/src/chitai/schemas/book.py

@@ -30,7 +30,11 @@ class FileMetadataRead(BaseModel):
     path: str
     hash: str
     size: int
-    content_type: str
+
+    # Nullable, though every ingest path now writes one through
+    # `guess_content_type`. Rows predating it can hold null, and a required field here
+    # turns one of those into a 500 on a book the reader can otherwise open.
+    content_type: str | None = None
 
     @computed_field
     @property
@@ -129,6 +133,83 @@ class BooksCreateFromFiles(BaseModel):
     model_config = ConfigDict(arbitrary_types_allowed=True)
 
 
+class FileFingerprint(BaseModel):
+    """What a client can say about a file it has not uploaded yet."""
+
+    hash: str
+    size: int
+    filename: str = ""
+
+
+class DuplicateFileRead(BaseModel):
+    """A file that was not stored because the library already holds its bytes."""
+
+    model_config = ConfigDict(from_attributes=True)
+
+    filename: str
+    hash: str
+    size: int
+    library_id: int
+
+    # Null when the match was another file in the same upload, which has no row yet.
+    book_id: int | None = None
+    book_title: str | None = None
+
+
+class DuplicateBookRead(BaseModel):
+    """
+    A stored book that may be the same book as another one.
+
+    Unlike `DuplicateFileRead` this is a guess: the evidence is metadata two editions
+    of one work legitimately share. Nothing was refused on the strength of it.
+    """
+
+    model_config = ConfigDict(from_attributes=True)
+
+    book_id: int
+    title: str
+    authors: list[str]
+    library_id: int
+    cover_image: Path | None = None
+
+    # Why it matched: "identifier" and/or "title-author".
+    matched_on: list[str]
+
+
+class PossibleDuplicateRead(BaseModel):
+    """A book that was imported, together with what it might be a second copy of."""
+
+    model_config = ConfigDict(from_attributes=True)
+
+    book_id: int
+    title: str
+    candidates: list[DuplicateBookRead]
+
+
+class DuplicateBookGroupRead(BaseModel):
+    """Books the library holds that all look like copies of one book."""
+
+    books: list[DuplicateBookRead]
+
+
+class DuplicateDismissal(BaseModel):
+    """Two books a reader is saying are not the same book."""
+
+    book_a_id: int
+    book_b_id: int
+
+
+class BooksUploadResult(BaseModel):
+    """The outcome of a multi-file upload: what was created, and what was skipped."""
+
+    created: list["BookRead"]
+    skipped: list[DuplicateFileRead]
+
+    # Created, not skipped — these are books that went in and look like something the
+    # library already had. The reader decides what to do about it.
+    possible_duplicates: list[PossibleDuplicateRead] = Field(default_factory=list)
+
+
 class BookMetadataUpdate(BaseModel):
     title: str | None = None
     subtitle: str | None = None
@@ -170,6 +251,20 @@ class BookMetadataUpdate(BaseModel):
         return v
 
 
+class BookMerge(BaseModel):
+    """
+    Fold several books into one.
+
+    `metadata` is the reader's resolution of the fields the records disagreed on.
+    Anything it does not name keeps the survivor's value — merging metadata is a
+    judgement, so nothing is guessed on the caller's behalf.
+    """
+
+    survivor_id: int
+    merged_ids: list[int]
+    metadata: Optional["BookMetadataUpdate"] = None
+
+
 class BookProgressCreate(BaseModel):
     percentage: float
     epub_cfi: str | None = None

backend/src/chitai/schemas/library.py

@@ -1,6 +1,7 @@
 from pathlib import Path
 from typing import Annotated
-from pydantic import BaseModel, Field, computed_field
+from pydantic import BaseModel, ConfigDict, Field, SkipValidation, computed_field
+from litestar.datastructures import UploadFile
 from  advanced_alchemy.utils.text import slugify
 
 class LibraryCreate(BaseModel):
@@ -27,6 +28,57 @@ class LibraryRead(BaseModel):
     total: int | None = None
 
 
+class CalibreArchiveUpload(BaseModel):
+    """
+    A zipped Calibre library.
+
+    The only way in through the API: a desktop Calibre install is usually not on the
+    server at all. Importing from a path the server can already see is a server-side
+    operation, and stays one — `litestar calibre-import` does that.
+    """
+
+    archive: Annotated[UploadFile, SkipValidation]
+    allow_duplicates: bool = False
+
+    model_config = ConfigDict(arbitrary_types_allowed=True)
+
+
+class ImportFailureRead(BaseModel):
+    """A book the import could not store."""
+
+    model_config = ConfigDict(from_attributes=True)
+
+    calibre_id: int
+    title: str
+    reason: str
+
+
+class CalibreImportRead(BaseModel):
+    """A running or finished import."""
+
+    model_config = ConfigDict(from_attributes=True)
+
+    id: str
+    library_id: int
+    source: str
+    state: str
+
+    total: int
+    processed: int
+    created: int
+    skipped: int
+    failed: int
+
+    current_title: str | None = None
+    failures: list[ImportFailureRead] = Field(default_factory=list)
+
+    # A count, not the records. The library's duplicates screen is what shows them.
+    possible_duplicates: int = 0
+
+    # Set when the run itself broke, as opposed to individual books failing.
+    error: str | None = None
+
+
 class LibraryUpdate(BaseModel):
     name: str | None
     root_path: str | None

backend/src/chitai/services/calibre.py

@@ -0,0 +1,570 @@
+# src/chitai/services/calibre.py
+
+"""
+Read a Calibre library.
+
+This knows about `metadata.db` and the tree beside it, and nothing about `Book`,
+`BookService` or a database session — it is a file-format reader, and it is testable
+without Postgres or an app. Interpretation belongs to whoever imports what it returns:
+identifiers come back exactly as Calibre wrote them, not folded onto Chitai's schemes.
+
+Things about Calibre that are load-bearing here:
+
+- **Never query the views.** `meta` and the `tag_browser_*` family call SQLite functions
+  Calibre registers from Python at connection time, so `SELECT * FROM meta` fails with
+  `no such function: sortconcat`. Only base tables are touched below.
+- **An unknown date is a sentinel, not a null** — `0101-01-01`, Calibre's
+  `UNDEFINED_DATE`. It parses fine as a date, so nothing complains; it just makes every
+  book without a publication date look like it was published in the year 101.
+- **`data.name` is not the title.** It is the on-disk stem, truncated to Calibre's
+  filename limit and sanitised, so the file is `The Project Gutenberg eBook #33283_
+  Calcul - Silvanus Phillips Thompson.pdf` for a book titled `The Project Gutenberg
+  eBook #33283: Calculus Made Easy, 2nd Edition`. Names locate files; the database
+  carries the metadata.
+- **`authors.name` escapes a comma as `|`**, which Calibre reverses on read
+  (`AuthorsTable.unserialize` in its `db/tables.py`).
+- **`series_index` defaults to 1.0 whether or not the book is in a series**, so a
+  position is only meaningful alongside a series.
+- **`books_pages_link` is recent and often empty.** It is treated as optional both ways:
+  the table may not exist, and where it does the rows are frequently `pages = 0` with
+  `needs_scan = 1`.
+
+Nothing walks the tree: every file is located through `books.path`, which is why
+`.caltrash` — where Calibre keeps deleted books, still on disk — cannot be picked up by
+accident.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import shutil
+import sqlite3
+import tempfile
+import zipfile
+from collections import defaultdict
+from collections.abc import Callable
+from dataclasses import dataclass, field
+from datetime import date, datetime
+from html.parser import HTMLParser
+from pathlib import Path, PurePosixPath
+
+
+METADATA_DB = "metadata.db"
+COVER_FILENAME = "cover.jpg"
+
+# Calibre writes `0101-01-01` for "no date". Any year this early is that sentinel rather
+# than a publication date somebody meant.
+EARLIEST_REAL_YEAR = 1000
+
+# The sidecars a WAL-mode database keeps beside itself. Copied along with it so the
+# snapshot can be recovered, since Calibre may be running while this reads.
+_DATABASE_SIDECARS = ("-wal", "-shm")
+
+
+class CalibreLibraryError(Exception):
+    """The library cannot be read at all — wrong directory, or no catalogue in it."""
+
+
+# How far down an archive to look for `metadata.db`. Zipping a Calibre library gives
+# either the directory itself or its contents, and a file manager may add a wrapper
+# folder on top, so two levels of nesting is normal and more is somebody's backup tree.
+_ARCHIVE_SEARCH_DEPTH = 3
+
+# Extraction is refused unless the destination has the uncompressed size plus this
+# much headroom. Filling the disk would take the whole application down, not just the
+# import.
+_DISK_HEADROOM = 256 * 1024 * 1024  # 256 MiB
+
+
+def _archive_members(archive: zipfile.ZipFile) -> list[zipfile.ZipInfo]:
+    """
+    The entries worth extracting, refusing any that would escape the destination.
+
+    `ZipFile.extract` does sanitise names, but relying on that silently is how the next
+    person to swap the extraction call reintroduces zip-slip. An archive naming
+    `../../etc/anything` is malformed or hostile, and either way there is nothing to
+    salvage by continuing.
+
+    Raises:
+        CalibreLibraryError: If any entry points outside the archive root.
+    """
+    members = []
+
+    for member in archive.infolist():
+        if member.is_dir():
+            continue
+
+        name = PurePosixPath(member.filename)
+
+        if name.is_absolute() or ".." in name.parts:
+            raise CalibreLibraryError(
+                f"The archive contains an entry outside itself: {member.filename!r}"
+            )
+
+        members.append(member)
+
+    return members
+
+
+async def extract_calibre_archive(archive: Path, destination: Path) -> Path:
+    """
+    Unpack a zipped Calibre library and find the catalogue inside it.
+
+    Args:
+        archive: The `.zip` to unpack.
+        destination: An empty directory to unpack into. The caller owns it and is
+            responsible for removing it.
+
+    Returns:
+        The directory holding `metadata.db`, which is what `CalibreLibrary` takes.
+
+    Raises:
+        CalibreLibraryError: If the file is not a zip, names an entry outside itself,
+            would not fit on disk, or holds no `metadata.db`.
+    """
+    return await asyncio.to_thread(_extract_calibre_archive, archive, destination)
+
+
+def _extract_calibre_archive(archive: Path, destination: Path) -> Path:
+    if not zipfile.is_zipfile(archive):
+        raise CalibreLibraryError(
+            "That is not a zip file. A Calibre library has to be zipped, not tarred."
+        )
+
+    with zipfile.ZipFile(archive) as opened:
+        members = _archive_members(opened)
+
+        if not any(
+            PurePosixPath(member.filename).name == METADATA_DB for member in members
+        ):
+            raise CalibreLibraryError(
+                f"The archive holds no {METADATA_DB}, so it is not a Calibre library"
+            )
+
+        # Checked before writing rather than discovered part-way through: a full disk
+        # takes the whole application down, and the number is in the archive already.
+        needed = sum(member.file_size for member in members)
+        free = shutil.disk_usage(destination).free
+
+        if needed + _DISK_HEADROOM > free:
+            raise CalibreLibraryError(
+                f"Unpacking needs {needed // (1024 * 1024)} MiB and only "
+                f"{free // (1024 * 1024)} MiB is free"
+            )
+
+        opened.extractall(destination, members=members)
+
+    return _find_catalogue(destination)
+
+
+def _find_catalogue(root: Path) -> Path:
+    """The shallowest directory under `root` holding a `metadata.db`."""
+    candidates = sorted(
+        (path.parent for path in root.rglob(METADATA_DB) if path.is_file()),
+        key=lambda path: len(path.relative_to(root).parts),
+    )
+
+    for candidate in candidates:
+        if len(candidate.relative_to(root).parts) <= _ARCHIVE_SEARCH_DEPTH:
+            return candidate
+
+    raise CalibreLibraryError(
+        f"No {METADATA_DB} within {_ARCHIVE_SEARCH_DEPTH} levels of the archive root"
+    )
+
+
+@dataclass(frozen=True)
+class CalibreFile:
+    """One row of Calibre's `data` table: a book in one format."""
+
+    path: Path
+    """Absolute path, resolved against the library root. Not checked for existence."""
+
+    format: str
+    """As Calibre stores it, upper case: `EPUB`, `PDF`, `AZW3`."""
+
+    size: int
+    """`data.uncompressed_size` — Calibre's claim, not a fresh stat."""
+
+
+@dataclass(frozen=True)
+class CalibreBook:
+    """One book, with everything Chitai has a column for and nothing it does not."""
+
+    calibre_id: int
+    uuid: str
+    title: str
+    authors: list[str] = field(default_factory=list)
+    description: str | None = None
+    published_date: date | None = None
+    series: str | None = None
+    series_position: str | None = None
+    tags: list[str] = field(default_factory=list)
+    publisher: str | None = None
+    language: str | None = None
+
+    identifiers: dict[str, str] = field(default_factory=dict)
+    """Keyed by `identifiers.type` verbatim — `isbn`, `mobi-asin`, `amazon`."""
+
+    pages: int | None = None
+    cover: Path | None = None
+    files: list[CalibreFile] = field(default_factory=list)
+
+
+class CalibreLibrary:
+    """
+    A Calibre library on disk, opened for reading.
+
+    The catalogue is **copied** before it is read. Calibre may be running and writing,
+    and opening the live file either sees a torn state or needs to recover a write-ahead
+    log, which read-only access cannot do. The copy is small — hundreds of kilobytes for
+    a handful of books, single-digit megabytes for thousands — so this costs nothing and
+    removes the question. The original is never opened by SQLite at all.
+    """
+
+    def __init__(self, root: Path | str) -> None:
+        self.root = Path(root)
+        self._connection: sqlite3.Connection | None = None
+        self._workspace: Path | None = None
+
+        # Every query runs in a worker thread, and `asyncio.to_thread` hands out
+        # whichever one is free — so the connection outlives the thread that opened it
+        # and `check_same_thread` has to be off. The lock is what makes that safe: it
+        # keeps two queries off the connection at once, which is the thing that check
+        # was standing in for.
+        self._lock = asyncio.Lock()
+
+    @property
+    def database(self) -> Path:
+        return self.root / METADATA_DB
+
+    async def open(self) -> None:
+        """
+        Copy the catalogue aside and connect to the copy.
+
+        Raises:
+            CalibreLibraryError: If there is no `metadata.db` under the root.
+        """
+        if self._connection is not None:
+            return
+
+        if not await asyncio.to_thread(self.database.is_file):
+            raise CalibreLibraryError(
+                f"No {METADATA_DB} in '{self.root}' — that is not a Calibre library"
+            )
+
+        self._workspace = Path(await asyncio.to_thread(tempfile.mkdtemp))
+        copy = self._workspace / METADATA_DB
+
+        await asyncio.to_thread(self._copy_database, copy)
+
+        # Read-write on our own copy, deliberately: that is what lets SQLite recover a
+        # write-ahead log the source may have been mid-way through.
+        self._connection = sqlite3.connect(str(copy), check_same_thread=False)
+
+    def _copy_database(self, destination: Path) -> None:
+        shutil.copy2(self.database, destination)
+
+        for suffix in _DATABASE_SIDECARS:
+            sidecar = self.database.with_name(self.database.name + suffix)
+            if sidecar.is_file():
+                shutil.copy2(sidecar, destination.with_name(destination.name + suffix))
+
+    async def close(self) -> None:
+        """
+        Disconnect and remove the copy. Safe to call more than once.
+
+        The copy is removed even if closing the connection fails — otherwise a failure
+        here leaves a catalogue-sized file in the temp directory, and the caller that
+        failed is exactly the one that will not come back to tidy up.
+        """
+        try:
+            if self._connection is not None:
+                async with self._lock:
+                    self._connection.close()
+                    self._connection = None
+        finally:
+            if self._workspace is not None:
+                await asyncio.to_thread(shutil.rmtree, self._workspace, True)
+                self._workspace = None
+
+    async def __aenter__(self) -> CalibreLibrary:
+        await self.open()
+        return self
+
+    async def __aexit__(self, *_exception: object) -> None:
+        await self.close()
+
+    async def count(self) -> int:
+        """How many books the catalogue holds, without reading any of them."""
+        rows = await self._in_thread(lambda: self._execute("SELECT count(*) FROM books"))
+        return int(rows[0][0])
+
+    async def books(self) -> list[CalibreBook]:
+        """
+        Read the whole catalogue.
+
+        One query per table and the joining done in Python, rather than a per-book query
+        across ten tables. Everything Chitai stores about a book is small, so a
+        self-hosted catalogue fits in memory comfortably.
+
+        Returns:
+            Every book, in Calibre id order.
+        """
+        return await self._in_thread(self._read_books)
+
+    async def _in_thread[T](self, work: Callable[[], T]) -> T:
+        """Run one unit of SQLite work off the event loop, and only one at a time."""
+        async with self._lock:
+            return await asyncio.to_thread(work)
+
+    def _execute(self, statement: str) -> list[tuple]:
+        if self._connection is None:
+            raise CalibreLibraryError("The library is not open")
+
+        return self._connection.execute(statement).fetchall()
+
+    def _has_table(self, name: str) -> bool:
+        return bool(
+            self._execute(
+                f"SELECT 1 FROM sqlite_master WHERE type = 'table' AND name = '{name}'"
+            )
+        )
+
+    def _read_books(self) -> list[CalibreBook]:
+        authors = self._grouped(
+            "SELECT bal.book, a.name FROM books_authors_link bal "
+            "JOIN authors a ON a.id = bal.author ORDER BY bal.id"
+        )
+        tags = self._grouped(
+            "SELECT btl.book, t.name FROM books_tags_link btl "
+            "JOIN tags t ON t.id = btl.tag ORDER BY t.name"
+        )
+        # Calibre's link tables are unique per book for these two, so the last write
+        # wins and there is nothing to choose between.
+        series = self._mapped(
+            "SELECT bsl.book, s.name FROM books_series_link bsl "
+            "JOIN series s ON s.id = bsl.series"
+        )
+        publishers = self._mapped(
+            "SELECT bpl.book, p.name FROM books_publishers_link bpl "
+            "JOIN publishers p ON p.id = bpl.publisher"
+        )
+        # A book can carry several languages; Chitai holds one, so the first wins.
+        languages = self._grouped(
+            "SELECT bll.book, l.lang_code FROM books_languages_link bll "
+            "JOIN languages l ON l.id = bll.lang_code ORDER BY bll.item_order"
+        )
+        descriptions = self._mapped("SELECT book, text FROM comments")
+
+        identifiers: dict[int, dict[str, str]] = defaultdict(dict)
+        for book_id, name, value in self._execute(
+            "SELECT book, type, val FROM identifiers"
+        ):
+            if name and value:
+                identifiers[book_id][str(name)] = str(value)
+
+        files: dict[int, list[tuple[str, str, int]]] = defaultdict(list)
+        for book_id, format, name, size in self._execute(
+            "SELECT book, format, name, uncompressed_size FROM data ORDER BY id"
+        ):
+            files[book_id].append((str(format), str(name), int(size or 0)))
+
+        pages: dict[int, int] = {}
+        if self._has_table("books_pages_link"):
+            pages = {
+                book_id: int(count)
+                for book_id, count in self._execute(
+                    "SELECT book, pages FROM books_pages_link WHERE pages > 0"
+                )
+            }
+
+        books = []
+        for row in self._execute(
+            "SELECT id, title, pubdate, series_index, path, uuid, has_cover "
+            "FROM books ORDER BY id"
+        ):
+            book_id, title, pubdate, series_index, path, uuid, has_cover = row
+            directory = self.root / Path(str(path))
+            in_series = series.get(book_id)
+
+            books.append(
+                CalibreBook(
+                    calibre_id=book_id,
+                    uuid=str(uuid or ""),
+                    title=str(title or ""),
+                    authors=[unescape_author(name) for name in authors.get(book_id, [])],
+                    description=strip_html(descriptions.get(book_id)),
+                    published_date=parse_date(pubdate),
+                    series=in_series,
+                    # Meaningless without a series: Calibre defaults the index to 1.0 for
+                    # every book, in a series or not.
+                    series_position=(
+                        format_series_index(series_index) if in_series else None
+                    ),
+                    tags=tags.get(book_id, []),
+                    publisher=publishers.get(book_id),
+                    language=next(iter(languages.get(book_id, [])), None),
+                    identifiers=dict(identifiers.get(book_id, {})),
+                    pages=pages.get(book_id),
+                    cover=directory / COVER_FILENAME if has_cover else None,
+                    files=[
+                        CalibreFile(
+                            path=directory / f"{name}.{format.lower()}",
+                            format=format,
+                            size=size,
+                        )
+                        for format, name, size in files.get(book_id, [])
+                    ],
+                )
+            )
+
+        return books
+
+    def _grouped(self, statement: str) -> dict[int, list[str]]:
+        """Run a `(book, value)` query into one list per book, keeping row order."""
+        grouped: dict[int, list[str]] = defaultdict(list)
+
+        for book_id, value in self._execute(statement):
+            if value is not None:
+                grouped[book_id].append(str(value))
+
+        return grouped
+
+    def _mapped(self, statement: str) -> dict[int, str]:
+        """Run a `(book, value)` query into one value per book."""
+        return {
+            book_id: str(value)
+            for book_id, value in self._execute(statement)
+            if value is not None
+        }
+
+
+def unescape_author(name: str) -> str:
+    """
+    Undo Calibre's comma escaping.
+
+    `authors.name` stores a comma as `|`, and Calibre reverses it on the way out. Left
+    alone, `Doyle, Sir Arthur Conan` comes back as `Doyle| Sir Arthur Conan`.
+    """
+    return name.replace("|", ",").strip()
+
+
+def parse_date(value: object) -> date | None:
+    """
+    Read one of Calibre's timestamps, discarding its "unknown" sentinel.
+
+    Args:
+        value: The stored column, which is text in practice but need not be.
+
+    Returns:
+        The date, or None for a null, an unparseable value, or Calibre's
+        `0101-01-01` placeholder.
+    """
+    if value is None:
+        return None
+
+    if isinstance(value, datetime):
+        parsed = value.date()
+    elif isinstance(value, date):
+        parsed = value
+    else:
+        text = str(value).strip()
+        if not text:
+            return None
+
+        try:
+            parsed = datetime.fromisoformat(text).date()
+        except ValueError:
+            try:
+                parsed = date.fromisoformat(text[:10])
+            except ValueError:
+                return None
+
+    return parsed if parsed.year >= EARLIEST_REAL_YEAR else None
+
+
+def format_series_index(index: object) -> str | None:
+    """
+    Render `series_index` as the string `Book.series_position` holds.
+
+    Calibre stores a REAL, so volume seven arrives as `7.0` — which would be stored
+    verbatim and then compared as a string against the `7` everything else writes.
+    Fractional positions are real and are kept: `1.5` is a novella between two novels.
+    """
+    if index is None:
+        return None
+
+    try:
+        number = float(index)
+    except (TypeError, ValueError):
+        return None
+
+    return str(int(number)) if number.is_integer() else f"{number:g}"
+
+
+class _TextExtractor(HTMLParser):
+    """Flatten markup to text, keeping the line breaks that carried meaning."""
+
+    # Tags whose boundaries are a line break rather than nothing at all. Without these
+    # a description of three paragraphs comes out as one run-on sentence.
+    _BREAKS = frozenset(
+        {
+            "p", "br", "div", "li", "tr", "blockquote", "hr",
+            "h1", "h2", "h3", "h4", "h5", "h6",
+        }
+    )
+
+    def __init__(self) -> None:
+        super().__init__(convert_charrefs=True)
+        self._parts: list[str] = []
+
+    def handle_data(self, data: str) -> None:
+        self._parts.append(data)
+
+    def handle_starttag(self, tag: str, _attrs: object) -> None:
+        self._break(tag)
+
+    def handle_endtag(self, tag: str) -> None:
+        self._break(tag)
+
+    def _break(self, tag: str) -> None:
+        """
+        End the current line, once.
+
+        Both halves of `</p><p>` are a boundary, and the open tag of the very first
+        block is not one at all — so emitting a newline per tag turns two paragraphs
+        into two blank-line-separated ones with a leading gap. One break per boundary
+        is what the plain text wants.
+        """
+        if tag in self._BREAKS and self._parts and not self._parts[-1].endswith("\n"):
+            self._parts.append("\n")
+
+    @property
+    def text(self) -> str:
+        lines = [line.strip() for line in "".join(self._parts).splitlines()]
+
+        return "\n".join(line for line in lines if line).strip()
+
+
+def strip_html(html: str | None) -> str | None:
+    """
+    Turn Calibre's `comments` into plain text.
+
+    `comments.text` is always HTML, and `Book.description` is rendered as text — so the
+    tags would show literally on the book page.
+
+    Args:
+        html: The stored comment, if there is one.
+
+    Returns:
+        The text, or None when there was nothing or nothing survived.
+    """
+    if not html:
+        return None
+
+    parser = _TextExtractor()
+    parser.feed(html)
+    parser.close()
+
+    return parser.text or None

backend/src/chitai/services/calibre_import.py

@@ -0,0 +1,239 @@
+# src/chitai/services/calibre_import.py
+
+"""
+Run a Calibre import in the background and report on it.
+
+The import outlives its request — a real library takes minutes to hours — so the handler
+starts a task and hands back a handle to poll. The work itself is
+`BookService.create_many_from_calibre`; everything here is lifecycle: state, progress,
+cancellation, and a session of its own.
+
+**This registry lives in memory, and therefore assumes one worker process.** That holds
+today: the production `CMD` is `litestar run`, which is single-process, and the consume
+watcher is already an in-process singleton with the same constraint. `TODO.md` records
+that the production image should move to uvicorn with a worker count — the day that
+happens, a poll can land on a worker that has never heard of the job, and this needs an
+`import_jobs` table instead. It is written down here because nothing else will say so.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import shutil
+import uuid
+from dataclasses import dataclass, field
+from enum import StrEnum
+from pathlib import Path
+
+from chitai.config import settings
+from chitai.database.models import Library
+from chitai.services.book import (
+    BookService,
+    CalibreImportProgress,
+    CalibreImportResult,
+    UnimportedBook,
+)
+from chitai.services.calibre import CalibreLibrary
+
+
+class ImportState(StrEnum):
+    """Where a job has got to."""
+
+    RUNNING = "running"
+    FINISHED = "finished"
+
+    # Stopped on request. What it imported is complete.
+    CANCELLED = "cancelled"
+
+    # The run itself broke — an unreadable catalogue, a missing library. Distinct from
+    # individual books failing, which `failures` carries and which never stops the run.
+    FAILED = "failed"
+
+
+@dataclass
+class ImportJob:
+    """One import, running or finished."""
+
+    id: str
+    library_id: int
+    source: str
+
+    state: ImportState = ImportState.RUNNING
+    total: int = 0
+    processed: int = 0
+    created: int = 0
+    skipped: int = 0
+    failed: int = 0
+
+    current_title: str | None = None
+    failures: list[UnimportedBook] = field(default_factory=list)
+
+    # Books imported that look like something the library already had. A count, not the
+    # records: the duplicates screen is what shows them, and a big import would make
+    # this the largest thing in the response for no benefit.
+    possible_duplicates: int = 0
+
+    # Why the whole run stopped, when `state` is FAILED.
+    error: str | None = None
+
+    # The directory the job owns and must delete when it ends: what the uploaded archive
+    # was unpacked into.
+    workspace: Path | None = None
+
+    _stop: bool = False
+
+    @property
+    def finished(self) -> bool:
+        return self.state is not ImportState.RUNNING
+
+    def absorb(self, result: CalibreImportResult) -> None:
+        """Take the final counts from a finished run."""
+        self.total = result.total
+        self.created = len(result.created)
+        self.skipped = len(result.skipped)
+        self.failed = len(result.failed)
+        self.failures = list(result.failed)
+        self.possible_duplicates = len(result.possible_duplicates)
+        self.current_title = None
+
+        self.state = ImportState.CANCELLED if result.stopped else ImportState.FINISHED
+
+
+class CalibreImportRegistry:
+    """
+    The imports this process knows about.
+
+    One instance, held at module scope below. Jobs are kept after they finish so the
+    screen that started one can still read its result; nothing evicts them, which is
+    fine for a handful of one-time migrations and is the other reason a table would be
+    the answer if this ever needed to be durable.
+    """
+
+    def __init__(self) -> None:
+        self._jobs: dict[str, ImportJob] = {}
+
+        # Held only to keep the tasks referenced. Without this the event loop is free to
+        # garbage-collect a running task mid-import.
+        self._tasks: set[asyncio.Task] = set()
+
+    def get(self, job_id: str) -> ImportJob | None:
+        return self._jobs.get(job_id)
+
+    def running_for(self, library_id: int) -> ImportJob | None:
+        """The unfinished import for a library, if it has one."""
+        return next(
+            (
+                job
+                for job in self._jobs.values()
+                if job.library_id == library_id and not job.finished
+            ),
+            None,
+        )
+
+    def cancel(self, job_id: str) -> ImportJob | None:
+        """
+        Ask a job to stop after the book it is on.
+
+        Not `task.cancel()`: that would abandon a book mid-copy, leaving files on disk
+        with no row describing them. The flag is read between books.
+        """
+        job = self._jobs.get(job_id)
+
+        if job is not None and not job.finished:
+            job._stop = True
+
+        return job
+
+    def start(
+        self,
+        library: Library,
+        source: Path,
+        workspace: Path,
+        label: str,
+        allow_duplicates: bool = False,
+    ) -> ImportJob:
+        """
+        Begin importing, and return the handle to poll.
+
+        Args:
+            library: The library to import into.
+            source: The unpacked Calibre library's directory.
+            workspace: A directory the job owns and deletes when it ends — what the
+                uploaded archive was unpacked into. The books have been copied into the
+                library by then, so nothing is lost with it.
+            label: What to report as the source. `source` is a temp directory that would
+                mean nothing to the reader, so this is the archive's own name.
+            allow_duplicates: Import books whose files are already stored.
+
+        Returns:
+            The job, already running.
+        """
+        job = ImportJob(
+            id=str(uuid.uuid4()),
+            library_id=library.id,
+            source=label,
+            workspace=workspace,
+        )
+        self._jobs[job.id] = job
+
+        task = asyncio.create_task(self._run(job, library.id, source, allow_duplicates))
+        self._tasks.add(task)
+        task.add_done_callback(self._tasks.discard)
+
+        return job
+
+    async def _run(
+        self, job: ImportJob, library_id: int, source: Path, allow_duplicates: bool
+    ) -> None:
+        """
+        Do the import, recording everything on the job.
+
+        Opens a **session of its own**: the request that started this is long gone, and
+        its session was closed with it.
+        """
+        from chitai.services.library import LibraryService
+
+        def on_progress(progress: CalibreImportProgress) -> None:
+            job.total = progress.total
+            job.processed = progress.processed
+            job.current_title = progress.title
+
+            if progress.outcome == "created":
+                job.created += 1
+            elif progress.outcome == "skipped":
+                job.skipped += 1
+            else:
+                job.failed += 1
+
+        catalogue = CalibreLibrary(source)
+
+        try:
+            await catalogue.open()
+
+            async with settings.alchemy_config.get_session() as session:
+                library = await LibraryService(session=session).get(library_id)
+
+                result = await BookService(session=session).create_many_from_calibre(
+                    catalogue,
+                    library,
+                    allow_duplicates=allow_duplicates,
+                    on_progress=on_progress,
+                    should_stop=lambda: job._stop,
+                )
+
+            job.absorb(result)
+        except Exception as exc:
+            job.state = ImportState.FAILED
+            job.error = f"{type(exc).__name__}: {exc}"
+        finally:
+            await catalogue.close()
+
+            # An unpacked archive is a second copy of the whole library, and the books
+            # worth keeping have been copied into the library proper by now. Removed
+            # even when the run failed — especially then, since nothing will come back
+            # for it.
+            if job.workspace is not None:
+                await asyncio.to_thread(shutil.rmtree, job.workspace, True)
+
+
+registry = CalibreImportRegistry()

backend/src/chitai/services/consume.py

@@ -1,6 +1,7 @@
 import asyncio
 from pathlib import Path
 from collections import defaultdict
+from chitai.config import settings
 from chitai.database.models.library import Library
 from chitai.services import BookService, LibraryService
 from chitai.services.metadata_extractor import Extractor
@@ -111,13 +112,32 @@ class ConsumeDirectoryWatcher:
         """Process a batch of files."""
         try:
 
-            books = await self.book_service.create_many_from_existing_files(
+            result = await self.book_service.create_many_from_existing_files(
                 list(file_paths),
                 self.watch_path / Path(library_slug),
                 library=await self._get_library(library_slug),
             )
 
-            print(f"Created {len(books)} books!")
+            print(f"Created {len(result.books)} books!")
+
+            if result.duplicates:
+                print(
+                    f"Moved {len(result.duplicates)} already-stored file(s) "
+                    f"to {settings.duplicate_path}"
+                )
+
+            # Imported all the same — a metadata match is a guess, and there is nobody
+            # here to ask. The library's duplicates screen is where these get decided.
+            for possible in result.possible_duplicates:
+                names = ", ".join(
+                    f"{candidate.title} (#{candidate.book_id})"
+                    for candidate in possible.candidates
+                )
+                print(
+                    f"Imported {possible.title!r} (#{possible.book_id}), which may "
+                    f"already be in the library as: {names}"
+                )
+
 
         except Exception as e:
             print(f"Error processing batch: {e}")

backend/src/chitai/services/dependencies.py

@@ -127,6 +127,31 @@ def create_book_filter_dependencies(
     # Get base filters first
     filters = create_filter_dependencies(config, dep_defaults)
 
+    # OVERRIDE: id filter typed by the configured id type, not always `str`
+    #
+    # advanced_alchemy's `provide_id_filter` annotates `ids` as `list[str]` and
+    # ignores `config["id_filter"]` entirely, so `?ids=12` reaches the database as
+    # the string "12" and Postgres refuses `bigint = character varying`. Nothing
+    # called `?ids=` until the duplicates screen needed to fetch a handful of books
+    # by id, which is why it went unnoticed.
+    if id_type := config.get("id_filter"):
+        id_field = config.get("id_field", "id")
+
+        def provide_typed_id_filter(
+            ids=Parameter(query="ids", default=None, required=False),
+        ) -> CollectionFilter:
+            return CollectionFilter(field_name=id_field, values=ids)
+
+        # Attached as a type object rather than written as an annotation: this module
+        # has `from __future__ import annotations`, so a written one is stored as the
+        # string "Optional[list[id_type]]" and resolved against module globals, where
+        # a local named `id_type` does not exist.
+        provide_typed_id_filter.__annotations__["ids"] = Optional[list[id_type]]
+
+        filters[dep_defaults.ID_FILTER_DEPENDENCY_KEY] = Provide(
+            provide_typed_id_filter, sync_to_thread=False
+        )
+
     # OVERRIDE: Custom search filter with trigram search
     if config.get("search"):
         search_fields = config.get("search")

backend/src/chitai/services/filesystem_library.py

@@ -16,6 +16,43 @@ import chitai.database.models as m
 #   - Auto-handle missing values (e.g., skip {series}/ if series is empty)
 
 
+# Characters that cannot survive being interpolated into a path. The forward slash is
+# the one that matters: titles legitimately contain it — "AC/DC", "Him/Her" — and the
+# template writes the title straight into a directory name, so an unsanitised one
+# silently adds a level and puts the book somewhere `book.path` does not describe.
+# Calibre strips these from its own on-disk names and keeps the real title in its
+# database, which is how an import surfaces them.
+_UNSAFE_IN_PATH = re.compile(r"[/\\\x00-\x1f]")
+
+
+def sanitize_path_component(value: str) -> str:
+    """Make one metadata value safe to use as a single directory or file name."""
+    return _UNSAFE_IN_PATH.sub("_", value).strip()
+
+
+def _safe_components(book_data: dict) -> dict:
+    """
+    A shallow copy of the metadata with the values a path is built from sanitised.
+
+    Only strings are touched, and only the fields the default template interpolates. A
+    caller's own template can reach anything else in the dict, which is a reason to keep
+    this conservative rather than to walk the whole structure.
+    """
+    safe = dict(book_data)
+
+    for key in ("title", "series", "series_position"):
+        if isinstance(safe.get(key), str):
+            safe[key] = sanitize_path_component(safe[key])
+
+    if isinstance(safe.get("authors"), list):
+        safe["authors"] = [
+            sanitize_path_component(author) if isinstance(author, str) else author
+            for author in safe["authors"]
+        ]
+
+    return safe
+
+
 default_path_template = """
 /{{book.authors[0] if book.authors else 'Unknown'}}
 {%- if book.series -%}
@@ -101,7 +138,12 @@ class BookPathGenerator:
 
         """
 
-        result = self.root_path / Path(self.path_template.render(book=book_data))
+        # Sanitised per value, never on the rendered result: the separators the template
+        # puts *between* author, series and title are the whole point of it, and only the
+        # values interpolated into it must not contribute any of their own.
+        result = self.root_path / Path(
+            self.path_template.render(book=_safe_components(book_data))
+        )
 
         # Clean up
         result = re.sub(r"/+", "/", str(result))  # Remove consecutive backslashes

backend/src/chitai/services/matching.py

@@ -0,0 +1,207 @@
+# src/chitai/services/matching.py
+
+"""
+Normalisation for book-level duplicate detection.
+
+Two copies of one book rarely agree on how it is written down. One says
+`The Metamorphosis`, the other `Metamorphosis`; one credits `Kafka, Franz`, the other
+`Franz Kafka`; one carries the ISBN-10 and the other the ISBN-13 of the same edition.
+These functions reduce each of those to a single key, so the comparison is an equality
+test the database can index rather than a similarity score nobody can explain.
+
+Everything here is pure: the keys are computed once and stored on the row (see
+`Book.normalized_title`, `Author.normalized_name`, `Identifier.normalized_value`), so
+no Postgres extension is needed at query time.
+"""
+
+from __future__ import annotations
+
+import re
+import unicodedata
+
+from chitai.services.utils import is_valid_isbn, isbn10_to_isbn13
+
+
+# Asides a title carries that say nothing about which book it is:
+# "Frankenstein (Illustrated)", "Dune [Deluxe]".
+_BRACKETED = re.compile(r"[(\[{][^)\]}]*[)\]}]")
+
+# Edition and format qualifiers, matched only as a *trailing* run of words. Anchoring
+# to the end is what keeps "The Illustrated Man" a book and "Moby Dick Illustrated" a
+# format note — a qualifier trails the title, it is never the thing the title is about.
+_EDITION_NOISE = re.compile(
+    r"\s+(?:"
+    # "2nd edition", but also the compact forms publishers actually print on a
+    # cover: "2E", "3 Ed", "5e". The number alone is never enough — "Catch 22" is
+    # a title and must survive.
+    r"\d+(?:st|nd|rd|th)?\s*(?:edition|edn|ed|e)"
+    r"|(?:first|second|third|fourth|fifth|sixth|new|revised|expanded|updated|"
+    r"annotated|illustrated|unabridged|abridged|complete|definitive|deluxe|"
+    r"anniversary|collectors|international|kindle|paperback|hardcover|hardback|"
+    r"ebook|audiobook)"
+    r"(?:\s+(?:and|&)\s+\w+)*"
+    r"(?:\s+ed(?:ition|n)?)?"
+    r")$"
+)
+
+_LEADING_ARTICLE = re.compile(r"^(?:the|a|an)\s+")
+
+# Anything that is not a letter, a digit or a space, once accents are gone.
+_PUNCTUATION = re.compile(r"[^0-9a-z ]+")
+
+_WHITESPACE = re.compile(r"\s+")
+
+# Junk an extractor leaves on the end of a name: the `;` from a `DC:creator` list, the
+# `.epub` from a filename the author's name was read out of.
+_TRAILING_SEPARATORS = " ;,&/"
+_FILE_EXTENSION = re.compile(
+    r"\.(?:epub|pdf|mobi|azw3?|djvu|fb2|txt|cbz|cbr)$", re.IGNORECASE
+)
+
+# `J. R. R.` survives punctuation stripping as three one-letter words; `J.R.R.` as one.
+# Joining any run of them makes both `jrr`.
+_INITIAL_RUN = re.compile(r"\b(?:[a-z] )+[a-z]\b")
+
+# ISBNs are the same number under several names; everything else keeps its own.
+_ISBN_NAMES = {"isbn", "isbn-10", "isbn10", "isbn-13", "isbn13"}
+
+# Generated fresh for every build of a file, so two copies of one book never share one.
+# Matching on them would only re-find files the hash check already catches.
+_PER_BUILD_NAMES = {"uuid", "urn:uuid"}
+
+# Below this an identifier is not specific enough to be evidence: a Calibre `id` of
+# "42" would otherwise pair two unrelated books.
+_MIN_IDENTIFIER_LENGTH = 4
+
+
+def _fold(text: str) -> str:
+    """Casefolded, accent-free, punctuation-free, single-spaced."""
+    decomposed = unicodedata.normalize("NFKD", text.casefold())
+    unaccented = "".join(c for c in decomposed if not unicodedata.combining(c))
+
+    return _WHITESPACE.sub(" ", _PUNCTUATION.sub(" ", unaccented)).strip()
+
+
+def normalize_title(title: str | None) -> str:
+    """
+    Reduce a title to the key two copies of one book should share.
+
+    `Book.subtitle` is already split off by `Extractor.format_book_title`, so only what
+    is left in the title column is considered here.
+
+    Args:
+        title: The title as it was stored.
+
+    Returns:
+        The comparison key, or an empty string if nothing survives normalisation —
+        which is the signal not to match on the title at all.
+    """
+    if not title:
+        return ""
+
+    folded = _fold(_BRACKETED.sub(" ", title).replace("&", " and "))
+
+    # Repeated because qualifiers stack: "Dune Deluxe Edition Illustrated".
+    while (trimmed := _EDITION_NOISE.sub("", folded)) != folded:
+        folded = trimmed
+
+    # An article says nothing, but a title that is only an article is not improved by
+    # having none, and neither is one that noise removal emptied out.
+    return _LEADING_ARTICLE.sub("", folded, count=1) or folded
+
+
+def format_author_name(name: str | None) -> str:
+    """
+    Tidy an author's name into the one form the library writes them in.
+
+    Distinct from `normalize_author`, which throws away case, accents and spacing to
+    build a comparison key. This one is what a reader sees, so it keeps everything
+    that belongs to the name and only removes what an extractor added: a trailing
+    separator left over from a creator list, a file extension carried in from a
+    filename, and the `Surname, Given` ordering that EPUBs file names under.
+
+    Args:
+        name: The name as the file or filename gave it.
+
+    Returns:
+        The name to store, or an empty string if there is nothing left of it.
+    """
+    if not name:
+        return ""
+
+    tidied = _FILE_EXTENSION.sub("", name.strip().strip(_TRAILING_SEPARATORS).strip())
+
+    if tidied.count(",") == 1:
+        surname, given = (part.strip() for part in tidied.split(","))
+
+        # Only when the part before the comma is a single word. "Dave Thomas, Andy
+        # Hunt" is two people in one string, and flipping it would invent a third
+        # person who does not exist. Leaving an unrecognised form alone is the safe
+        # failure; rewriting it wrongly is not.
+        if surname and given and " " not in surname:
+            tidied = f"{given} {surname}"
+
+    return _WHITESPACE.sub(" ", tidied).strip()
+
+
+def normalize_author(name: str | None) -> str:
+    """
+    Reduce an author's name to the key their other books should share.
+
+    Deliberately not reduced to surname plus initial: that collides unrelated people,
+    and a wrong match here is a book pointed at a stranger's shelf.
+
+    Args:
+        name: The name as it was stored, in either `Franz Kafka` or `Kafka, Franz` form.
+
+    Returns:
+        The comparison key, or an empty string if nothing survives normalisation.
+    """
+    if not name:
+        return ""
+
+    # `Kafka, Franz` is one name written backwards. More than one comma is a list, or a
+    # suffix, and guessing at either does more harm than leaving it alone.
+    if name.count(",") == 1:
+        surname, forename = name.split(",")
+        name = f"{forename.strip()} {surname.strip()}"
+
+    folded = _fold(name)
+
+    return _INITIAL_RUN.sub(lambda run: run.group().replace(" ", ""), folded)
+
+
+def normalize_identifier(name: str, value: str) -> str | None:
+    """
+    Reduce one identifier to a `scheme:value` key, if it can carry a match at all.
+
+    Args:
+        name: What kind of identifier it is, as stored.
+        value: The identifier itself.
+
+    Returns:
+        The key, or None when the identifier is no use for matching: a per-build UUID,
+        something too short to be evidence, or an ISBN that fails its own checksum.
+    """
+    name = (name or "").strip().casefold()
+    value = (value or "").strip()
+
+    if not name or not value or name in _PER_BUILD_NAMES:
+        return None
+
+    if name in _ISBN_NAMES:
+        digits = re.sub(r"[^0-9Xx]", "", value).upper()
+
+        if not is_valid_isbn(digits):
+            return None
+
+        # One scheme for both forms: a publisher prints whichever it likes, and the
+        # ISBN-10 and ISBN-13 of an edition are the same number written twice.
+        isbn = digits if len(digits) == 13 else isbn10_to_isbn13(digits)
+        return f"isbn:{isbn}" if isbn else None
+
+    folded = _fold(value) or value.casefold()
+    if len(folded) < _MIN_IDENTIFIER_LENGTH:
+        return None
+
+    return f"{name}:{folded}"

backend/src/chitai/services/metadata_extractor.py

@@ -3,7 +3,6 @@
 # TODO: Code is a mess. Clean it up and add docstrings
 
 # Standard library
-from abc import ABC, abstractmethod
 import datetime
 from pathlib import Path
 from io import BytesIO
@@ -31,6 +30,147 @@ from chitai.services.utils import (
 logger = logging.getLogger(__name__)
 
 
+# Identifier schemes an EPUB can declare, mapped onto the names the rest of the app
+# uses. A scheme arrives either as an `opf:scheme` attribute or as a prefix on the
+# value itself (`urn:isbn:…`, `calibre:…`), and the two say the same thing.
+_IDENTIFIER_SCHEMES = {
+    "isbn": "isbn",
+    "isbn10": "isbn-10",
+    "isbn-10": "isbn-10",
+    "isbn13": "isbn-13",
+    "isbn-13": "isbn-13",
+    "uuid": "uuid",
+    "calibre": "calibre",
+    "doi": "doi",
+    "asin": "asin",
+    "amazon": "asin",
+    "mobi-asin": "asin",
+    "google": "google",
+    "goodreads": "goodreads",
+}
+
+# `scheme:rest`, with an optional `urn:` in front of it.
+_SCHEME_PREFIX = re.compile(r"^(?:urn:)?([A-Za-z][A-Za-z0-9.-]*):(.+)$")
+
+_UUID = re.compile(r"^[0-9a-f]{8}(?:-[0-9a-f]{4}){3}-[0-9a-f]{12}$", re.IGNORECASE)
+
+
+def parse_identifier(value: str, scheme: str | None = None) -> tuple[str, str] | None:
+    """
+    Work out what one raw identifier is, and what it is worth storing as.
+
+    EPUBs write the same ISBN as `9780486282114`, `978-0-486-28211-4` and
+    `urn:isbn:978-0-486-28211-4`, and carry plenty of identifiers that are not ISBNs
+    at all. Validating the string verbatim keeps only the first form and throws the
+    rest away, so normalise first and name whatever survives.
+
+    Args:
+        value: The identifier as the file wrote it.
+        scheme: What the file said it is, if it said anything — an `opf:scheme`
+            attribute. A prefix on the value takes precedence over this.
+
+    Returns:
+        The `(name, value)` to store, or `None` when there is nothing usable: an
+        empty value, or one declared to be an ISBN that fails its own checksum.
+    """
+    value = (value or "").strip()
+    if not value:
+        return None
+
+    name = _IDENTIFIER_SCHEMES.get((scheme or "").strip().casefold())
+
+    # An unrecognised prefix is part of the value rather than a scheme —
+    # "http://example.com/book" is not an identifier called "http".
+    if (match := _SCHEME_PREFIX.match(value)) and (
+        prefixed := _IDENTIFIER_SCHEMES.get(match.group(1).casefold())
+    ):
+        name = prefixed
+        value = match.group(2).strip()
+
+    if name is None or name.startswith("isbn"):
+        digits = re.sub(r"[^0-9Xx]", "", value).upper()
+        if is_valid_isbn(digits):
+            return f"isbn-{len(digits)}", digits
+
+        # Something that announced itself as an ISBN and is not one carries no
+        # information: storing it would link the reader to a page that does not exist.
+        if name is not None:
+            return None
+
+    return name or ("uuid" if _UUID.match(value) else "id"), value
+
+
+# Numbered editions, in the forms covers and catalogue records actually use:
+# "3rd Edition", "2E", "4e", "8_e", "/6e", "(2nd edition)", "Third International Edition".
+_ORDINAL_WORDS = {
+    "first": 1, "second": 2, "third": 3, "fourth": 4, "fifth": 5, "sixth": 6,
+    "seventh": 7, "eighth": 8, "ninth": 9, "tenth": 10, "eleventh": 11, "twelfth": 12,
+}
+
+# Words that sit between the number and "Edition" and belong to the same statement.
+_EDITION_QUALIFIER = (
+    r"(?:international|global|revised|updated|expanded|anniversary|deluxe|student|"
+    r"instructors?|annotated|illustrated|reprint)"
+)
+
+_EDITION = re.compile(
+    rf"""
+    [\s,;:/\-–—(\[]+                        # the separator the statement hangs off
+    (?:
+        (?P<num>\d{{1,2}})\s*(?:st|nd|rd|th)?[\s_]*
+            (?:{_EDITION_QUALIFIER}\s+)*(?:edition\b|edn\b|ed\b\.?|e\b)
+      | (?P<word>{"|".join(_ORDINAL_WORDS)})\s+
+            (?:{_EDITION_QUALIFIER}\s+)*(?:edition\b|edn\b|ed\b\.?)
+    )
+    [\s)\]]*                                # and its closing bracket, if it had one
+    """,
+    re.IGNORECASE | re.VERBOSE,
+)
+
+
+def split_edition(title: str | None) -> tuple[str | None, int | None]:
+    """
+    Separate a numbered edition statement from the title it is written into.
+
+    "Fluent Python, 2nd Edition" is one book with a field for the edition, not a
+    title. Left in place it also splits the library: the second edition never looks
+    like the first, and neither matches the copy whose file simply did not mention it.
+
+    The number is what makes this safe. Nothing is stripped without one, so
+    "Catch 22" and "Blade Runner 2049" keep their numbers and "Global Edition" —
+    which is a variant, not a numbered edition, and has nowhere to go in an
+    integer column — is left in the title where it can still be read.
+
+    Args:
+        title: The title as the file or filename gave it.
+
+    Returns:
+        The title without the edition statement, and the edition number. The title
+        unchanged and None when there is no numbered edition in it, or when removing
+        it would leave nothing behind.
+    """
+    if not title:
+        return title, None
+
+    if (match := _EDITION.search(title)) is None:
+        return title, None
+
+    edition = (
+        int(match["num"]) if match["num"] else _ORDINAL_WORDS[match["word"].casefold()]
+    )
+
+    stripped = _EDITION.sub(" ", title)
+    stripped = re.sub(r"\s{2,}", " ", stripped)
+    stripped = re.sub(r"\s+([,;:.!?])", r"\1", stripped)  # "Works : What" → "Works: What"
+    stripped = stripped.strip(" ,;:-–—/")
+
+    # A title that is only an edition statement is not improved by having none.
+    if not stripped:
+        return title, None
+
+    return stripped, edition
+
+
 class FileExtractor(Protocol):
     @classmethod
     async def extract_metadata(
@@ -54,15 +194,36 @@ class Extractor:
         # EPUB tends to give better metadata results over pdf
         sorted_files = sorted(files, key=lambda f: Extractor._get_file_priority(f))
 
+        # Identifiers accumulate across formats instead of replacing each other. Every
+        # other field is a single value where the later, better-trusted format simply
+        # wins, but identifiers are a *collection*: a book holding an EPUB and a PDF
+        # genuinely carries what both of them declare, and merging the dict wholesale
+        # threw away everything the earlier format found. An EPUB that declares an
+        # ASIN, a Google volume id and a Calibre id kept none of them once a PDF
+        # contributed a single ISBN.
+        identifiers: dict[str, str] = {}
+
         for file in sorted_files:
             match get_file_extension(file):
                 case "epub":
-                    metadata = metadata | await EpubExtractor.extract_metadata(file)
+                    extracted = await EpubExtractor.extract_metadata(file)
                 case "pdf":
-                    metadata = metadata | await PdfExtractor.extract_metadata(file)
+                    extracted = await PdfExtractor.extract_metadata(file)
                 case _:
                     break
 
+            # First writer wins per name, and the files are already ordered by how
+            # far their metadata can be trusted. A `dc:identifier` the publisher
+            # declared outranks an ISBN scraped out of a PDF's copyright page, which
+            # routinely prints the ISBNs of other formats and older editions too.
+            for name, value in (extracted.pop("identifiers", None) or {}).items():
+                identifiers.setdefault(name, value)
+
+            metadata = metadata | extracted
+
+        if identifiers:
+            metadata["identifiers"] = identifiers
+
         # Get metadata from file names
         for file in files:
             metadata = FilenameExtractor.extract_metadata(file) | metadata
@@ -75,7 +236,14 @@ class Extractor:
 
         # format the title
         if metadata.get('title', None):
-            title, subtitle = Extractor.format_book_title(metadata["title"])
+            # Before the subtitle split, so the edition cannot be mistaken for one:
+            # "How Linux Works, 3rd Edition: What Every Superuser Should Know" has to
+            # lose the edition first for the colon count to mean anything.
+            title, edition = split_edition(metadata["title"])
+            if edition is not None:
+                metadata.setdefault("edition", edition)
+
+            title, subtitle = Extractor.format_book_title(title)
             metadata["title"] = title
             metadata["subtitle"] = subtitle
 
@@ -236,7 +404,7 @@ class PdfExtractor(FileExtractor):
         try:
             return datetime.datetime.strptime(date_portion, "%Y%m%d").date()
 
-        except Exception as e:
+        except Exception:
             return None
 
     @classmethod
@@ -363,15 +531,23 @@ class EpubExtractor(FileExtractor):
 
     @classmethod
     def _extract_identifiers(cls, epub: epub.EpubBook) -> dict[str, str]:
+        """
+        Every `DC:identifier` the file carries, keyed by what kind of thing it is.
+
+        Non-ISBN identifiers are kept: `Identifier` is a free-form name/value pair, so
+        a Calibre id or an ASIN costs nothing to store and is one more thing two copies
+        of a book can be recognised by.
+        """
         identifiers = {}
 
-        for id in epub.get_metadata("DC", "identifier"):
-            if is_valid_isbn(id[0]):
-                if len(id[0]) == 13:
-                    identifiers.update({"isbn-13": id[0]})
+        for value, attributes in epub.get_metadata("DC", "identifier"):
+            scheme = None
+            if isinstance(attributes, dict):
+                scheme = attributes.get("opf:scheme") or attributes.get("scheme")
 
-                elif len(id[0]) == 10:
-                    identifiers.update({"isbn-10": id[0]})
+            if (parsed := parse_identifier(value, scheme)) is not None:
+                name, parsed_value = parsed
+                identifiers[name] = parsed_value
 
         return identifiers
 
@@ -380,7 +556,7 @@ class EpubExtractor(FileExtractor):
         try:
             return epub.get_metadata("DC", "description")[0][0]
 
-        except:
+        except Exception:
             return None
 
     @classmethod
@@ -389,15 +565,15 @@ class EpubExtractor(FileExtractor):
             date_str = epub.get_metadata("DC", "date")[0][0].split("T")[0]
             return datetime.date.fromisoformat(date_str)
 
-        except:
+        except Exception:
             return None
 
     @classmethod
     def _extract_publisher(cls, epub: epub.EpubBook) -> str | None:
         try:
-            epub.get_metadata("DC", "publisher")[0][0]
+            return epub.get_metadata("DC", "publisher")[0][0]
 
-        except:
+        except Exception:
             return None
 
     @classmethod
@@ -421,7 +597,7 @@ class EpubExtractor(FileExtractor):
                             cover_item = epub.get_item_with_id(cover_id)
                             if cover_item:
                                 return PIL.Image.open(BytesIO(cover_item.content))
-        except Exception as e:
+        except Exception:
             pass  # Fallback to next strategy
 
         # Strategy 2: Search image filenames for "cover" keyword
@@ -489,7 +665,11 @@ class FilenameExtractor(FileExtractor):
         elif isinstance(input, Path):
             filename = get_filename(input, ext=False)
         elif isinstance(input, UploadFile):
-            filename = Path(input.filename).name
+            # `.stem`, not `.name`: the extension is not part of the metadata, and
+            # this is the browser upload path, so keeping it is how a library fills
+            # up with authors called "Sam Newman.epub". The other two branches have
+            # always stripped it.
+            filename = Path(input.filename).stem
         else:
             raise ValueError("Input type not supported")
 

backend/src/chitai/services/opds/opds.py

@@ -47,8 +47,13 @@ def convert_book_to_entry(book: m.Book) -> Entry:
         link=[
             ImageLink(href=f"/{book.cover_image}", type="image/webp"),
             *[
+                # The only place a content type has to be a string: `Link.type` is
+                # required, and a null fails the whole feed rather than one entry.
+                # `application/octet-stream` is the registered way to say "opaque
+                # bytes", which is exactly what an unnamed format is.
                 AcquisitionLink(
-                    href=f"/opds/download/{book.id}/{file.id}", type=file.content_type
+                    href=f"/opds/download/{book.id}/{file.id}",
+                    type=file.content_type or "application/octet-stream",
                 )
                 for file in book.files
             ],

backend/src/chitai/services/utils.py

@@ -3,13 +3,12 @@
 # Standard library
 from __future__ import annotations
 
+import errno
 import hashlib
+import mimetypes
 from pathlib import Path
 import shutil
-from typing import TYPE_CHECKING, BinaryIO
-
-if TYPE_CHECKING:
-    from hashlib import _Hash
+from typing import BinaryIO
 
 # Third-party libraries
 import PIL
@@ -32,6 +31,9 @@ KO_STEP = 1024
 KO_SAMPLE_SIZE = 1024
 KO_INDICES = range(-1, 11)  # -1 to 10 inclusive
 
+# How much is read at a time while hashing.
+HASH_CHUNK_SIZE = 262144  # 256 KiB
+
 
 def _lshift32(val: int, shift: int) -> int:
     """
@@ -100,10 +102,9 @@ async def calculate_koreader_hash(file_path: Path) -> str:
     offsets = _get_koreader_offsets()
 
     file_pos = 0
-    chunk_size = 262144  # 256 KiB
 
     async with aiofiles.open(file_path, "rb") as f:
-        while chunk := await f.read(chunk_size):
+        while chunk := await f.read(HASH_CHUNK_SIZE):
             _partial_md5_from_chunk(chunk, hasher, offsets, file_pos)
             file_pos += len(chunk)
 
@@ -132,6 +133,49 @@ class StreamingHasher:
         """Return the final hash."""
         return self.hasher.hexdigest()
 
+    @property
+    def size(self) -> int:
+        """Total number of bytes fed in so far."""
+        return self.position
+
+
+async def fingerprint_upload(file: UploadFile) -> tuple[str, int]:
+    """
+    Calculate the hash and byte size of an uploaded file without storing it.
+
+    Duplicate detection has to answer before anything is written to the library, so
+    the file is read here and rewound for whoever writes it afterwards.
+
+    Args:
+        file: The uploaded file to read.
+
+    Returns:
+        The file's `(hash, size)` pair.
+    """
+    hasher = StreamingHasher()
+
+    await file.seek(0)
+    while chunk := await file.read(HASH_CHUNK_SIZE):
+        hasher.update(chunk)
+    await file.seek(0)
+
+    return hasher.hexdigest(), hasher.size
+
+
+async def fingerprint_file(file_path: Path) -> tuple[str, int]:
+    """
+    Calculate the hash and byte size of a file already on disk.
+
+    Args:
+        file_path: Path to the file to read.
+
+    Returns:
+        The file's `(hash, size)` pair.
+    """
+    stats = await aios.stat(file_path)
+    return await calculate_koreader_hash(file_path), stats.st_size
+
+
 ##################################
 #  Filesystem related utilities  #
 ##################################
@@ -184,7 +228,39 @@ async def move_file(src_path: Path, dest_path: Path, create_dirs=True) -> None:
         if dest_dir:  # Only create if there's a directory path
             await aios.makedirs(dest_dir, exist_ok=True)
 
+    try:
         await aios.rename(src_path, dest_path)
+    except OSError as exc:
+        if exc.errno != errno.EXDEV:
+            raise
+
+        # Source and destination are on different filesystems, which rename cannot
+        # cross. Libraries, the consume directory and the duplicates directory are
+        # all configured separately, so they can easily be separate mounts.
+        shutil.move(str(src_path), str(dest_path))
+
+async def copy_file(src_path: Path, dest_path: Path, create_dirs: bool = True) -> None:
+    """
+    Copy a file, streaming it rather than reading it whole.
+
+    `shutil.copy` would block the event loop for as long as the read takes, which for a
+    40 MB ebook — and a few thousand of them in a row — is not acceptable.
+
+    Args:
+        src_path: The file to copy. Left exactly as it is.
+        dest_path: Where the copy goes.
+        create_dirs: Create the destination's parent directories first.
+    """
+    if create_dirs and dest_path.parent:
+        await aios.makedirs(dest_path.parent, exist_ok=True)
+
+    async with (
+        aiofiles.open(src_path, "rb") as source,
+        aiofiles.open(dest_path, "wb") as destination,
+    ):
+        while chunk := await source.read(HASH_CHUNK_SIZE):
+            await destination.write(chunk)
+
 
 async def move_dir_contents(source_dir: Path | str, target_dir: Path | str) -> None:
     """
@@ -407,6 +483,66 @@ def get_filename(file: Path | str, ext: bool = True) -> str:
     return filename.stem
 
 
+# Content types for the ebook formats `mimetypes` does not know. Python's built-in map
+# covers `.epub`, `.pdf`, `.azw3`, `.cbz`, `.cbr` and `.djvu`, and answers `None` for
+# every format below — so a library imported from elsewhere, which is where MOBI and
+# AZW files come from, stores nothing for them.
+#
+# That matters downstream because an OPDS acquisition link is how a reader app decides
+# whether it can open a file at all.
+
+# What a client sends when it does not know either. Treated as an absence rather than
+# an answer: storing it would be indistinguishable from having determined a format, and
+# it is the value browsers post for every extension they do not recognise.
+_UNSPECIFIED = "application/octet-stream"
+
+EBOOK_CONTENT_TYPES = {
+    "mobi": "application/x-mobipocket-ebook",
+    "prc": "application/x-mobipocket-ebook",
+    "azw": "application/vnd.amazon.ebook",
+    "fb2": "application/x-fictionbook+xml",
+    "fbz": "application/x-zip-compressed-fb2",
+    "lit": "application/x-ms-reader",
+    "lrf": "application/x-sony-bbeb",
+    "cb7": "application/x-cb7",
+}
+
+
+def guess_content_type(
+    file: Path | str | UploadFile, fallback: str | None = None
+) -> str | None:
+    """
+    Name a file's format from its extension.
+
+    The extension is trusted ahead of anything a client said: a browser posts
+    `application/octet-stream` for every format it does not recognise, which is most
+    ebook formats, and that answer is worth less than the `.mobi` on the end of the
+    name.
+
+    Args:
+        file: The file to name, as a path or an upload.
+        fallback: What to use when neither table knows the extension — a client-supplied
+            content type, if there is one. `application/octet-stream` is discarded: it
+            is the client saying it does not know, which is not information.
+
+    Returns:
+        The content type, or None when nothing can name it. Null is the honest answer
+        and the column is nullable: a caller that structurally needs a string should
+        substitute one where it needs it, rather than have an invented value stored.
+    """
+    extension = get_file_extension(file)
+
+    if known := EBOOK_CONTENT_TYPES.get(extension):
+        return known
+
+    guessed, _ = mimetypes.guess_type(get_filename(file))
+
+    if fallback == _UNSPECIFIED:
+        fallback = None
+
+    return guessed or fallback
+
+
 ###############################
 #  ISBN Validation utilities  #
 ###############################
@@ -432,7 +568,7 @@ def is_valid_isbn(isbn: str) -> bool:
             return is_valid_isbn13(isbn)
         else:
             return False
-    except:
+    except Exception:
         return False
 
 
@@ -457,6 +593,29 @@ def is_valid_isbn10(isbn: str) -> bool:
     return str(check_digit) == isbn[-1] or (check_digit == 10 and isbn[-1] in "Xx")
 
 
+def isbn10_to_isbn13(isbn: str) -> str | None:
+    """
+    Convert an ISBN-10 to the ISBN-13 naming the same edition.
+
+    The two are the same number written twice: prefix `978`, drop the ISBN-10 check
+    digit, recompute the check digit under the ISBN-13 rule. Matching only works if
+    both forms collapse onto one, since a publisher may print either.
+
+    Args:
+        isbn: A 10-character ISBN, digits and an optional trailing `X` only.
+
+    Returns:
+        The equivalent ISBN-13, or None if the input is not a valid ISBN-10.
+    """
+    if not is_valid_isbn(isbn) or len(isbn) != 10:
+        return None
+
+    digits = f"978{isbn[:9]}"
+    total = sum(int(digit) * (1 if i % 2 == 0 else 3) for i, digit in enumerate(digits))
+
+    return f"{digits}{(10 - total % 10) % 10}"
+
+
 def is_valid_isbn13(isbn: str) -> bool:
     """
     Validate an ISBN-13 number using its check digit.

backend/tests/calibre_fixtures.py

@@ -0,0 +1,282 @@
+"""
+Build a Calibre library on disk, for tests to read.
+
+Generated rather than committed as a binary `metadata.db`, because the rows worth
+testing are the awkward ones — the year-101 pubdate, a `|` in an author name, a REAL
+series index, HTML in a comment — and those are clearer written out in Python than
+hidden inside a blob.
+
+The schema below is Calibre's own, copied from a real library's `sqlite_master`, reduced
+to the tables the reader touches. `books_pages_link` is created separately by
+`add_pages`: it is recent, and a library made by an older Calibre will not have it.
+"""
+
+from __future__ import annotations
+
+import shutil
+import sqlite3
+from pathlib import Path
+
+
+SCHEMA = """
+CREATE TABLE books (
+    id INTEGER PRIMARY KEY AUTOINCREMENT,
+    title TEXT NOT NULL DEFAULT 'Unknown' COLLATE NOCASE,
+    sort TEXT COLLATE NOCASE,
+    timestamp TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+    pubdate TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+    series_index REAL NOT NULL DEFAULT 1.0,
+    author_sort TEXT COLLATE NOCASE,
+    path TEXT NOT NULL DEFAULT '',
+    uuid TEXT,
+    has_cover BOOL DEFAULT 0,
+    last_modified TIMESTAMP NOT NULL DEFAULT '2000-01-01 00:00:00+00:00'
+);
+CREATE TABLE authors (
+    id INTEGER PRIMARY KEY, name TEXT NOT NULL COLLATE NOCASE,
+    sort TEXT COLLATE NOCASE, link TEXT NOT NULL DEFAULT '', UNIQUE(name)
+);
+CREATE TABLE books_authors_link (
+    id INTEGER PRIMARY KEY, book INTEGER NOT NULL, author INTEGER NOT NULL,
+    UNIQUE(book, author)
+);
+CREATE TABLE publishers (
+    id INTEGER PRIMARY KEY, name TEXT NOT NULL COLLATE NOCASE,
+    sort TEXT COLLATE NOCASE, link TEXT NOT NULL DEFAULT '', UNIQUE(name)
+);
+CREATE TABLE books_publishers_link (
+    id INTEGER PRIMARY KEY, book INTEGER NOT NULL, publisher INTEGER NOT NULL,
+    UNIQUE(book)
+);
+CREATE TABLE tags (
+    id INTEGER PRIMARY KEY, name TEXT NOT NULL COLLATE NOCASE,
+    link TEXT NOT NULL DEFAULT '', UNIQUE (name)
+);
+CREATE TABLE books_tags_link (
+    id INTEGER PRIMARY KEY, book INTEGER NOT NULL, tag INTEGER NOT NULL,
+    UNIQUE(book, tag)
+);
+CREATE TABLE series (
+    id INTEGER PRIMARY KEY, name TEXT NOT NULL COLLATE NOCASE,
+    sort TEXT COLLATE NOCASE, link TEXT NOT NULL DEFAULT '', UNIQUE (name)
+);
+CREATE TABLE books_series_link (
+    id INTEGER PRIMARY KEY, book INTEGER NOT NULL, series INTEGER NOT NULL,
+    UNIQUE(book)
+);
+CREATE TABLE languages (
+    id INTEGER PRIMARY KEY, lang_code TEXT NOT NULL COLLATE NOCASE,
+    link TEXT NOT NULL DEFAULT '', UNIQUE(lang_code)
+);
+CREATE TABLE books_languages_link (
+    id INTEGER PRIMARY KEY, book INTEGER NOT NULL, lang_code INTEGER NOT NULL,
+    item_order INTEGER NOT NULL DEFAULT 0, UNIQUE(book, lang_code)
+);
+CREATE TABLE comments (
+    id INTEGER PRIMARY KEY, book INTEGER NOT NULL,
+    text TEXT NOT NULL COLLATE NOCASE, UNIQUE(book)
+);
+CREATE TABLE identifiers (
+    id INTEGER PRIMARY KEY, book INTEGER NOT NULL,
+    type TEXT NOT NULL DEFAULT 'isbn' COLLATE NOCASE,
+    val TEXT NOT NULL COLLATE NOCASE, UNIQUE(book, type)
+);
+CREATE TABLE data (
+    id INTEGER PRIMARY KEY, book INTEGER NOT NULL,
+    format TEXT NOT NULL COLLATE NOCASE, uncompressed_size INTEGER NOT NULL,
+    name TEXT NOT NULL, UNIQUE(book, format)
+);
+"""
+
+# Calibre's own "no date". Stored, never null, and a valid date — which is exactly why
+# it has to be recognised rather than parsed.
+UNDEFINED_DATE = "0101-01-01 00:00:00+00:00"
+
+# What a `cover.jpg` that PIL cannot read looks like. Real libraries hold these, from
+# an interrupted download or a failed conversion.
+CORRUPT_COVER = b"\xff\xd8\xff\xe0 not really a jpeg"
+
+
+def write_cover(path: Path) -> None:
+    """
+    Write a real, readable JPEG.
+
+    Generated with PIL rather than embedded as a hex blob: a hand-rolled JPEG that is
+    subtly malformed fails inside the import as an unrelated error, which is exactly the
+    confusion this avoids.
+    """
+    from PIL import Image
+
+    Image.new("RGB", (2, 3), (10, 20, 30)).save(path, "JPEG")
+
+
+class CalibreFixture:
+    """A Calibre library being assembled under `root`."""
+
+    def __init__(self, root: Path) -> None:
+        self.root = root
+        self.root.mkdir(parents=True, exist_ok=True)
+
+        self.connection = sqlite3.connect(self.root / "metadata.db")
+        self.connection.executescript(SCHEMA)
+
+    def add_pages_table(self) -> None:
+        """Add `books_pages_link`, which only a recent Calibre creates."""
+        self.connection.executescript(
+            """
+            CREATE TABLE books_pages_link (
+                book INTEGER PRIMARY KEY,
+                pages INTEGER DEFAULT 0 NOT NULL,
+                algorithm INTEGER DEFAULT 0 NOT NULL,
+                format TEXT DEFAULT '' NOT NULL COLLATE NOCASE,
+                format_size INTEGER DEFAULT 0 NOT NULL,
+                timestamp TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
+                needs_scan INTEGER NOT NULL DEFAULT 0
+            );
+            """
+        )
+
+    def add_book(
+        self,
+        book_id: int,
+        title: str,
+        *,
+        authors: list[str] | None = None,
+        pubdate: str = UNDEFINED_DATE,
+        series: str | None = None,
+        series_index: float = 1.0,
+        tags: list[str] | None = None,
+        publisher: str | None = None,
+        languages: list[str] | None = None,
+        comment: str | None = None,
+        identifiers: dict[str, str] | None = None,
+        uuid: str | None = None,
+        pages: int | None = None,
+        cover: bool = False,
+        corrupt_cover: bool = False,
+        formats: dict[str, Path] | None = None,
+        directory: str | None = None,
+    ) -> Path:
+        """
+        Add one book, with its files laid out the way Calibre lays them out.
+
+        Args:
+            formats: Format name (`EPUB`) to a real file to copy in. Its on-disk stem is
+                Calibre's, not the title — that is the point of the `data` table.
+            directory: Override the `books.path` value, for testing a row whose
+                directory is not where the convention would put it.
+
+        Returns:
+            The book's directory.
+        """
+        author_names = authors or ["Unknown"]
+        relative = directory or f"{author_names[0]}/{title} ({book_id})"
+        book_directory = self.root / relative
+        book_directory.mkdir(parents=True, exist_ok=True)
+
+        self.connection.execute(
+            "INSERT INTO books (id, title, pubdate, series_index, path, uuid, has_cover) "
+            "VALUES (?, ?, ?, ?, ?, ?, ?)",
+            (
+                book_id,
+                title,
+                pubdate,
+                series_index,
+                relative,
+                uuid or f"uuid-{book_id}",
+                int(cover or corrupt_cover),
+            ),
+        )
+
+        for name in author_names:
+            self._link("authors", "books_authors_link", "author", book_id, name)
+
+        for name in tags or []:
+            self._link("tags", "books_tags_link", "tag", book_id, name)
+
+        if series:
+            self._link("series", "books_series_link", "series", book_id, series)
+
+        if publisher:
+            self._link(
+                "publishers", "books_publishers_link", "publisher", book_id, publisher
+            )
+
+        for order, code in enumerate(languages or []):
+            language_id = self._lookup("languages", "lang_code", code)
+            self.connection.execute(
+                "INSERT INTO books_languages_link (book, lang_code, item_order) "
+                "VALUES (?, ?, ?)",
+                (book_id, language_id, order),
+            )
+
+        if comment is not None:
+            self.connection.execute(
+                "INSERT INTO comments (book, text) VALUES (?, ?)", (book_id, comment)
+            )
+
+        for name, value in (identifiers or {}).items():
+            self.connection.execute(
+                "INSERT INTO identifiers (book, type, val) VALUES (?, ?, ?)",
+                (book_id, name, value),
+            )
+
+        if pages is not None:
+            self.connection.execute(
+                "INSERT INTO books_pages_link (book, pages) VALUES (?, ?)",
+                (book_id, pages),
+            )
+
+        if corrupt_cover:
+            (book_directory / "cover.jpg").write_bytes(CORRUPT_COVER)
+        elif cover:
+            write_cover(book_directory / "cover.jpg")
+
+        for format, origin in (formats or {}).items():
+            # Calibre's on-disk stem: sanitised, truncated, and not the title.
+            stem = f"{title[:40]} - {author_names[0]}".replace(":", "_")
+            destination = book_directory / f"{stem}.{format.lower()}"
+            shutil.copy(origin, destination)
+
+            self.connection.execute(
+                "INSERT INTO data (book, format, uncompressed_size, name) "
+                "VALUES (?, ?, ?, ?)",
+                (book_id, format, destination.stat().st_size, stem),
+            )
+
+        return book_directory
+
+    def add_missing_format(self, book_id: int, format: str, stem: str) -> None:
+        """Record a file in the catalogue without putting one on disk."""
+        self.connection.execute(
+            "INSERT INTO data (book, format, uncompressed_size, name) VALUES (?, ?, ?, ?)",
+            (book_id, format, 1234, stem),
+        )
+
+    def _link(
+        self, table: str, link_table: str, column: str, book_id: int, name: str
+    ) -> None:
+        item_id = self._lookup(table, "name", name)
+        self.connection.execute(
+            f"INSERT INTO {link_table} (book, {column}) VALUES (?, ?)",
+            (book_id, item_id),
+        )
+
+    def _lookup(self, table: str, column: str, value: str) -> int:
+        row = self.connection.execute(
+            f"SELECT id FROM {table} WHERE {column} = ?", (value,)
+        ).fetchone()
+
+        if row:
+            return int(row[0])
+
+        cursor = self.connection.execute(
+            f"INSERT INTO {table} ({column}) VALUES (?)", (value,)
+        )
+        return int(cursor.lastrowid or 0)
+
+    def commit(self) -> Path:
+        """Finish writing and return the library root."""
+        self.connection.commit()
+        self.connection.close()
+        return self.root

backend/tests/conftest.py

@@ -40,7 +40,13 @@ pytest_plugins = [
 
 @pytest.fixture(autouse=True)
 def _patch_settings(tmp_path: Path, monkeypatch: pytest.MonkeyPatch) -> None:
-    monkeypatch.setattr(settings, "book_cover_path", f"{tmp_path}/covers")
+    # PIL will not create the directory it is asked to save into, so anything that
+    # imports a file carrying a cover needs it to exist first.
+    covers = tmp_path / "covers"
+    covers.mkdir()
+
+    monkeypatch.setattr(settings, "book_cover_path", str(covers))
+    monkeypatch.setattr(settings, "duplicate_path", str(tmp_path / "duplicates"))
 
 
 @pytest.fixture(name="engine")

backend/tests/integration/test_book.py

@@ -37,7 +37,8 @@ from pathlib import Path
         (
             Path("tests/data_files/Calculus Made Easy - Silvanus Thompson.pdf"),
             2,
-            "The Project Gutenberg eBook #33283: Calculus Made Easy, 2nd Edition",
+            # The ", 2nd Edition" is split off into `edition`, not kept in the title.
+            "The Project Gutenberg eBook #33283: Calculus Made Easy",
             ["Silvanus Phillips Thompson"],
         ),
     ],
@@ -208,6 +209,21 @@ async def test_get_book_file(
     assert downloaded_content == file_content
 
 
+async def test_list_books_by_id(populated_authenticated_client: AsyncClient) -> None:
+    """
+    `?ids=` has to reach the database as integers.
+
+    advanced_alchemy's stock id filter annotates the parameter as `list[str]` whatever
+    the configured id type, so the ids arrived as strings and Postgres refused to
+    compare a bigint primary key against them. Nothing called it until a screen needed
+    to fetch a handful of books by id.
+    """
+    response = await populated_authenticated_client.get("/books?ids=1&ids=2&pageSize=10")
+
+    assert response.status_code == 200
+    assert sorted(book["id"] for book in response.json()["items"]) == [1, 2]
+
+
 async def test_get_book_by_id(populated_authenticated_client: AsyncClient) -> None:
     """Test retrieving a specific book by ID."""
 
@@ -366,7 +382,8 @@ async def test_create_multiple_books_from_directory(
 
     assert response.status_code == 201
     data = response.json()
-    assert len(data.get("items") or data.get("data")) >= 1
+    assert len(data["created"]) == 2
+    assert data["skipped"] == []
 
 
 async def test_create_books_from_parent_directory_keeps_embedded_title(
@@ -396,7 +413,7 @@ async def test_create_books_from_parent_directory_keeps_embedded_title(
 
     assert response.status_code == 201
 
-    books = response.json()["items"]
+    books = response.json()["created"]
     assert len(books) == 1
     assert books[0]["title"] == "Metamorphosis"
 
@@ -419,11 +436,336 @@ async def test_create_books_groups_formats_within_one_folder(
 
     assert response.status_code == 201
 
-    books = response.json()["items"]
+    books = response.json()["created"]
     assert len(books) == 1
     assert len(books[0]["files"]) == 2
 
 
+class TestDuplicateHandling:
+    """A file the library already holds must not be stored a second time."""
+
+    epub_path = Path("tests/data_files/Metamorphosis - Franz Kafka.epub")
+
+    def upload(self, name: str | None = None) -> list[tuple[str, tuple]]:
+        return [
+            (
+                "files",
+                (
+                    name or self.epub_path.name,
+                    self.epub_path.read_bytes(),
+                    "application/epub+zip",
+                ),
+            )
+        ]
+
+    async def test_bulk_upload_reports_skipped_files(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        """Re-dropping a folder must import what is new and name what was not."""
+        first = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload()
+        )
+        assert first.status_code == 201
+        created = first.json()["created"][0]
+
+        second = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload()
+        )
+
+        assert second.status_code == 201
+        result = second.json()
+        assert result["created"] == []
+        assert len(result["skipped"]) == 1
+
+        skipped = result["skipped"][0]
+        assert skipped["filename"] == self.epub_path.name
+        assert skipped["book_id"] == created["id"]
+        assert skipped["book_title"] == created["title"]
+
+    async def test_bulk_upload_can_be_forced(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload()
+        )
+
+        response = await authenticated_client.post(
+            "/books/fromFiles?library_id=1&allow_duplicates=true", files=self.upload()
+        )
+
+        assert response.status_code == 201
+        assert len(response.json()["created"]) == 1
+        assert response.json()["skipped"] == []
+
+    async def test_single_book_create_conflicts(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        """Naming files deliberately earns a refusal rather than a silent drop."""
+        await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload()
+        )
+
+        response = await authenticated_client.post(
+            "/books?library_id=1", files=self.upload(), data={"library_id": 1}
+        )
+
+        assert response.status_code == 409
+        assert response.json()["extra"][0]["filename"] == self.epub_path.name
+
+        forced = await authenticated_client.post(
+            "/books?library_id=1&allow_duplicates=true",
+            files=self.upload(),
+            data={"library_id": 1},
+        )
+        assert forced.status_code == 201
+
+    async def test_adding_another_books_file_conflicts(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        created = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload()
+        )
+        book_id = created.json()["created"][0]["id"]
+
+        other = await authenticated_client.post(
+            "/books?library_id=1",
+            files=[
+                (
+                    "files",
+                    (
+                        "war.epub",
+                        Path("tests/data_files/The Art of War - Sun Tzu.epub").read_bytes(),
+                        "application/epub+zip",
+                    ),
+                )
+            ],
+            data={"library_id": 1},
+        )
+        other_id = other.json()["id"]
+
+        response = await authenticated_client.post(
+            f"/books/{book_id}/files",
+            files=[
+                (
+                    "files",
+                    (
+                        "war.epub",
+                        Path("tests/data_files/The Art of War - Sun Tzu.epub").read_bytes(),
+                        "application/epub+zip",
+                    ),
+                )
+            ],
+        )
+
+        assert response.status_code == 409
+        assert response.json()["extra"][0]["book_id"] == other_id
+
+    async def test_resending_a_books_own_file_changes_nothing(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        created = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload()
+        )
+        book_id = created.json()["created"][0]["id"]
+
+        response = await authenticated_client.post(
+            f"/books/{book_id}/files", files=self.upload()
+        )
+
+        assert response.status_code == 201
+        assert len(response.json()["files"]) == 1
+
+    async def test_duplicates_can_be_checked_before_uploading(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        """The pre-flight check answers from hashes alone, with no file sent."""
+        created = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload()
+        )
+        book = created.json()["created"][0]
+        stored = book["files"][0]
+
+        response = await authenticated_client.post(
+            "/books/duplicate-files?library_id=1",
+            json=[
+                {
+                    "hash": stored["hash"],
+                    "size": stored["size"],
+                    "filename": "local-copy.epub",
+                },
+                {"hash": stored["hash"], "size": stored["size"] + 1},
+                {"hash": "0" * 32, "size": 1234},
+            ],
+        )
+
+        assert response.status_code == 200
+
+        matches = response.json()
+        assert len(matches) == 1
+        assert matches[0]["filename"] == "local-copy.epub"
+        assert matches[0]["book_id"] == book["id"]
+
+
+@pytest.mark.asyncio
+class TestDuplicateBooks:
+    """A second copy of a book is imported and reported, never refused."""
+
+    epub_path = Path("tests/data_files/Metamorphosis - Franz Kafka.epub")
+
+    def upload(self, name: str, pad: bool = False) -> list[tuple[str, tuple]]:
+        """
+        The fixture, optionally padded so it is a different file and the same book.
+
+        Padding the archive changes its size and its sampled hash without disturbing
+        the metadata, which is exactly the case the file-level check cannot see.
+        """
+        data = self.epub_path.read_bytes() + (b"\0" * 64 if pad else b"")
+
+        return [("files", (name, data, "application/epub+zip"))]
+
+    async def test_a_second_edition_is_created_and_reported(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        first = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload("first.epub")
+        )
+        original = first.json()["created"][0]
+
+        second = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload("second.epub", pad=True)
+        )
+
+        assert second.status_code == 201
+        body = second.json()
+
+        # Created, not skipped: a metadata match is a guess, and refusing a legitimate
+        # second edition costs more than a note does.
+        assert len(body["created"]) == 1
+        assert body["skipped"] == []
+
+        assert len(body["possible_duplicates"]) == 1
+        possible = body["possible_duplicates"][0]
+        assert possible["book_id"] == body["created"][0]["id"]
+
+        candidate = possible["candidates"][0]
+        assert candidate["book_id"] == original["id"]
+        assert candidate["title"] == original["title"]
+        assert candidate["authors"] == ["Franz Kafka"]
+        assert "title-author" in candidate["matched_on"]
+
+    async def test_the_review_screen_groups_them(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        first = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload("first.epub")
+        )
+        second = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload("second.epub", pad=True)
+        )
+
+        book_ids = sorted(
+            [first.json()["created"][0]["id"], second.json()["created"][0]["id"]]
+        )
+
+        response = await authenticated_client.get("/books/duplicate-books?library_id=1")
+
+        assert response.status_code == 200
+        groups = response.json()
+        assert len(groups) == 1
+        assert [book["book_id"] for book in groups[0]["books"]] == book_ids
+
+    async def test_a_dismissed_group_stays_dismissed(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        first = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload("first.epub")
+        )
+        second = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload("second.epub", pad=True)
+        )
+
+        pair = {
+            "book_a_id": first.json()["created"][0]["id"],
+            "book_b_id": second.json()["created"][0]["id"],
+        }
+
+        dismissed = await authenticated_client.post(
+            "/books/duplicate-books/dismissals", json=pair
+        )
+        assert dismissed.status_code == 204
+
+        response = await authenticated_client.get("/books/duplicate-books?library_id=1")
+        assert response.json() == []
+
+        restored = await authenticated_client.delete(
+            "/books/duplicate-books/dismissals"
+            f"?book_a_id={pair['book_b_id']}&book_b_id={pair['book_a_id']}"
+        )
+        assert restored.status_code == 204
+
+        response = await authenticated_client.get("/books/duplicate-books?library_id=1")
+        assert len(response.json()) == 1
+
+    async def test_two_books_merge_into_one(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        """The survivor keeps its id and gains the other's file; the other is gone."""
+        first = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload("first.epub")
+        )
+        second = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload("second.epub", pad=True)
+        )
+        keep = first.json()["created"][0]
+        fold = second.json()["created"][0]
+
+        response = await authenticated_client.post(
+            "/books/merge?library_id=1",
+            json={
+                "survivor_id": keep["id"],
+                "merged_ids": [fold["id"]],
+                "metadata": {"title": "Metamorphosis", "edition": 2},
+            },
+        )
+
+        assert response.status_code == 201
+        merged = response.json()
+
+        assert merged["id"] == keep["id"]
+        assert merged["edition"] == 2
+        assert len(merged["files"]) == 2
+
+        # The folded record is gone, and the group it formed with it.
+        assert (await authenticated_client.get(f"/books/{fold['id']}")).status_code == 404
+        groups = await authenticated_client.get("/books/duplicate-books?library_id=1")
+        assert groups.json() == []
+
+    async def test_merging_an_unknown_book_is_refused(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        created = await authenticated_client.post(
+            "/books/fromFiles?library_id=1", files=self.upload("first.epub")
+        )
+        keep = created.json()["created"][0]["id"]
+
+        response = await authenticated_client.post(
+            "/books/merge?library_id=1",
+            json={"survivor_id": keep, "merged_ids": [9999]},
+        )
+
+        assert response.status_code == 400
+
+    async def test_dismissing_an_unknown_book_is_refused(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        response = await authenticated_client.post(
+            "/books/duplicate-books/dismissals",
+            json={"book_a_id": 1, "book_b_id": 9999},
+        )
+
+        assert response.status_code == 400
+
+
 # NOTE: the multi-book ZIP download is covered at the service level, in
 # tests/unit/test_services/test_book_service.py. Driving `/books/download` through
 # AsyncTestClient hangs in fixture teardown: it is the only `Stream` endpoint in the
@@ -930,3 +1272,94 @@ class TestFileManagement:
         )
         # Should succeed (idempotent)
         assert response2.status_code == 204
+
+
+@pytest.mark.asyncio
+class TestUnnameableFormats:
+    """
+    A file whose format nothing can name must still round-trip.
+
+    `mimetypes.guess_type` answers None for `.mobi`, `.azw`, `.fb2` and `.lit`, which
+    is most of what a library imported from elsewhere carries alongside its EPUBs.
+    `FileMetadataRead.content_type` used to be a required string, so such a book was
+    created and then failed serialisation on its way back out — a 500 on a book the
+    reader can otherwise download.
+    """
+
+    def upload(self, name: str) -> list[tuple[str, tuple]]:
+        # `application/octet-stream` is what a browser posts for these, and it is not
+        # an answer — the extension is what names the format.
+        return [("files", (name, b"BOOKMOBI\x00 payload", "application/octet-stream"))]
+
+    async def test_a_mobi_is_named_from_its_extension(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        response = await authenticated_client.post(
+            "/books?library_id=1", files=self.upload("Dune.mobi"), data={"library_id": 1}
+        )
+
+        assert response.status_code == 201
+        book = response.json()
+        assert book["files"][0]["content_type"] == "application/x-mobipocket-ebook"
+
+        detail = await authenticated_client.get(f"/books/{book['id']}")
+        assert detail.status_code == 200
+
+    async def test_an_unknown_extension_stores_no_content_type(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        """Null, not a placeholder — and the book still serialises either way."""
+        response = await authenticated_client.post(
+            "/books?library_id=1",
+            files=self.upload("Notes.xyzzy"),
+            data={"library_id": 1},
+        )
+
+        assert response.status_code == 201
+        book = response.json()
+        assert book["files"][0]["content_type"] is None
+
+        detail = await authenticated_client.get(f"/books/{book['id']}")
+        assert detail.status_code == 200
+        assert detail.json()["files"][0]["content_type"] is None
+
+    async def test_the_file_downloads(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        """Litestar supplies its own media type when the row carries none."""
+        created = await authenticated_client.post(
+            "/books?library_id=1",
+            files=self.upload("Notes.xyzzy"),
+            data={"library_id": 1},
+        )
+        book = created.json()
+
+        response = await authenticated_client.get(
+            f"/books/download/{book['id']}/{book['files'][0]['id']}"
+        )
+
+        assert response.status_code == 200
+        assert response.headers["content-type"] == "application/octet-stream"
+
+    async def test_the_opds_feed_survives_a_null_content_type(
+        self, authenticated_client: AsyncClient
+    ) -> None:
+        """
+        The one place the type has to be a string.
+
+        `Link.type` is required, so a null fails the whole feed rather than one entry.
+        OPDS clients speak Basic, not the JWT the rest of the API uses.
+        """
+        await authenticated_client.post(
+            "/books?library_id=1",
+            files=self.upload("Notes.xyzzy"),
+            data={"library_id": 1},
+        )
+
+        feed = await authenticated_client.get(
+            "/opds/acquisition?feed_id=all&feed_title=All+Books",
+            auth=("user1@example.com", "password123"),
+        )
+
+        assert feed.status_code == 200
+        assert 'type="application/octet-stream"' in feed.text

backend/tests/integration/test_calibre_import.py

@@ -0,0 +1,400 @@
+"""
+Tests for the Calibre import endpoints.
+
+The API takes a zipped library and nothing else — a desktop Calibre install is usually
+not on the server, and importing from a path the server can already see stays a
+server-side operation (`litestar calibre-import`).
+"""
+
+import asyncio
+import tempfile
+import zipfile
+from pathlib import Path
+
+import pytest
+from httpx import AsyncClient
+
+from chitai.services.calibre_import import registry
+
+from tests.calibre_fixtures import CalibreFixture
+
+
+EPUB = Path("tests/data_files/Metamorphosis - Franz Kafka.epub")
+OTHER_EPUB = Path("tests/data_files/The Art of War - Sun Tzu.epub")
+
+
+@pytest.fixture(autouse=True)
+def _clear_registry():
+    """The registry is a module-level singleton, so it leaks between tests."""
+    registry._jobs.clear()
+    yield
+    registry._jobs.clear()
+
+
+@pytest.fixture(name="source")
+def fx_source(tmp_path: Path) -> Path:
+    fixture = CalibreFixture(tmp_path / "calibre")
+    fixture.add_book(
+        1,
+        "The Metamorphosis",
+        authors=["Franz Kafka"],
+        tags=["Fiction"],
+        cover=True,
+        formats={"EPUB": EPUB},
+    )
+    fixture.add_book(2, "The Art of War", authors=["Sun Tzu"], formats={"EPUB": OTHER_EPUB})
+    fixture.add_book(3, "Metadata Only", authors=["Nobody"])
+
+    return fixture.commit()
+
+
+def zip_of(root: Path, into: Path, prefix: str = "") -> Path:
+    """Zip a directory the way a file manager would."""
+    into.mkdir(parents=True, exist_ok=True)
+    archive = into / "library.zip"
+
+    with zipfile.ZipFile(archive, "w") as writing:
+        for path in sorted(root.rglob("*")):
+            if path.is_file():
+                writing.write(path, f"{prefix}{path.relative_to(root)}")
+
+    return archive
+
+
+async def upload(
+    client: AsyncClient,
+    archive: Path,
+    library_id: int = 1,
+    allow_duplicates: bool = False,
+) -> tuple[int, dict]:
+    response = await client.post(
+        f"/libraries/{library_id}/imports/calibre/upload",
+        files=[("archive", (archive.name, archive.read_bytes(), "application/zip"))],
+        data={"allow_duplicates": str(allow_duplicates).lower()},
+    )
+
+    return response.status_code, response.json()
+
+
+async def wait_for(client: AsyncClient, job_id: str) -> dict:
+    """Poll until the job is no longer running, the way the screen does."""
+    for _ in range(200):
+        response = await client.get(f"/libraries/imports/{job_id}")
+        assert response.status_code == 200
+
+        job = response.json()
+        if job["state"] != "running":
+            return job
+
+        await asyncio.sleep(0.05)
+
+    raise AssertionError("the import never finished")
+
+
+async def test_an_uploaded_library_imports(
+    authenticated_client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    status, job = await upload(
+        authenticated_client, zip_of(source, tmp_path / "out", prefix="Calibre Library/")
+    )
+
+    assert status == 202
+    assert job["state"] == "running"
+    assert job["library_id"] == 1
+
+    # The archive's name, not the temp directory it was unpacked into.
+    assert job["source"] == "library.zip"
+
+    finished = await wait_for(authenticated_client, job["id"])
+
+    assert finished["state"] == "finished"
+    assert finished["total"] == 3
+    assert finished["created"] == 2
+    assert finished["skipped"] == 1
+    assert finished["failed"] == 0
+    assert finished["error"] is None
+    assert finished["current_title"] is None
+
+    listed = await authenticated_client.get("/books?library_id=1")
+    titles = [book["title"] for book in listed.json()["items"]]
+    assert "The Metamorphosis" in titles
+    assert "The Art of War" in titles
+
+
+async def test_a_library_zipped_without_a_wrapping_folder(
+    authenticated_client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    """Zipping the contents is as common as zipping the folder."""
+    status, job = await upload(authenticated_client, zip_of(source, tmp_path / "out"))
+
+    assert status == 202
+    finished = await wait_for(authenticated_client, job["id"])
+    assert finished["created"] == 2
+
+
+async def test_the_unpacked_copy_is_cleaned_up(
+    authenticated_client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    """
+    An unpacked archive is a second copy of the whole library.
+
+    The books worth keeping have been copied into the library by the time the job ends,
+    so nothing is lost with it — and nothing will come back for it.
+    """
+    status, job = await upload(authenticated_client, zip_of(source, tmp_path / "out"))
+    assert status == 202
+
+    workspace = registry.get(job["id"]).workspace
+    assert workspace is not None
+
+    await wait_for(authenticated_client, job["id"])
+
+    assert not workspace.exists()
+
+
+async def test_uploading_the_same_library_twice_imports_nothing_new(
+    authenticated_client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    """Re-running is safe, which is what makes an interrupted import resumable."""
+    archive = zip_of(source, tmp_path / "out")
+
+    for _ in range(2):
+        _, job = await upload(authenticated_client, archive)
+        finished = await wait_for(authenticated_client, job["id"])
+
+    assert finished["created"] == 0
+    assert finished["skipped"] == 3
+
+
+async def test_allow_duplicates_stores_the_files_again(
+    authenticated_client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    """
+    The one option the screen offers, and it has to reach the import.
+
+    Without it the second pass skips everything, which is the previous test.
+    """
+    archive = zip_of(source, tmp_path / "out")
+
+    _, first = await upload(authenticated_client, archive)
+    await wait_for(authenticated_client, first["id"])
+
+    _, second = await upload(authenticated_client, archive, allow_duplicates=True)
+    finished = await wait_for(authenticated_client, second["id"])
+
+    assert finished["created"] == 2
+    assert finished["skipped"] == 1  # still the book with no files
+
+
+async def test_two_imports_into_one_library_conflict(
+    authenticated_client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    archive = zip_of(source, tmp_path / "out")
+
+    first_status, first = await upload(authenticated_client, archive)
+    assert first_status == 202
+
+    second_status, second = await upload(authenticated_client, archive)
+
+    assert second_status == 409
+    assert second["extra"]["job_id"] == first["id"]
+
+    await wait_for(authenticated_client, first["id"])
+
+
+async def test_a_finished_import_does_not_block_the_next_one(
+    authenticated_client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    archive = zip_of(source, tmp_path / "out")
+
+    _, first = await upload(authenticated_client, archive)
+    await wait_for(authenticated_client, first["id"])
+
+    status, second = await upload(authenticated_client, archive)
+
+    assert status == 202
+    await wait_for(authenticated_client, second["id"])
+
+
+async def test_cancelling_stops_after_the_current_book(
+    authenticated_client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    """
+    Cancelling is not aborting: a book abandoned mid-copy would leave files with no row.
+
+    Whether this cancels before any book, after one, or after the lot is a race — the
+    catalogue is three books long. What must hold either way is that the state is
+    terminal and every book it did import is complete.
+    """
+    _, job = await upload(authenticated_client, zip_of(source, tmp_path / "out"))
+
+    cancelled = await authenticated_client.delete(f"/libraries/imports/{job['id']}")
+    assert cancelled.status_code == 200
+
+    final = await wait_for(authenticated_client, job["id"])
+
+    assert final["state"] in {"cancelled", "finished"}
+
+    listed = await authenticated_client.get("/books?library_id=1")
+    for book in listed.json()["items"]:
+        assert book["files"]
+
+
+async def test_failures_are_reported_on_the_job(
+    authenticated_client: AsyncClient, tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    """One book failing is recorded and does not stop the run."""
+    fixture = CalibreFixture(tmp_path / "calibre")
+    fixture.add_book(1, "Fine", authors=["A"], formats={"EPUB": EPUB})
+    fixture.add_book(2, "Doomed", authors=["B"], formats={"EPUB": OTHER_EPUB})
+    root = fixture.commit()
+
+    from chitai.services.book import BookService
+
+    original = BookService.create
+
+    async def fail_on_the_second(self, data, *args, **kwargs):
+        if isinstance(data, dict) and data.get("title") == "Doomed":
+            raise RuntimeError("no room on the shelf")
+        return await original(self, data, *args, **kwargs)
+
+    monkeypatch.setattr(BookService, "create", fail_on_the_second)
+
+    _, job = await upload(authenticated_client, zip_of(root, tmp_path / "out"))
+    finished = await wait_for(authenticated_client, job["id"])
+
+    assert finished["state"] == "finished"
+    assert finished["created"] == 1
+    assert finished["failed"] == 1
+    assert finished["failures"][0]["calibre_id"] == 2
+    assert "no room on the shelf" in finished["failures"][0]["reason"]
+
+
+async def test_a_second_copy_is_counted_as_a_possible_duplicate(
+    authenticated_client: AsyncClient, tmp_path: Path
+) -> None:
+    """A count, not the records — the duplicates screen is what shows them."""
+    padded = tmp_path / "padded.epub"
+    padded.write_bytes(EPUB.read_bytes() + b"\0" * 64)
+
+    fixture = CalibreFixture(tmp_path / "calibre")
+    fixture.add_book(1, "The Metamorphosis", authors=["Franz Kafka"], formats={"EPUB": EPUB})
+    fixture.add_book(
+        2, "The Metamorphosis", authors=["Franz Kafka"], formats={"EPUB": padded}
+    )
+    root = fixture.commit()
+
+    _, job = await upload(authenticated_client, zip_of(root, tmp_path / "out"))
+    finished = await wait_for(authenticated_client, job["id"])
+
+    assert finished["created"] == 2
+    assert finished["possible_duplicates"] == 1
+
+
+async def test_polling_an_unknown_job(authenticated_client: AsyncClient) -> None:
+    response = await authenticated_client.get("/libraries/imports/not-a-job")
+    assert response.status_code == 404
+
+
+async def test_cancelling_an_unknown_job(authenticated_client: AsyncClient) -> None:
+    response = await authenticated_client.delete("/libraries/imports/not-a-job")
+    assert response.status_code == 404
+
+
+async def test_importing_into_a_library_that_does_not_exist(
+    authenticated_client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    status, _ = await upload(
+        authenticated_client, zip_of(source, tmp_path / "out"), library_id=999
+    )
+
+    assert status == 404
+
+
+async def test_importing_into_a_read_only_library(
+    authenticated_client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    """A read-only library points at a tree Chitai does not own."""
+    root = tmp_path / "read-only"
+    root.mkdir()
+
+    created = await authenticated_client.post(
+        "/libraries",
+        json={"name": "Read Only", "root_path": str(root), "read_only": True},
+    )
+    assert created.status_code == 201
+
+    status, body = await upload(
+        authenticated_client,
+        zip_of(source, tmp_path / "out"),
+        library_id=created.json()["id"],
+    )
+
+    assert status == 400
+    assert "read-only" in body["detail"]
+
+
+async def test_an_import_needs_authentication(
+    client: AsyncClient, source: Path, tmp_path: Path
+) -> None:
+    status, _ = await upload(client, zip_of(source, tmp_path / "out"))
+
+    assert status == 401
+
+
+class TestRefusedArchives:
+    """Everything wrong with an archive is answered now, not as a job that fails later."""
+
+    async def test_a_hostile_archive(
+        self, authenticated_client: AsyncClient, tmp_path: Path
+    ) -> None:
+        """Zip slip."""
+        archive = tmp_path / "hostile.zip"
+
+        with zipfile.ZipFile(archive, "w") as writing:
+            writing.writestr("metadata.db", "not really")
+            writing.writestr("../../escaped.txt", "gotcha")
+
+        status, body = await upload(authenticated_client, archive)
+
+        assert status == 400
+        assert "outside itself" in body["detail"]
+        assert registry._jobs == {}
+
+    async def test_something_that_is_not_a_zip(
+        self, authenticated_client: AsyncClient, tmp_path: Path
+    ) -> None:
+        archive = tmp_path / "notes.txt"
+        archive.write_bytes(b"just some text")
+
+        status, body = await upload(authenticated_client, archive)
+
+        assert status == 400
+        assert "not a zip" in body["detail"]
+
+    async def test_an_archive_with_no_catalogue(
+        self, authenticated_client: AsyncClient, tmp_path: Path
+    ) -> None:
+        archive = tmp_path / "books.zip"
+
+        with zipfile.ZipFile(archive, "w") as writing:
+            writing.writestr("Some Book.epub", "content")
+
+        status, body = await upload(authenticated_client, archive)
+
+        assert status == 400
+        assert "no metadata.db" in body["detail"]
+
+    async def test_a_refusal_leaves_no_temp_files(
+        self, authenticated_client: AsyncClient, tmp_path: Path
+    ) -> None:
+        """Every refusal path removes the workspace it had already made."""
+        before = set(Path(tempfile.gettempdir()).glob("tmp*"))
+
+        archive = tmp_path / "books.zip"
+        with zipfile.ZipFile(archive, "w") as writing:
+            writing.writestr("Some Book.epub", "content")
+
+        await upload(authenticated_client, archive)
+
+        assert set(Path(tempfile.gettempdir()).glob("tmp*")) == before

backend/tests/unit/test_calibre.py

@@ -0,0 +1,392 @@
+import zipfile
+from datetime import date
+from pathlib import Path
+from types import SimpleNamespace
+
+import pytest
+
+from chitai.services.calibre import (
+    CalibreLibrary,
+    CalibreLibraryError,
+    extract_calibre_archive,
+    format_series_index,
+    parse_date,
+    strip_html,
+    unescape_author,
+)
+
+from tests.calibre_fixtures import UNDEFINED_DATE, CalibreFixture
+
+
+EPUB = Path("tests/data_files/Metamorphosis - Franz Kafka.epub")
+PDF = Path("tests/data_files/Calculus Made Easy - Silvanus Thompson.pdf")
+
+
+@pytest.fixture(name="library_root")
+def fx_library_root(tmp_path: Path) -> Path:
+    """A small Calibre library covering the rows that are easy to read wrongly."""
+    fixture = CalibreFixture(tmp_path / "Calibre Library")
+    fixture.add_pages_table()
+
+    fixture.add_book(
+        1,
+        "The Metamorphosis",
+        authors=["Franz Kafka"],
+        pubdate="1915-10-15 00:00:00+00:00",
+        tags=["Fiction", "Absurdist"],
+        publisher="Kurt Wolff Verlag",
+        languages=["deu", "eng"],
+        comment="<p>A travelling salesman.</p><p>He wakes up <i>changed</i>.</p>",
+        identifiers={"isbn": "978-0-486-29030-0", "amazon": "B01N5IB20Q"},
+        uuid="11111111-2222-3333-4444-555555555555",
+        pages=201,
+        cover=True,
+        formats={"EPUB": EPUB},
+    )
+
+    # Volume seven of a series, and no publication date — the two values most likely to
+    # be carried through verbatim when they should not be.
+    fixture.add_book(
+        2,
+        "Persepolis Rising",
+        # Calibre escapes the comma and nothing else, so the space after it is stored
+        # as-is: `Corey, Jr.` is written `Corey| Jr.`.
+        authors=["Corey| Jr., James S. A."],
+        series="The Expanse",
+        series_index=7.0,
+        formats={"EPUB": EPUB},
+    )
+
+    # A novella between two novels: a fractional position is real and must survive.
+    fixture.add_book(
+        3, "Strange Dogs", series="The Expanse", series_index=6.5, formats={"PDF": PDF}
+    )
+
+    # Every row Calibre will happily hold and Chitai cannot use: no files at all.
+    fixture.add_book(4, "Metadata Only")
+
+    # A catalogue row whose file is not on disk.
+    fixture.add_book(5, "Lost Book")
+    fixture.add_missing_format(5, "EPUB", "Lost Book - Unknown")
+
+    return fixture.commit()
+
+
+async def test_reads_a_book_whole(library_root: Path) -> None:
+    async with CalibreLibrary(library_root) as library:
+        assert await library.count() == 5
+        books = await library.books()
+
+    book = books[0]
+
+    assert book.calibre_id == 1
+    assert book.title == "The Metamorphosis"
+    assert book.authors == ["Franz Kafka"]
+    assert book.published_date == date(1915, 10, 15)
+    assert book.tags == ["Absurdist", "Fiction"]
+    assert book.publisher == "Kurt Wolff Verlag"
+    assert book.pages == 201
+    assert book.uuid == "11111111-2222-3333-4444-555555555555"
+
+    # One language, and the one Calibre put first.
+    assert book.language == "deu"
+
+    # Reported as Calibre wrote them: folding `amazon` onto `asin` is the importer's
+    # job, not the reader's.
+    assert book.identifiers == {"isbn": "978-0-486-29030-0", "amazon": "B01N5IB20Q"}
+
+    assert book.cover is not None
+    assert book.cover.is_file()
+
+    assert len(book.files) == 1
+    assert book.files[0].format == "EPUB"
+    assert book.files[0].path.is_file()
+    # The stem is Calibre's, truncated and sanitised — never the title.
+    assert book.files[0].path.name != f"{book.title}.epub"
+
+
+async def test_the_undefined_date_is_not_a_date(library_root: Path) -> None:
+    """`0101-01-01` parses fine, which is exactly the problem."""
+    async with CalibreLibrary(library_root) as library:
+        books = {book.calibre_id: book for book in await library.books()}
+
+    assert books[2].published_date is None
+
+
+async def test_series_position_is_a_plain_string(library_root: Path) -> None:
+    async with CalibreLibrary(library_root) as library:
+        books = {book.calibre_id: book for book in await library.books()}
+
+    assert books[2].series == "The Expanse"
+    assert books[2].series_position == "7"
+
+    assert books[3].series_position == "6.5"
+
+    # `series_index` defaults to 1.0 for every book, so a position without a series
+    # would invent a volume one out of nothing.
+    assert books[1].series is None
+    assert books[1].series_position is None
+
+
+async def test_author_commas_are_unescaped(library_root: Path) -> None:
+    async with CalibreLibrary(library_root) as library:
+        books = {book.calibre_id: book for book in await library.books()}
+
+    assert books[2].authors == ["Corey, Jr., James S. A."]
+
+
+async def test_comments_come_back_as_text(library_root: Path) -> None:
+    async with CalibreLibrary(library_root) as library:
+        books = {book.calibre_id: book for book in await library.books()}
+
+    assert books[1].description == "A travelling salesman.\nHe wakes up changed."
+    assert books[2].description is None
+
+
+async def test_files_are_reported_whether_or_not_they_exist(library_root: Path) -> None:
+    """
+    The reader says what the catalogue says. Whether the bytes are there is a question
+    for whoever is about to copy them, which stats them anyway.
+    """
+    async with CalibreLibrary(library_root) as library:
+        books = {book.calibre_id: book for book in await library.books()}
+
+    assert books[4].files == []
+
+    assert len(books[5].files) == 1
+    assert not books[5].files[0].path.exists()
+
+
+async def test_a_library_without_the_pages_table_still_reads(tmp_path: Path) -> None:
+    """`books_pages_link` is recent; an older library simply does not have it."""
+    fixture = CalibreFixture(tmp_path / "Old Library")
+    fixture.add_book(1, "Old Book", formats={"EPUB": EPUB})
+    root = fixture.commit()
+
+    async with CalibreLibrary(root) as library:
+        books = await library.books()
+
+    assert books[0].pages is None
+
+
+async def test_the_original_is_never_opened(library_root: Path) -> None:
+    """
+    The catalogue is copied before it is read, and the copy goes away afterwards.
+
+    Calibre may be running and writing; this is what keeps a live library out of it.
+    """
+    before = (library_root / "metadata.db").read_bytes()
+
+    library = CalibreLibrary(library_root)
+    await library.open()
+    workspace = library._workspace
+
+    assert workspace is not None and (workspace / "metadata.db").is_file()
+
+    await library.close()
+
+    assert not workspace.exists()
+    assert (library_root / "metadata.db").read_bytes() == before
+
+
+async def test_closing_twice_is_harmless(library_root: Path) -> None:
+    library = CalibreLibrary(library_root)
+    await library.open()
+    await library.close()
+    await library.close()
+
+
+async def test_a_directory_that_is_not_a_calibre_library(tmp_path: Path) -> None:
+    with pytest.raises(CalibreLibraryError, match="not a Calibre library"):
+        await CalibreLibrary(tmp_path).open()
+
+
+@pytest.mark.parametrize(
+    ("stored", "expected"),
+    [
+        ("2017-12-04 04:00:00+00:00", date(2017, 12, 4)),
+        ("2001-07-02 00:00:00+00:00", date(2001, 7, 2)),
+        ("1999-01-31", date(1999, 1, 31)),
+        # Calibre's sentinel, and anything else implausibly early.
+        (UNDEFINED_DATE, None),
+        ("0101-01-01", None),
+        (None, None),
+        ("", None),
+        ("not a date", None),
+    ],
+)
+def test_parse_date(stored: str | None, expected: date | None) -> None:
+    assert parse_date(stored) == expected
+
+
+@pytest.mark.parametrize(
+    ("index", "expected"),
+    [
+        (7.0, "7"),
+        (1.0, "1"),
+        (6.5, "6.5"),
+        (0.0, "0"),
+        (12.25, "12.25"),
+        (None, None),
+    ],
+)
+def test_format_series_index(index: float | None, expected: str | None) -> None:
+    assert format_series_index(index) == expected
+
+
+@pytest.mark.parametrize(
+    ("stored", "expected"),
+    [
+        ("Doyle| Sir Arthur Conan", "Doyle, Sir Arthur Conan"),
+        ("Franz Kafka", "Franz Kafka"),
+        ("  Herman Melville  ", "Herman Melville"),
+    ],
+)
+def test_unescape_author(stored: str, expected: str) -> None:
+    assert unescape_author(stored) == expected
+
+
+@pytest.mark.parametrize(
+    ("html", "expected"),
+    [
+        ("<p>One.</p><p>Two.</p>", "One.\nTwo."),
+        ("Plain text", "Plain text"),
+        ("<div>A<br>B</div>", "A\nB"),
+        ("<p>Caf&eacute; &amp; bar</p>", "Café & bar"),
+        ("<ul><li>One</li><li>Two</li></ul>", "One\nTwo"),
+        # Markup carrying no text at all is nothing, not an empty description.
+        ("<p></p>", None),
+        ("", None),
+        (None, None),
+    ],
+)
+def test_strip_html(html: str | None, expected: str | None) -> None:
+    assert strip_html(html) == expected
+
+
+class TestArchives:
+    """A Calibre library that arrives zipped rather than as a path."""
+
+    def zipped(self, root: Path, into: Path, prefix: str = "") -> Path:
+        """Zip a directory the way a file manager would."""
+        archive = into / "library.zip"
+
+        with zipfile.ZipFile(archive, "w") as writing:
+            for path in sorted(root.rglob("*")):
+                if path.is_file():
+                    writing.write(path, f"{prefix}{path.relative_to(root)}")
+
+        return archive
+
+    async def test_a_library_zipped_at_its_root(
+        self, library_root: Path, tmp_path: Path
+    ) -> None:
+        archive = self.zipped(library_root, tmp_path)
+        destination = tmp_path / "unpacked"
+        destination.mkdir()
+
+        catalogue = await extract_calibre_archive(archive, destination)
+
+        assert catalogue == destination
+        async with CalibreLibrary(catalogue) as library:
+            assert await library.count() == 5
+
+    async def test_a_library_zipped_inside_a_folder(
+        self, library_root: Path, tmp_path: Path
+    ) -> None:
+        """Zipping the folder itself is at least as common as zipping its contents."""
+        archive = self.zipped(library_root, tmp_path, prefix="Calibre Library/")
+        destination = tmp_path / "unpacked"
+        destination.mkdir()
+
+        catalogue = await extract_calibre_archive(archive, destination)
+
+        assert catalogue == destination / "Calibre Library"
+        async with CalibreLibrary(catalogue) as library:
+            assert await library.count() == 5
+
+    async def test_an_entry_pointing_outside_the_archive_is_refused(
+        self, tmp_path: Path
+    ) -> None:
+        """
+        Zip slip. `ZipFile.extract` sanitises names itself, but relying on that silently
+        is how the next person to change the extraction call reintroduces it.
+        """
+        archive = tmp_path / "hostile.zip"
+
+        with zipfile.ZipFile(archive, "w") as writing:
+            writing.writestr("metadata.db", "not really")
+            writing.writestr("../../escaped.txt", "gotcha")
+
+        destination = tmp_path / "unpacked"
+        destination.mkdir()
+
+        with pytest.raises(CalibreLibraryError, match="outside itself"):
+            await extract_calibre_archive(archive, destination)
+
+        assert not (tmp_path.parent / "escaped.txt").exists()
+
+    async def test_something_that_is_not_a_zip(self, tmp_path: Path) -> None:
+        archive = tmp_path / "not.zip"
+        archive.write_bytes(b"PK-ish, but no")
+
+        destination = tmp_path / "unpacked"
+        destination.mkdir()
+
+        with pytest.raises(CalibreLibraryError, match="not a zip file"):
+            await extract_calibre_archive(archive, destination)
+
+    async def test_an_archive_with_no_catalogue(self, tmp_path: Path) -> None:
+        archive = tmp_path / "books.zip"
+
+        with zipfile.ZipFile(archive, "w") as writing:
+            writing.writestr("Some Book.epub", "content")
+
+        destination = tmp_path / "unpacked"
+        destination.mkdir()
+
+        with pytest.raises(CalibreLibraryError, match="no metadata.db"):
+            await extract_calibre_archive(archive, destination)
+
+        # Refused before anything was written.
+        assert list(destination.iterdir()) == []
+
+    async def test_a_catalogue_buried_too_deep(self, tmp_path: Path) -> None:
+        """Somebody's whole backup tree is not a library, however much it contains one."""
+        archive = tmp_path / "backup.zip"
+
+        with zipfile.ZipFile(archive, "w") as writing:
+            writing.writestr("backups/2026/january/library/metadata.db", "not really")
+
+        destination = tmp_path / "unpacked"
+        destination.mkdir()
+
+        with pytest.raises(CalibreLibraryError, match="within 3 levels"):
+            await extract_calibre_archive(archive, destination)
+
+    async def test_an_archive_too_big_for_the_disk(
+        self, tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+    ) -> None:
+        """
+        Checked before writing, not discovered part-way through.
+
+        A full disk takes the whole application down, and the size is in the archive
+        already.
+        """
+        archive = tmp_path / "huge.zip"
+
+        with zipfile.ZipFile(archive, "w") as writing:
+            writing.writestr("metadata.db", "not really")
+
+        destination = tmp_path / "unpacked"
+        destination.mkdir()
+
+        monkeypatch.setattr(
+            "chitai.services.calibre.shutil.disk_usage",
+            lambda _path: SimpleNamespace(total=1024, used=1024, free=0),
+        )
+
+        with pytest.raises(CalibreLibraryError, match="only"):
+            await extract_calibre_archive(archive, destination)
+
+        assert list(destination.iterdir()) == []

backend/tests/unit/test_content_type.py

@@ -0,0 +1,64 @@
+from pathlib import Path
+
+import pytest
+
+from chitai.services.utils import guess_content_type
+
+
+@pytest.mark.parametrize(
+    ("filename", "expected"),
+    [
+        # What `mimetypes` already knows, kept here so a host with a thin
+        # /etc/mime.types cannot change the answer without a test noticing.
+        ("Frankenstein.epub", "application/epub+zip"),
+        ("Calculus.pdf", "application/pdf"),
+        ("Persepolis.azw3", "application/vnd.amazon.mobi8-ebook"),
+        ("Watchmen.cbz", "application/vnd.comicbook+zip"),
+        # What it does not, and where a Calibre library's older formats live.
+        ("Dune.mobi", "application/x-mobipocket-ebook"),
+        ("Dune.prc", "application/x-mobipocket-ebook"),
+        ("Dune.azw", "application/vnd.amazon.ebook"),
+        ("Voyna i Mir.fb2", "application/x-fictionbook+xml"),
+        ("Voyna i Mir.fbz", "application/x-zip-compressed-fb2"),
+        ("Reader.lit", "application/x-ms-reader"),
+        ("Reader.lrf", "application/x-sony-bbeb"),
+        ("Watchmen.cb7", "application/x-cb7"),
+        # Case is not part of the answer, and Calibre writes formats uppercase.
+        ("Dune.MOBI", "application/x-mobipocket-ebook"),
+        # Nothing can name these, and None is the answer rather than a placeholder.
+        ("Notes.xyzzy", None),
+        ("README", None),
+    ],
+)
+def test_guess_content_type(filename: str, expected: str | None) -> None:
+    assert guess_content_type(Path(filename)) == expected
+    # A str and a Path must agree, and an upload's `filename` carries its relative
+    # path, so a name with directories in front of it has to resolve the same way.
+    assert guess_content_type(filename) == expected
+    assert guess_content_type(f"Some Author/Some Book/{filename}") == expected
+
+
+def test_fallback_is_used_only_when_the_extension_says_nothing() -> None:
+    """A client's claim fills a gap; it never overrides the name."""
+    assert (
+        guess_content_type(Path("Dune.mobi"), fallback="application/pdf")
+        == "application/x-mobipocket-ebook"
+    )
+    assert (
+        guess_content_type(Path("Notes.xyzzy"), fallback="application/epub+zip")
+        == "application/epub+zip"
+    )
+
+
+def test_an_unspecified_fallback_is_not_an_answer() -> None:
+    """
+    `application/octet-stream` from a client is it saying it does not know.
+
+    Browsers post exactly that for every extension they do not recognise, which is most
+    ebook formats. Storing it would be indistinguishable from having determined a
+    format, so it is discarded and the column keeps its null.
+    """
+    assert (
+        guess_content_type(Path("Notes.xyzzy"), fallback="application/octet-stream")
+        is None
+    )

backend/tests/unit/test_filesystem_library.py

@@ -0,0 +1,69 @@
+"""Tests for BookPathGenerator."""
+
+from pathlib import Path
+
+from chitai.services.filesystem_library import BookPathGenerator, sanitize_path_component
+
+
+ROOT = Path("/library")
+
+
+def path_for(**book) -> Path:
+    return BookPathGenerator(ROOT).generate_path(book)
+
+
+def test_author_and_title() -> None:
+    assert path_for(title="Dune", authors=["Frank Herbert"]) == (
+        ROOT / "Frank Herbert" / "Dune"
+    )
+
+
+def test_a_book_with_no_authors() -> None:
+    assert path_for(title="Beowulf", authors=[]) == ROOT / "Unknown" / "Beowulf"
+
+
+def test_a_series_adds_a_level_and_pads_the_position() -> None:
+    assert path_for(
+        title="Persepolis Rising",
+        authors=["James S. A. Corey"],
+        series="The Expanse",
+        series_position="7",
+    ) == ROOT / "James S. A. Corey" / "The Expanse" / "07 - Persepolis Rising"
+
+
+def test_a_slash_in_a_title_does_not_add_a_directory() -> None:
+    """
+    The separators in the path come from the template, never from the metadata.
+
+    A title with a slash in it — "AC/DC", "Him/Her" — would otherwise put the book one
+    level below where `book.path` says it is, which is what deletes, moves and file
+    lookups all act on. Calibre keeps the real title in its database and strips this
+    from its own directory names, so an import is where they surface.
+    """
+    generated = path_for(title="Back in Black: AC/DC", authors=["Murray Engleheart"])
+
+    assert generated == ROOT / "Murray Engleheart" / "Back in Black: AC_DC"
+    assert generated.relative_to(ROOT).parts == ("Murray Engleheart", "Back in Black: AC_DC")
+
+
+def test_a_slash_in_an_author_or_series_is_handled_too() -> None:
+    assert path_for(title="Split", authors=["A/B Collective"]) == (
+        ROOT / "A_B Collective" / "Split"
+    )
+    assert path_for(
+        title="Volume One", authors=["Someone"], series="Either/Or", series_position="1"
+    ) == ROOT / "Someone" / "Either_Or" / "01 - Volume One"
+
+
+def test_control_characters_are_removed() -> None:
+    assert path_for(title="Line\nBreak", authors=["Someone"]) == (
+        ROOT / "Someone" / "Line_Break"
+    )
+
+
+def test_sanitize_path_component() -> None:
+    assert sanitize_path_component("AC/DC") == "AC_DC"
+    assert sanitize_path_component("back\\slash") == "back_slash"
+    assert sanitize_path_component("  padded  ") == "padded"
+    # Colons and other punctuation are legal in a path and are left alone.
+    assert sanitize_path_component("Title: Subtitle") == "Title: Subtitle"

backend/tests/unit/test_matching.py

@@ -0,0 +1,233 @@
+"""Tests for the normalization behind book-level duplicate detection."""
+
+import pytest
+
+from chitai.services.matching import (
+    format_author_name,
+    normalize_author,
+    normalize_identifier,
+    normalize_title,
+)
+from chitai.services.metadata_extractor import parse_identifier
+from chitai.services.utils import isbn10_to_isbn13
+
+
+class TestNormalizeTitle:
+    """Two copies of one book rarely agree on how the title is written."""
+
+    @pytest.mark.parametrize(
+        ("title", "expected"),
+        [
+            ("The Metamorphosis", "metamorphosis"),
+            ("Metamorphosis", "metamorphosis"),
+            ("METAMORPHOSIS", "metamorphosis"),
+            ("A Tale of Two Cities", "tale of two cities"),
+            ("An Enquiry", "enquiry"),
+            # Accents, punctuation and ampersands are spelling, not identity.
+            ("Les Misérables", "les miserables"),
+            ("Moby Dick; Or, The Whale", "moby dick or the whale"),
+            ("Sense & Sensibility", "sense and sensibility"),
+            # Bracketed asides and trailing edition noise say nothing about the book.
+            ("Frankenstein (Illustrated)", "frankenstein"),
+            ("Frankenstein [Kindle Edition]", "frankenstein"),
+            ("Frankenstein, 2nd Edition", "frankenstein"),
+            # The compact forms a cover actually carries.
+            ("Building Microservices, 2E", "building microservices"),
+            ("Building Microservices 2e", "building microservices"),
+            ("Frankenstein 3 Ed", "frankenstein"),
+            ("Dungeons & Dragons 5e", "dungeons and dragons"),
+            ("Frankenstein Revised Edition", "frankenstein"),
+            ("Dune Deluxe Edition Illustrated", "dune"),
+            ("", ""),
+            (None, ""),
+        ],
+    )
+    def test_titles_that_should_agree(self, title: str | None, expected: str) -> None:
+        assert normalize_title(title) == expected
+
+    def test_a_qualifier_that_is_the_title_survives(self) -> None:
+        """A trailing qualifier is noise; the same word at the front is the book."""
+        assert normalize_title("The Illustrated Man") == "illustrated man"
+
+    def test_normalization_never_empties_a_title(self) -> None:
+        """An article-only title is not improved by having no article left."""
+        assert normalize_title("The") == "the"
+
+    @pytest.mark.parametrize(
+        "title", ["Catch 22", "Fahrenheit 451", "Blade Runner 2049", "1984", "Apollo 13"]
+    )
+    def test_a_number_is_not_an_edition(self, title: str) -> None:
+        """Edition stripping keys on the `e`; a bare number is part of the title."""
+        assert normalize_title(title) == title.casefold()
+
+
+class TestNormalizeAuthor:
+    """One person, written down several ways."""
+
+    @pytest.mark.parametrize(
+        ("name", "expected"),
+        [
+            ("Franz Kafka", "franz kafka"),
+            ("Kafka, Franz", "franz kafka"),
+            ("KAFKA, FRANZ", "franz kafka"),
+            ("Émile Zola", "emile zola"),
+            ("Doyle, Arthur Conan", "arthur conan doyle"),
+            # Runs of initials are joined, so spacing them out changes nothing.
+            ("J.R.R. Tolkien", "jrr tolkien"),
+            ("J. R. R. Tolkien", "jrr tolkien"),
+            ("JRR Tolkien", "jrr tolkien"),
+            ("", ""),
+            (None, ""),
+        ],
+    )
+    def test_names_that_should_agree(self, name: str | None, expected: str) -> None:
+        assert normalize_author(name) == expected
+
+    def test_two_people_are_not_reduced_together(self) -> None:
+        """Surname plus initial would collide unrelated writers; it is not used."""
+        assert normalize_author("Charles Dickens") != normalize_author("Colin Dexter")
+
+    def test_a_list_is_left_alone(self) -> None:
+        """More than one comma is a list or a suffix, and guessing does more harm."""
+        assert normalize_author("Smith, John, Jr.") == "smith john jr"
+
+
+class TestFormatAuthorName:
+    """What gets stored and shown, as opposed to what gets compared."""
+
+    @pytest.mark.parametrize(
+        ("written", "expected"),
+        [
+            # A leftover separator from a `DC:creator` list.
+            ("Newman, Sam;", "Sam Newman"),
+            ("Sam Newman ", "Sam Newman"),
+            ("  Dan   Vanderkam  ", "Dan Vanderkam"),
+            # `Surname, Given` is how EPUBs file a name, not how anyone reads it.
+            ("Kleppmann, Martin", "Martin Kleppmann"),
+            ("Huxley, Aldous", "Aldous Huxley"),
+            ("Liu, Cixin", "Cixin Liu"),
+            # An extension carried in from the filename the name was read out of.
+            ("Sam Newman.epub", "Sam Newman"),
+            ("Franz Kafka.mobi", "Franz Kafka"),
+            ("Brian W. Kernighan.epub", "Brian W. Kernighan"),
+            ("", ""),
+            (None, ""),
+        ],
+    )
+    def test_names_are_tidied(self, written: str | None, expected: str) -> None:
+        assert format_author_name(written) == expected
+
+    @pytest.mark.parametrize(
+        "written",
+        [
+            # Two people in one string. Flipping it would invent a third.
+            "Dave Thomas, Andy Hunt",
+            "Mark Richards, Neal Ford",
+            # A compound surname is not recognised, and is left alone rather than
+            # rearranged wrongly.
+            "García Márquez, Gabriel",
+        ],
+    )
+    def test_an_unrecognised_form_is_left_alone(self, written: str) -> None:
+        assert format_author_name(written) == written
+
+    def test_case_and_accents_belong_to_the_author(self) -> None:
+        """Tidying removes what an extractor added; it does not correct spelling."""
+        assert format_author_name("Michał Płachta.epub") == "Michał Płachta"
+        assert format_author_name("Steve McConnell") == "Steve McConnell"
+
+    @pytest.mark.parametrize(
+        "written", ["Newman, Sam;", "Sam Newman.epub", "Kleppmann, Martin"]
+    )
+    def test_tidying_is_idempotent(self, written: str) -> None:
+        """`unique_filter` tidies a name that may already be tidy; it must not drift."""
+        once = format_author_name(written)
+        assert format_author_name(once) == once
+
+
+class TestNormalizeIdentifier:
+    """Identifiers only help if the same edition produces the same key."""
+
+    def test_isbn_10_and_isbn_13_are_one_key(self) -> None:
+        assert normalize_identifier("isbn-10", "0486282112") == "isbn:9780486282114"
+        assert normalize_identifier("isbn-13", "9780486282114") == "isbn:9780486282114"
+
+    @pytest.mark.parametrize(
+        "written", ["978-0-486-28211-4", "978 0 486 28211 4", "9780486282114"]
+    )
+    def test_formatting_is_not_part_of_an_isbn(self, written: str) -> None:
+        assert normalize_identifier("isbn", written) == "isbn:9780486282114"
+
+    def test_an_isbn_that_fails_its_checksum_is_no_evidence(self) -> None:
+        assert normalize_identifier("isbn-13", "9780486282115") is None
+
+    def test_uuids_are_refused(self) -> None:
+        """Generated per build, so they only re-find what the hash check catches."""
+        assert normalize_identifier("uuid", "3f2b1c4e-1111-2222-3333-444455556666") is None
+        assert normalize_identifier("urn:uuid", "3f2b1c4e-1111-2222-3333-444455556666") is None
+
+    def test_other_schemes_keep_their_own_key(self) -> None:
+        assert normalize_identifier("asin", "B000FC0PDA") == "asin:b000fc0pda"
+        assert normalize_identifier("ASIN", "b000fc0pda") == "asin:b000fc0pda"
+
+    def test_something_too_short_is_not_evidence(self) -> None:
+        """A Calibre id of "42" would otherwise pair two unrelated books."""
+        assert normalize_identifier("calibre", "42") is None
+
+    @pytest.mark.parametrize(("name", "value"), [("", "1234567"), ("asin", "")])
+    def test_half_an_identifier_is_no_identifier(self, name: str, value: str) -> None:
+        assert normalize_identifier(name, value) is None
+
+
+class TestIsbnConversion:
+    def test_isbn_10_converts_to_its_isbn_13(self) -> None:
+        assert isbn10_to_isbn13("0486282112") == "9780486282114"
+
+    def test_a_trailing_x_is_a_digit(self) -> None:
+        assert isbn10_to_isbn13("043942089X") == "9780439420891"
+
+    @pytest.mark.parametrize("isbn", ["0486282113", "9780486282114", "nonsense"])
+    def test_anything_that_is_not_an_isbn_10_converts_to_nothing(self, isbn: str) -> None:
+        assert isbn10_to_isbn13(isbn) is None
+
+
+class TestParseIdentifier:
+    """What an EPUB writes, and what is worth storing for it."""
+
+    @pytest.mark.parametrize(
+        "written",
+        [
+            "9780486282114",
+            "978-0-486-28211-4",
+            "urn:isbn:9780486282114",
+            "urn:isbn:978-0-486-28211-4",
+            "ISBN:978-0-486-28211-4",
+        ],
+    )
+    def test_isbns_survive_however_they_are_written(self, written: str) -> None:
+        assert parse_identifier(written) == ("isbn-13", "9780486282114")
+
+    def test_the_scheme_attribute_is_read_too(self) -> None:
+        assert parse_identifier("0-486-28211-2", "ISBN") == ("isbn-10", "0486282112")
+
+    def test_non_isbn_identifiers_are_kept(self) -> None:
+        assert parse_identifier("urn:uuid:3f2b1c4e-1111-2222-3333-444455556666") == (
+            "uuid",
+            "3f2b1c4e-1111-2222-3333-444455556666",
+        )
+        assert parse_identifier("calibre:1234") == ("calibre", "1234")
+        assert parse_identifier("B000FC0PDA", "mobi-asin") == ("asin", "B000FC0PDA")
+
+    def test_an_unrecognised_prefix_is_part_of_the_value(self) -> None:
+        """"http://example.com/book" is not an identifier called "http"."""
+        assert parse_identifier("http://www.gutenberg.org/5200") == (
+            "id",
+            "http://www.gutenberg.org/5200",
+        )
+
+    def test_a_declared_isbn_that_is_not_one_is_dropped(self) -> None:
+        assert parse_identifier("urn:isbn:not-an-isbn") is None
+
+    @pytest.mark.parametrize("written", ["", "   ", None])
+    def test_nothing_yields_nothing(self, written: str | None) -> None:
+        assert parse_identifier(written) is None

backend/tests/unit/test_metadata_extractor.py

@@ -1,7 +1,13 @@
 import pytest
+from ebooklib import epub
 from pathlib import Path
 from datetime import date
-from chitai.services.metadata_extractor import EpubExtractor
+from chitai.services.metadata_extractor import (
+    EpubExtractor,
+    Extractor,
+    PdfExtractor,
+    split_edition,
+)
 
 
 @pytest.mark.asyncio()
@@ -15,3 +21,172 @@ class TestEpubExtractor:
         assert metadata["authors"] == ["Herman Melville"]
         assert metadata["published_date"] == date(year=2001, month=7, day=1)
 
+
+
+EPUB = Path("tests/data_files/Metamorphosis - Franz Kafka.epub")
+PDF = Path("tests/data_files/Calculus Made Easy - Silvanus Thompson.pdf")
+
+
+@pytest.mark.asyncio()
+class TestIdentifierMerging:
+    """A book's formats each contribute identifiers; none of them replaces the rest."""
+
+    async def test_every_format_contributes(self, monkeypatch: pytest.MonkeyPatch) -> None:
+        """
+        Identifiers are a collection, not a single value.
+
+        Merging the whole dict meant the last format to report won outright: an EPUB
+        declaring an ASIN, a Google volume id and a Calibre id kept none of them once
+        a PDF contributed one ISBN.
+        """
+
+        async def epub(_file):
+            return {
+                "title": "How Linux Works",
+                "identifiers": {"isbn-13": "9781718500419", "asin": "1718500408"},
+            }
+
+        async def pdf(_file):
+            return {"identifiers": {"isbn-13": "9781718500402", "isbn-10": "1593270356"}}
+
+        monkeypatch.setattr(EpubExtractor, "extract_metadata", epub)
+        monkeypatch.setattr(PdfExtractor, "extract_metadata", pdf)
+
+        metadata = await Extractor.extract_metadata(
+            [Path("How Linux Works.epub"), Path("How Linux Works.pdf")]
+        )
+
+        assert metadata["identifiers"] == {
+            # Declared by the publisher's toolchain, so it outranks the PDF's, which
+            # was scraped off a copyright page that also prints the print edition's.
+            "isbn-13": "9781718500419",
+            "asin": "1718500408",
+            "isbn-10": "1593270356",
+        }
+
+    async def test_a_second_format_that_finds_nothing_erases_nothing(self) -> None:
+        """The PDF fixture carries no ISBN, so it must leave the EPUB's alone."""
+        metadata = await Extractor.extract_metadata([EPUB, PDF])
+
+        assert metadata["identifiers"] == {"id": "http://www.gutenberg.org/5200"}
+
+    async def test_one_format_on_its_own_is_unaffected(self) -> None:
+        metadata = await Extractor.extract_metadata([EPUB])
+
+        assert metadata["identifiers"] == {"id": "http://www.gutenberg.org/5200"}
+
+    async def test_no_identifiers_anywhere_leaves_the_field_absent(self) -> None:
+        """An empty dict would count as extracted metadata and overwrite nothing."""
+        metadata = await Extractor.extract_metadata([PDF])
+
+        assert "identifiers" not in metadata
+
+
+class TestSplitEdition:
+    """An edition is a field on the book, not part of what the book is called."""
+
+    @pytest.mark.parametrize(
+        ("title", "stripped", "edition"),
+        [
+            # Every form below is one that turned up in a real library.
+            ("Fluent Python, 2nd Edition", "Fluent Python", 2),
+            ("Building Microservices, 2E", "Building Microservices", 2),
+            ("Digital Image Processing, 4e", "Digital Image Processing", 4),
+            (
+                "Network Security Essentials: Applications and Standards/6e",
+                "Network Security Essentials: Applications and Standards",
+                6,
+            ),
+            (
+                "Refactoring: Improving the Design of Existing Code (2nd edition)",
+                "Refactoring: Improving the Design of Existing Code",
+                2,
+            ),
+            # Ordinal words, including a qualifier sitting inside the statement.
+            (
+                "The Art of Computer Programming: Volume 1 / Fundamental Algorithms, Third Edition",
+                "The Art of Computer Programming: Volume 1 / Fundamental Algorithms",
+                3,
+            ),
+            (
+                "Introduction to the Theory of Computation, Third International Edition",
+                "Introduction to the Theory of Computation",
+                3,
+            ),
+            # Mid-title, before a subtitle and before a trailing author.
+            (
+                "How Linux Works, 3rd Edition: What Every Superuser Should Know",
+                "How Linux Works: What Every Superuser Should Know",
+                3,
+            ),
+            (
+                "Code Complete, 2nd Edition - Steve McConnell",
+                "Code Complete - Steve McConnell",
+                2,
+            ),
+            # An underscore between the number and the "e", beside an unnumbered
+            # qualifier that has nowhere to go in an integer column and so stays put.
+            (
+                "Cryptography and Network Security, Global Edition, 8_e - Stallings",
+                "Cryptography and Network Security, Global Edition - Stallings",
+                8,
+            ),
+        ],
+    )
+    def test_editions_are_split_out(self, title: str, stripped: str, edition: int) -> None:
+        assert split_edition(title) == (stripped, edition)
+
+    @pytest.mark.parametrize(
+        "title",
+        [
+            # A number alone is never an edition — these are titles.
+            "Catch 22",
+            "Fahrenheit 451",
+            "Blade Runner 2049",
+            "1984",
+            "Apollo 13",
+            "Slaughterhouse 5",
+            "The Art of Computer Programming: Volume 1",
+            # "Edition" with no number cannot be stored, so it stays where it can
+            # still be read.
+            "Cryptography and Network Security: Principles and Practice, Global Edition",
+            "Building Microservices",
+        ],
+    )
+    def test_titles_are_left_alone(self, title: str) -> None:
+        assert split_edition(title) == (title, None)
+
+    def test_a_title_that_is_only_an_edition_is_kept(self) -> None:
+        """Stripping must never leave a book with no title at all."""
+        assert split_edition("2nd Edition") == ("2nd Edition", None)
+
+    @pytest.mark.parametrize("title", ["", None])
+    def test_nothing_yields_nothing(self, title: str | None) -> None:
+        assert split_edition(title) == (title, None)
+
+
+@pytest.mark.asyncio()
+class TestEditionFromFiles:
+    async def test_extraction_moves_the_edition_off_the_title(self) -> None:
+        """The PDF fixture calls itself a 2nd edition in its own metadata title."""
+        metadata = await Extractor.extract_metadata([PDF])
+
+        assert metadata["title"] == "The Project Gutenberg eBook #33283: Calculus Made Easy"
+        assert metadata["edition"] == 2
+
+
+class TestEpubPublisher:
+    """The publisher was looked up and then dropped on the floor."""
+
+    def test_a_declared_publisher_is_returned(self) -> None:
+        """
+        The lookup discarded its own result and fell off the end of the function, so
+        every EPUB reported no publisher no matter what it said.
+        """
+        book = epub.EpubBook()
+        book.add_metadata("DC", "publisher", "No Starch Press")
+
+        assert EpubExtractor._extract_publisher(book) == "No Starch Press"
+
+    def test_no_publisher_is_none(self) -> None:
+        assert EpubExtractor._extract_publisher(epub.EpubBook()) is None

backend/tests/unit/test_services/test_calibre_import_service.py

@@ -0,0 +1,415 @@
+"""Tests for importing a Calibre library through BookService."""
+
+from pathlib import Path
+
+import pytest
+
+from chitai.config import settings
+from chitai.database import models as m
+from chitai.services import BookService
+from chitai.services.calibre import CalibreLibrary
+
+from tests.calibre_fixtures import CalibreFixture
+
+
+DATA_FILES = Path("tests/data_files")
+EPUB = DATA_FILES / "Metamorphosis - Franz Kafka.epub"
+OTHER_EPUB = DATA_FILES / "The Art of War - Sun Tzu.epub"
+PDF = DATA_FILES / "Calculus Made Easy - Silvanus Thompson.pdf"
+
+
+@pytest.fixture(name="calibre_root")
+def fx_calibre_root(tmp_path: Path) -> Path:
+    """Three books: one plain, one in two formats, one Chitai cannot use."""
+    fixture = CalibreFixture(tmp_path / "source")
+    fixture.add_pages_table()
+
+    fixture.add_book(
+        1,
+        "The Metamorphosis",
+        authors=["Franz Kafka"],
+        pubdate="1915-10-15 00:00:00+00:00",
+        tags=["Fiction", "Absurdist"],
+        publisher="Kurt Wolff Verlag",
+        languages=["deu"],
+        comment="<p>He wakes up <i>changed</i>.</p>",
+        identifiers={"isbn": "978-0-486-29030-0", "amazon": "B01N5IB20Q"},
+        uuid="11111111-2222-3333-4444-555555555555",
+        pages=201,
+        cover=True,
+        formats={"EPUB": EPUB},
+    )
+
+    fixture.add_book(
+        2,
+        "The Art of War",
+        authors=["Sun Tzu"],
+        series="Classics",
+        series_index=3.0,
+        formats={"EPUB": OTHER_EPUB, "PDF": PDF},
+    )
+
+    fixture.add_book(3, "Metadata Only", authors=["Nobody"])
+
+    return fixture.commit()
+
+
+async def library_of(root: Path) -> CalibreLibrary:
+    source = CalibreLibrary(root)
+    await source.open()
+    return source
+
+
+async def test_imports_a_catalogue(
+    books_service: BookService, test_library: m.Library, calibre_root: Path
+) -> None:
+    source = await library_of(calibre_root)
+    try:
+        result = await books_service.create_many_from_calibre(source, test_library)
+    finally:
+        await source.close()
+
+    assert result.total == 3
+    assert len(result.created) == 2
+
+    # The book with no files is left out: a record with nothing to read, and a directory
+    # to match, is worse than not importing it.
+    assert [skipped.calibre_id for skipped in result.skipped] == [3]
+    assert result.skipped[0].reason == "no files in the catalogue"
+    assert result.failed == []
+
+    book = await books_service.get(result.created[0])
+
+    assert book.title == "The Metamorphosis"
+    assert [author.name for author in book.authors] == ["Franz Kafka"]
+    assert sorted(tag.name for tag in book.tags) == ["Absurdist", "Fiction"]
+    assert book.publisher is not None and book.publisher.name == "Kurt Wolff Verlag"
+    assert book.published_date is not None and book.published_date.year == 1915
+    assert book.language == "deu"
+    assert book.pages == 201
+    assert book.cover_image is not None
+
+    # The HTML is gone; `Book.description` is rendered as text.
+    assert book.description == "He wakes up changed."
+
+
+async def test_two_formats_are_one_book(
+    books_service: BookService, test_library: m.Library, calibre_root: Path
+) -> None:
+    source = await library_of(calibre_root)
+    try:
+        result = await books_service.create_many_from_calibre(source, test_library)
+    finally:
+        await source.close()
+
+    book = await books_service.get(result.created[1])
+
+    assert book.title == "The Art of War"
+    assert sorted(Path(file.path).suffix for file in book.files) == [".epub", ".pdf"]
+
+    # A REAL series index reaches the column as the string everything else writes.
+    assert book.series is not None and book.series.title == "Classics"
+    assert book.series_position == "3"
+
+
+async def test_identifiers_are_folded_onto_chitai_schemes(
+    books_service: BookService, test_library: m.Library, calibre_root: Path
+) -> None:
+    """
+    `amazon` becomes `asin`, a hyphenated ISBN survives, and the Calibre uuid is kept.
+
+    The uuid is deliberately not stored under `uuid`, which duplicate matching ignores
+    because an EPUB regenerates one per build. Calibre's is stable, so it is the durable
+    link back to the row it came from.
+    """
+    source = await library_of(calibre_root)
+    try:
+        result = await books_service.create_many_from_calibre(source, test_library)
+    finally:
+        await source.close()
+
+    book = await books_service.get(result.created[0])
+    identifiers = {identifier.name: identifier.value for identifier in book.identifiers}
+
+    assert identifiers["asin"] == "B01N5IB20Q"
+    assert identifiers["isbn-13"] == "9780486290300"
+    assert identifiers["calibre-uuid"] == "11111111-2222-3333-4444-555555555555"
+
+    matching = {
+        identifier.name: identifier.normalized_value for identifier in book.identifiers
+    }
+
+    # Stored under its own name, matched under one scheme for both ISBN forms.
+    assert matching["isbn-13"] == "isbn:9780486290300"
+
+    # And the uuid carries a real matching key, which is the whole reason it is not
+    # filed under `uuid`.
+    assert matching["calibre-uuid"] is not None
+
+
+async def test_the_source_library_is_left_alone(
+    books_service: BookService, test_library: m.Library, calibre_root: Path
+) -> None:
+    """Files are copied. Moving them would leave `metadata.db` pointing at nothing."""
+    before = {
+        path: path.stat().st_mtime_ns
+        for path in sorted(calibre_root.rglob("*"))
+        if path.is_file()
+    }
+
+    source = await library_of(calibre_root)
+    try:
+        result = await books_service.create_many_from_calibre(source, test_library)
+    finally:
+        await source.close()
+
+    after = {
+        path: path.stat().st_mtime_ns
+        for path in sorted(calibre_root.rglob("*"))
+        if path.is_file()
+    }
+
+    assert after == before
+
+    # And the copies are really there, under the library's own layout.
+    for book_id in result.created:
+        book = await books_service.get(book_id)
+        for file in book.files:
+            assert (Path(book.path or "") / file.path).is_file()
+
+
+async def test_importing_twice_creates_nothing(
+    books_service: BookService, test_library: m.Library, calibre_root: Path
+) -> None:
+    """
+    Re-running is safe with no bookkeeping: the bytes are recognised wherever they sit.
+
+    This is what makes an interrupted import resumable by simply running it again.
+    """
+    for _ in range(2):
+        source = await library_of(calibre_root)
+        try:
+            result = await books_service.create_many_from_calibre(source, test_library)
+        finally:
+            await source.close()
+
+    assert result.created == []
+    assert sorted(skipped.reason for skipped in result.skipped) == [
+        "already stored",
+        "already stored",
+        "no files in the catalogue",
+    ]
+
+    held_by = [
+        skipped.book_id for skipped in result.skipped if skipped.reason == "already stored"
+    ]
+    assert all(book_id is not None for book_id in held_by)
+
+
+async def test_a_file_the_catalogue_lists_but_disk_does_not(
+    books_service: BookService, test_library: m.Library, tmp_path: Path
+) -> None:
+    """Calibre keeps the row when a file is moved away behind its back."""
+    fixture = CalibreFixture(tmp_path / "source")
+    fixture.add_book(1, "Present", authors=["A"], formats={"EPUB": EPUB})
+    fixture.add_book(2, "Absent", authors=["B"])
+    fixture.add_missing_format(2, "EPUB", "Absent - B")
+    root = fixture.commit()
+
+    source = await library_of(root)
+    try:
+        result = await books_service.create_many_from_calibre(source, test_library)
+    finally:
+        await source.close()
+
+    assert len(result.created) == 1
+    assert [(s.calibre_id, s.reason) for s in result.skipped] == [(2, "no files on disk")]
+
+
+async def test_one_broken_book_does_not_stop_the_import(
+    books_service: BookService, test_library: m.Library, tmp_path: Path
+) -> None:
+    """
+    A failure is recorded and the run continues, leaving no files behind for it.
+
+    An orphaned directory would make the next attempt reserve `title (2)` and look as
+    though it had worked.
+    """
+    fixture = CalibreFixture(tmp_path / "source")
+    fixture.add_book(1, "First", authors=["A"], formats={"EPUB": EPUB})
+    fixture.add_book(2, "Doomed", authors=["B"], formats={"EPUB": OTHER_EPUB})
+    fixture.add_book(3, "Third", authors=["C"], formats={"PDF": PDF})
+    root = fixture.commit()
+
+    original = books_service.create
+
+    async def fail_on_the_second(data, *args, **kwargs):
+        if isinstance(data, dict) and data.get("title") == "Doomed":
+            raise RuntimeError("no room on the shelf")
+        return await original(data, *args, **kwargs)
+
+    books_service.create = fail_on_the_second  # type: ignore[method-assign]
+
+    source = await library_of(root)
+    try:
+        result = await books_service.create_many_from_calibre(source, test_library)
+    finally:
+        await source.close()
+        books_service.create = original  # type: ignore[method-assign]
+
+    assert len(result.created) == 2
+    assert len(result.failed) == 1
+    assert result.failed[0].calibre_id == 2
+    assert "no room on the shelf" in result.failed[0].reason
+
+    # Nothing of the failed book was left in the library. Checked against the path the
+    # template would have produced, rather than by walking the root — the Calibre source
+    # sits under it in these tests, and its own files are meant to still be there.
+    assert not (Path(test_library.root_path) / "B").exists()
+
+    # And the books either side of it are where they should be.
+    for book_id in result.created:
+        book = await books_service.get(book_id)
+        for file in book.files:
+            assert (Path(book.path or "") / file.path).is_file()
+
+
+async def test_a_cover_that_cannot_be_read_is_not_fatal(
+    books_service: BookService, test_library: m.Library, tmp_path: Path
+) -> None:
+    """
+    A truncated `cover.jpg` costs the cover, not the book.
+
+    Real libraries hold them, from an interrupted download or a failed conversion, and
+    the cover is the one thing in the directory that can be replaced from the book page.
+    """
+    fixture = CalibreFixture(tmp_path / "source")
+    fixture.add_book(
+        1, "Unreadable Cover", authors=["A"], corrupt_cover=True, formats={"EPUB": EPUB}
+    )
+    root = fixture.commit()
+
+    source = await library_of(root)
+    try:
+        result = await books_service.create_many_from_calibre(source, test_library)
+    finally:
+        await source.close()
+
+    assert result.failed == []
+    assert len(result.created) == 1
+
+    book = await books_service.get(result.created[0])
+    assert book.cover_image is None
+    assert len(book.files) == 1
+
+
+async def test_shared_authors_and_tags_are_one_row_each(
+    books_service: BookService, test_library: m.Library, tmp_path: Path, session
+) -> None:
+    """Two books by one author must not produce two `Author` rows."""
+    fixture = CalibreFixture(tmp_path / "source")
+    fixture.add_book(
+        1, "One", authors=["Franz Kafka"], tags=["Fiction"], formats={"EPUB": EPUB}
+    )
+    fixture.add_book(
+        2, "Two", authors=["Franz Kafka"], tags=["Fiction"], formats={"EPUB": OTHER_EPUB}
+    )
+    root = fixture.commit()
+
+    source = await library_of(root)
+    try:
+        result = await books_service.create_many_from_calibre(source, test_library)
+    finally:
+        await source.close()
+
+    assert len(result.created) == 2
+
+    first, second = [await books_service.get(book_id) for book_id in result.created]
+
+    assert first.authors[0].id == second.authors[0].id
+    assert first.tags[0].id == second.tags[0].id
+
+
+async def test_a_second_copy_is_reported_not_refused(
+    books_service: BookService, test_library: m.Library, tmp_path: Path
+) -> None:
+    """
+    Two catalogue rows for one book, with different bytes, both import.
+
+    File-level dedupe cannot see it — the archives differ — so book-level detection
+    reports the pair and leaves the decision to the reader.
+    """
+    padded = tmp_path / "padded.epub"
+    padded.write_bytes(EPUB.read_bytes() + b"\0" * 64)
+
+    fixture = CalibreFixture(tmp_path / "source")
+    fixture.add_book(
+        1, "The Metamorphosis", authors=["Franz Kafka"], formats={"EPUB": EPUB}
+    )
+    fixture.add_book(
+        2, "The Metamorphosis", authors=["Franz Kafka"], formats={"EPUB": padded}
+    )
+    root = fixture.commit()
+
+    source = await library_of(root)
+    try:
+        result = await books_service.create_many_from_calibre(source, test_library)
+    finally:
+        await source.close()
+
+    assert len(result.created) == 2
+    assert len(result.possible_duplicates) == 1
+    assert result.possible_duplicates[0].candidates[0].book_id == result.created[0]
+
+
+async def test_allow_duplicates_stores_the_same_bytes_again(
+    books_service: BookService, test_library: m.Library, calibre_root: Path
+) -> None:
+    for allow in (False, True):
+        source = await library_of(calibre_root)
+        try:
+            result = await books_service.create_many_from_calibre(
+                source, test_library, allow_duplicates=allow
+            )
+        finally:
+            await source.close()
+
+    assert len(result.created) == 2
+
+
+async def test_duplicate_scope_off_imports_everything(
+    books_service: BookService,
+    test_library: m.Library,
+    calibre_root: Path,
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    monkeypatch.setattr(settings, "duplicate_scope", "off")
+
+    for _ in range(2):
+        source = await library_of(calibre_root)
+        try:
+            result = await books_service.create_many_from_calibre(source, test_library)
+        finally:
+            await source.close()
+
+    assert len(result.created) == 2
+
+
+async def test_progress_is_reported_per_book(
+    books_service: BookService, test_library: m.Library, calibre_root: Path
+) -> None:
+    """The import is long enough that its progress is the only thing worth watching."""
+    seen = []
+
+    source = await library_of(calibre_root)
+    try:
+        await books_service.create_many_from_calibre(
+            source, test_library, on_progress=seen.append
+        )
+    finally:
+        await source.close()
+
+    assert [progress.processed for progress in seen] == [1, 2, 3]
+    assert all(progress.total == 3 for progress in seen)
+    assert [progress.outcome for progress in seen] == ["created", "created", "skipped"]
+    assert seen[0].title == "The Metamorphosis"

docs/calibre-import.md

@@ -0,0 +1,407 @@
+# Implementation brief: importing a Calibre library
+
+Written for an agent picking this up cold. Read the repo-root `AGENTS.md` and
+`backend/AGENTS.md` first — this brief assumes both, particularly the **Filesystem behaviour**
+and **Duplicate detection** sections.
+
+Every claim about Calibre's schema and on-disk layout below was checked against a real library at
+`~/Documents/Calibre Library` (6 books, current Calibre). Where a fact came from Calibre's source
+rather than that library, it says so.
+
+## Feasibility: high, and most of the machinery already exists
+
+`metadata.db` is plain SQLite with a schema that has been stable for a decade, and the files sit
+beside it in a predictable tree. Chitai already has every piece an import needs:
+
+| Needed | Already in the tree |
+| --- | --- |
+| Ingest files that are already on disk | `BookService.create_many_from_existing_files` (`services/book.py:1280`) |
+| Deduplicate authors/tags/publishers/series | `_populate_with_unique_relationships` (`services/book.py:1752`), via `as_unique_async` |
+| Generic identifiers with a scheme map | `Identifier`, and `parse_identifier` (`services/metadata_extractor.py:58`) — whose map already covers `isbn`, `amazon`, `mobi-asin`, `google`, `goodreads`, `doi`, `calibre` |
+| Decide where a book lives on disk | `BookPathGenerator`, `_reserve_book_path` (`services/book.py:1066`) |
+| Not import the same book twice | `find_duplicate_files` (`:461`) and `find_duplicate_books` (`:535`) |
+| Store a cover | `_save_cover_image` (`:1994`) |
+
+So this is **a reader, not new ingest machinery**: turn Calibre rows into the metadata dict
+`BookService` already accepts, and hand it to a slightly generalised version of the consume-directory
+path. The parsing is the easy half.
+
+The hard parts are elsewhere, and all three are addressed below:
+
+1. Calibre libraries hold formats Chitai cannot describe, let alone read — and one of them
+   **currently 500s the book detail endpoint** (see prerequisites).
+2. A 5,000-book import is a long-running job, and the app has no job/progress concept.
+3. Whether files are **copied** into the library or **referenced in place** — which is a
+   product decision with a large blast radius, because in-place means Chitai's write paths point
+   at somebody's Calibre library.
+
+## What a Calibre library actually is
+
+```
+Calibre Library/
+├── metadata.db                       the whole catalogue
+├── metadata_db_prefs_backup.json     ignore
+├── .caltrash/  .calnotes/            ignore — deleted books still live in .caltrash
+└── <Author Name>/
+    └── <Title> (<book id>)/          == books.path
+        ├── cover.jpg                 iff books.has_cover
+        ├── metadata.opf              ignore; the db is authoritative
+        └── <data.name>.<format>      one per row in `data`
+```
+
+The tables that matter, and nothing else: `books`, `authors` + `books_authors_link`,
+`publishers` + `books_publishers_link`, `tags` + `books_tags_link`, `series` +
+`books_series_link`, `languages` + `books_languages_link`, `comments`, `identifiers`, `data`,
+`books_pages_link`, `last_read_positions`.
+
+Ten things that will produce wrong data if you do not know them:
+
+- **Never query the views.** `meta`, `tag_browser_*` and friends call SQLite functions Calibre
+  registers from Python at connection time. Verified: `SELECT * FROM meta` fails with
+  `no such function: sortconcat`. Query base tables only.
+
+- **`pubdate` has a sentinel, not a null.** An unknown publication date is stored as
+  `0101-01-01 00:00:00+00:00` (Calibre's `UNDEFINED_DATE`, year 101). It parses fine as a
+  `date`, so nothing will complain — two of the six books in the reference library carry it. Drop
+  any `pubdate` with year < 1000. The same sentinel appears in `timestamp`.
+
+- **`data.name` is lossy and is not the title.** It is the on-disk stem, truncated to Calibre's
+  filename limit and sanitised. Verified in the reference library: the book titled
+  `The Project Gutenberg eBook #33283: Calculus Made Easy, 2nd Edition` is stored as
+  `The Project Gutenberg eBook #33283_ Calcul - Silvanus Phillips Thompson.pdf`. So the file
+  extractors must not be consulted for metadata (see decision 2), and `books.path`/`data.name`
+  are for *locating* files only.
+
+- **`books.title` may contain characters Calibre strips from its own paths** — `:` became `_`
+  above, and titles legitimately contain `/` (`AC/DC`). `BookPathGenerator` interpolates the
+  title straight into a path and only collapses repeated slashes
+  (`services/filesystem_library.py`), so an unsanitised Calibre title can silently add a
+  directory level. Sanitise `/` and control characters out of `title` before path generation.
+
+- **`authors.name` escapes commas as `|`.** Calibre's `AuthorsTable` unserialises with
+  `name.replace('|', ',')` (from Calibre's `db/tables.py`; the reference library has no such
+  name, so this one is unverified locally). Do the same replacement, and pass `authors.name` —
+  **not** `authors.sort`, which is `Melville, Herman`. `format_author_name` would flip the sort
+  form correctly anyway, but there is no reason to hand it the worse input.
+
+- **`series_index` is a REAL.** `7.0` must become `"7"`, not `"7.0"` — `Book.series_position` is
+  a string, and `find_duplicate_books`'s series-position disqualifier compares it as one.
+
+- **`languages.lang_code` is ISO 639-2/B** (`eng`), while `EpubExtractor` stores raw
+  `DC:language` (`en`). Both will coexist in the column. `Book.language` is free text and the
+  edit form is a plain `<input>`, so nothing breaks; normalising to two letters is optional
+  polish, not part of this work.
+
+- **`comments.text` is HTML.** `Book.description` is rendered as plain text by
+  `CollapsibleText`, so `<p>` tags will show literally. Strip to text on import.
+
+- **`books_pages_link` is usually empty of real data.** It carries `needs_scan` and, in the
+  reference library, `pages = 0` for all six books. Only use it when `pages > 0`.
+
+- **`identifiers.type` is free text.** The reference library holds `isbn`, `amazon` and
+  `mobi-asin`, all of which `parse_identifier` already maps. Feed every identifier through it and
+  keep whatever survives; do not filter to a known list.
+
+## Decisions to settle before writing code
+
+1. **Copy files into the library. Do not move, do not reference in place** — for the first
+   version. Moving leaves `metadata.db` pointing at files that are gone, which quietly destroys a
+   library the user still uses. Referencing in place is genuinely desirable (nobody wants two
+   copies of 80 GB) but it points `book.path` at the Calibre tree, and `update_book` **moves
+   directories** while `delete_books` **deletes files** — so a metadata edit in Chitai would
+   rearrange somebody's Calibre library. `Library.read_only` exists but is enforced in exactly one
+   place (`services/library.py:54`, at creation). See phase 3.
+
+2. **Trust Calibre's metadata; do not run the extractors.** Calibre's catalogue is curated, its
+   filenames are truncated garbage, and running `Extractor.extract_metadata` over thousands of
+   files means opening every EPUB and rendering a cover page from every PDF. Take the cover from
+   `cover.jpg` directly. The one exception worth allowing: fill `pages` from the file when
+   Calibre has no useful value, behind a flag, off by default.
+
+3. ~~**The source is a server-side path, not an upload.**~~ **Reversed in review, and the reason
+   this brief was wrong is worth keeping.** The premise — "the library lives on the same host as
+   the backend in every realistic deployment" — is false for the common case: Calibre is a desktop
+   application, and its library is on the desktop. So the split is by *surface*, not by preference:
+
+   - **Over HTTP: an uploaded zip only.** `…/imports/calibre/upload`. There is no endpoint taking a
+     server path; one was built and then removed deliberately.
+   - **On the server: a path only.** `litestar calibre-import <path>`, which is where a very large
+     library or a headless migration belongs — an upload has to carry the whole archive across
+     first.
+
+   A CLI taking a path needs no justification. An *endpoint* taking one would have: it would let any
+   authenticated caller read any directory the backend can, and `TODO.md` records there is no
+   authorization tier at all. Not adding it is one less thing to gate later.
+
+4. **Import into an existing Chitai library**, chosen by the caller. Creating a library is
+   already one action, and the Calibre tree is rewritten by `BookPathGenerator` regardless.
+
+5. **Re-running an import must be safe, and file-level dedupe already makes it so.** The same
+   bytes are recognised by `(hash, size)` whatever their path, so a second run over the same
+   library skips everything. No import bookkeeping is needed for idempotency.
+
+## Prerequisite: a `.mobi` file breaks the book endpoint — **done**
+
+> Landed ahead of the import itself. `guess_content_type` in `services/utils.py` now names every
+> format from its extension, the column keeps a null when nothing can name one, and the OPDS feed
+> substitutes `application/octet-stream` at the one place a string is required. The Read control is
+> driven by `isReadable` rather than by the file count. See the section below for why it mattered,
+> and `backend/AGENTS.md` for the rule as it now stands.
+
+
+`FileMetadataRead.content_type` is a required `str` (`schemas/book.py:33`), but every ingest path
+fills it from `mimetypes.guess_type`, which returns `None` for `.mobi`, `.azw`, `.fb2`, `.lit`
+and `.htmlz` (verified). `FileMetadata.content_type` is nullable in the model, so the row stores
+fine and then fails response validation on the way out — a book whose only file is a MOBI would
+be unreadable through the API.
+
+Nothing in the tree hits this today because the browser upload path is used with EPUBs and PDFs.
+A Calibre library is full of MOBI and AZW3. Fix it first, either way round:
+
+- make the schema field `str | None`, and/or
+- add a small extension→MIME table for the ebook formats `mimetypes` does not know
+  (`application/x-mobipocket-ebook`, `application/vnd.amazon.ebook`, `application/x-fictionbook+xml`).
+
+Do both, in fact: the table is the right answer for OPDS clients, which choose an acquisition link
+by MIME type, and the nullable field is the safety net.
+
+**Related, but not a blocker:** Chitai reads EPUB and PDF only. `openBookInReader`
+(`book/[bookId]/+page.svelte:66`) branches on `getFileType(...) === 'EPUB' | 'PDF'` and does
+nothing for anything else, so an AZW3-only book gets a Read button that silently fails. Importing
+those files is still right — they are downloadable and they are the user's — but the button
+should be disabled for a book with no readable file. One `$derived` on the page, worth doing in
+the same branch.
+
+## Design
+
+### 1. `services/calibre.py` — a pure reader, no Chitai types
+
+```python
+@dataclass(frozen=True)
+class CalibreFile:
+    path: Path          # absolute, resolved against the library root
+    format: str         # "EPUB", as stored
+    size: int           # data.uncompressed_size, for a cheap sanity check
+
+@dataclass(frozen=True)
+class CalibreBook:
+    calibre_id: int
+    uuid: str
+    title: str
+    authors: list[str]
+    ...                 # one field per row of the mapping table below
+    cover: Path | None
+    files: list[CalibreFile]
+
+class CalibreLibrary:
+    def __init__(self, root: Path) -> None: ...
+    async def open(self) -> None: ...          # copy + connect, see below
+    async def books(self) -> AsyncIterator[CalibreBook]: ...
+    async def close(self) -> None: ...
+```
+
+Deliberately knows nothing about `Book`, `BookService` or the session — it is a file-format
+reader, unit-testable against a fixture database with no Postgres and no app.
+
+Two implementation notes:
+
+- **Copy `metadata.db` to a temp file and read the copy.** Calibre may be running and writing;
+  opening the live file read-only either sees a torn state or needs the `-wal` sidecar. The
+  database is small (438 KB for six books, single-digit MB for thousands), so a copy costs
+  nothing and removes the whole problem.
+- **`sqlite3` inside `asyncio.to_thread`, not a new dependency.** The connection is used for a
+  handful of queries. Do not add `aiosqlite` for this.
+
+Read the whole catalogue in **one query per table** and join in Python — six or so `SELECT`s and
+a few dicts, versus a per-book N+1 across ten tables. At self-hosted scale the entire catalogue
+minus descriptions fits in memory comfortably; if `comments.text` for 20k books is a concern,
+fetch that one table per batch.
+
+### 2. The mapping
+
+| Calibre | Chitai | Notes |
+| --- | --- | --- |
+| `books.title` | `title` | Sanitise `/` and control chars for path generation. `Extractor.format_book_title` may still be worth applying to split a subtitle at the second colon — but **do not** run `split_edition`, Calibre's title is the curated one. |
+| `authors.name` via `books_authors_link` | `authors` | `\|` → `,`. Order by `books_authors_link.id`; `Book.author_links` is an `ordering_list`, so insertion order is the displayed order. |
+| `comments.text` | `description` | Strip HTML to text. |
+| `books.pubdate` | `published_date` | Drop the year-101 sentinel. |
+| `series.name`, `books.series_index` | `series`, `series_position` | `7.0` → `"7"`. |
+| `tags.name` | `tags` | |
+| `publishers.name` | `publisher` | `books_publishers_link` is unique per book. |
+| `languages.lang_code` (lowest `item_order`) | `language` | Chitai holds one. |
+| `identifiers.type` / `.val` | `identifiers` | Through `parse_identifier`; keep what survives. |
+| `books.uuid` | `identifiers["calibre-uuid"]` | The one durable link back to the source row. `normalize_identifier` returns a key for it (it is not in `_PER_BUILD_NAMES`), which is *desirable*: a book re-imported from the same Calibre library matches on it exactly. |
+| `books_pages_link.pages` | `pages` | Only when `> 0`. |
+| `cover.jpg` when `has_cover` | `cover_image` | |
+| `data` rows | `files` | |
+| `books.timestamp` | — | `Book.created_at` is audit-managed; do not fight it. |
+| `ratings`, `annotations`, `custom_columns` | — | No home in the model. Out of scope. |
+| `last_read_positions` | `BookProgress` | Phase 2. |
+
+### 3. The ingest
+
+> **As built, this went on `BookService` as `create_many_from_calibre`, not into a separate
+> `services/calibre_import.py`.** The orchestration needs `_reserve_book_path`,
+> `_save_cover_image`, `_screen_for_duplicates` and `_record_possible_duplicates`, and reaching
+> into four privates from another module is worse than one more method in the file where the other
+> two ingest paths already live. `_record_possible_duplicates` was changed to take the list it
+> appends to rather than an `ImportResult`, so every ingest path can share it whatever its own
+> result type is. Two other deviations: `CalibreLibrary.books()` returns a list rather than an
+> async iterator, because the caller needs the total up front anyway; and the reader reports
+> identifiers exactly as Calibre keyed them, with the fold onto Chitai's schemes done by the
+> importer, which keeps the reader free of Chitai imports.
+
+Per book, in this order — it mirrors `create_many_from_existing_files`, which is the closest
+existing shape:
+
+1. `fingerprint_file` each source file (`services/utils.py:164`).
+2. `find_duplicate_files` against the target library. All files known → skip the book entirely,
+   recording it. Some known → import the rest.
+3. Build the metadata dict from the `CalibreBook`.
+4. `_reserve_book_path(path_gen.generate_path(data))`.
+5. **Copy** each file to `parent / _unused_path(...)`, building `FileMetadata` from the
+   fingerprint already computed. `services/utils.py` has `move_file` but no copy — add
+   `copy_file` beside it, streaming through `aiofiles` in `CHUNK_SIZE` blocks like
+   `_save_book_files` does, not `shutil.copy` (a 40 MB blocking read inside the event loop).
+6. Cover: open `cover.jpg` with PIL and hand the `Image` to `_save_cover_image`, which already
+   accepts one and converts to WebP.
+7. `super().create(data)` through `BookService`, then `find_duplicate_books` and record
+   candidates — same as `_record_possible_duplicates` (`services/book.py:1253`).
+8. **Commit per book.** A 5,000-book import inside one transaction is one failure away from
+   nothing, and per-book commits are what lets the library page show books arriving — which is
+   the behaviour commit `85367da` deliberately built.
+
+Report an `ImportResult`-shaped outcome; reuse `ImportResult` itself if it fits, extending it with
+a `failures: list[tuple[int, str]]` keyed by Calibre id. **One book must never fail the run** —
+a missing file, an unreadable cover or a `NOT NULL` violation gets recorded and skipped.
+
+### 4. Progress, and where the import runs
+
+> **As built, the HTTP surface is an uploaded archive and nothing else** — see decision 3.
+>
+> `POST …/imports/calibre/upload` takes a zipped library. The job owns the temp directory it is
+> unpacked into and deletes it when it ends. Extraction refuses zip slip, an archive too big for the
+> disk, and one with no `metadata.db` within three levels — all answered 400 before a job exists.
+>
+> This forced a fix to the SvelteKit proxy, which buffered request bodies with `arrayBuffer()`:
+> survivable for one book, not for a multi-gigabyte archive. POST and PATCH now stream
+> `request.body` through with `duplex: 'half'`.
+>
+> **A preview endpoint was built and then removed with the path route.** It read a server-side
+> catalogue and reported its size before writing anything, which is only useful when the caller
+> named a directory. An upload has already been carried across by the time anything can be read, so
+> unpacking it *is* the validation step — an archive that is not a Calibre library is refused there.
+
+The import outlives its request, so the handler starts it and returns a handle:
+
+- `POST /libraries/{library_id:int}/imports/calibre` — body `{path, copy_files: true}`, returns
+  `{job_id, total}`. 202.
+- `GET /libraries/imports/{job_id}` — `{state, total, processed, created, skipped, failed,
+  current_title, errors}`.
+- `DELETE /libraries/imports/{job_id}` — cancel; the task checks a flag between books.
+
+Keep the registry **in memory**, a `dict[str, ImportJob]` on a module-level singleton, with the
+task created by `asyncio.create_task`. This matches what the app already does — the consume
+watcher is an in-process singleton started from a lifespan hook — and it is roughly thirty lines
+against a model, a migration and a service for the alternative.
+
+State that limitation explicitly in the docstring: **it assumes one worker process.** The
+production `CMD` is `litestar run`, which is single-process, so this holds today; `TODO.md`
+already records that the production image should move to uvicorn with a worker count, and doing
+that would mean a poll landing on a worker that has never heard of the job. The consume watcher
+has the same problem, so this is not a new constraint — but the next person to add workers needs
+to find it written down. If import *history* is ever wanted, that is when an `import_jobs` table
+earns its migration.
+
+**Also add a CLI entry point.** A 200 GB library imported through a browser tab that must stay
+open is a bad experience, and `pyproject.toml` already declares a `chitai` script. A Litestar CLI
+command (`litestar --app-dir src/chitai/ calibre-import <path> --library <slug>`) is ~20 lines
+over the same service and is the right tool for the initial migration, which is the case this
+whole feature exists for. The endpoint is for people who would rather click.
+
+### 5. Frontend
+
+Model it on the duplicates screen, which is the closest precedent in shape and placement:
+
+- Route `(root)/settings/libraries/[libraryId]/import` — beside
+  `settings/libraries/[libraryId]/duplicates`, reached from the library settings page.
+- `getCalibreImport` (a `query`) and `cancelCalibreImport` (a `command`) in
+  `src/lib/api/calibre-import.remote.ts`, re-exported from `src/lib/api/index.ts`. **Starting an
+  import is not a remote function**: the archive goes to the backend through the proxy so the
+  browser streams straight through, where a remote function would put the whole thing through the
+  SvelteKit process first. The screen uses `XMLHttpRequest` for it, which is the only way to get
+  upload progress.
+- A file input, an upload progress bar, then a progress bar polling `getCalibreImport` every second
+  or two, a running count, and the failures listed at the end with their Calibre ids.
+- Finish with a link to the library's duplicates screen. An import into a non-empty library is
+  the single most likely way to produce duplicate books, and that screen already handles them.
+
+Do **not** route this through the upload tray. The tray reports on a client-driven queue it owns
+(`upload-queue.svelte.ts`); this is server-side work whose state survives a page reload, and
+conflating the two would mean teaching the tray to poll.
+
+Regenerate `src/lib/schema/openapi/schema.d.ts` against a backend running **your** branch —
+a stale server silently writes a stale file.
+
+## Testing
+
+The fixture is the interesting part. Build a Calibre library in a `tmp_path` fixture rather than
+committing a binary `metadata.db`: a helper that executes the subset of Calibre's `CREATE TABLE`
+statements (they are in this document's shape, and in any real library's `sqlite_master`), inserts
+a handful of books, and lays out `<Author>/<Title> (id)/` directories containing the existing
+EPUB and PDF fixtures from `backend/tests/data_files/` plus a copy of `cover.jpg`. Generated
+beats committed here because the tests need to assert on *odd* rows — the pubdate sentinel, a
+`|` in an author name, a title with a colon — and those are clearer written in Python than hidden
+in a blob.
+
+- **Unit** (`tests/unit/test_calibre.py`) — the reader alone: field mapping; the year-101 pubdate
+  dropped; `series_index` 7.0 → `"7"`; `|` unescaped in an author name; HTML stripped from
+  `comments`; `pages = 0` ignored; identifiers passed through `parse_identifier`; a `data` row
+  whose file is missing from disk reported rather than raised; `.caltrash` never walked.
+- **Service** (`tests/unit/test_services/test_calibre_import.py`) — a book with two formats lands
+  as one record with two files; the source files still exist afterwards; a second run over the
+  same library creates nothing; a library with one broken book imports the rest; authors and tags
+  shared between two books produce one `Author` / `Tag` row each; `possible_duplicates` reported
+  when the target library already holds the same book.
+- **Integration** (`tests/integration/test_calibre_import.py`) — `POST` returns 202 with a job id,
+  polling reaches a terminal state, and the books are then listable through `GET /books`. A
+  `.mobi`-only book must come back from `GET /books/{id}` without a 500 — that is the
+  prerequisite's regression test.
+
+`pytest` needs Docker (`pytest-databases`).
+
+## Verification
+
+```bash
+nix-shell                                                    # postgres + migrations applied
+cd backend
+pytest tests/                                                # take your own baseline first
+uv run litestar --app-dir src/chitai/ run --port 8001        # for the OpenAPI regeneration
+cd ../frontend && pnpm check                                 # baseline: 30 errors, 1 warning, 8 files
+```
+
+Neither `pnpm check` nor `pnpm lint` is clean on this repo — baseline before assuming an error is
+yours.
+
+End to end, against a real library (`~/Documents/Calibre Library` will do): import into an empty
+Chitai library and confirm the six books arrive with their covers, authors, tags, series
+positions and identifiers intact; that the AZW3 book is listed and downloadable; that the source
+library is byte-for-byte untouched (`diff -r` a copy taken beforehand); and that re-running the
+import creates nothing and reports six skipped books. Then import the same library into a
+library that already holds one of those books by upload, and confirm it lands on the duplicates
+screen rather than as a second copy.
+
+## Phasing
+
+| Phase | Scope |
+| --- | --- |
+| **0** ✅ | The `content_type` prerequisite, plus the Read button. **Done** — see below. |
+| **1** ✅ | `services/calibre.py`, the ingest, the CLI command, copy-only. **Done** — a headless one-time migration works today. |
+| **2** ◐ | The upload endpoint, the job registry and the settings screen — **done**. The HTTP surface is an uploaded zip only; see decision 3, which this reversed. `last_read_positions` → `BookProgress` is **not** done: it needs a Calibre-user → Chitai-user mapping, and the answer differs between the endpoint (which has a `current_user`) and the CLI (which has none). That decision is the next thing to make. |
+| **3** | Reference-in-place import. Its real content is **enforcing `Library.read_only`** across `update_book`, `delete_books`, `add_files` and `remove_files` — which is a feature of its own and should not be smuggled in under an import. |
+
+## Out of scope
+
+Annotations and highlights (no model to put them in), custom columns, ratings, virtual libraries
+and saved searches → bookshelves, format conversion, writing anything back to Calibre, and any
+form of continuing two-way sync. This is a one-way migration.

docs/duplicates-in-library-settings.md

@@ -0,0 +1,146 @@
+# Implementation brief: move Duplicates into library settings
+
+Written for an agent picking this up cold. Read the repo-root `AGENTS.md` and
+`frontend/AGENTS.md` first — this brief assumes both.
+
+**This is a frontend-only change.** The backend already scopes everything by library
+(`GET /books/duplicate-books?library_id=`), so no endpoint, schema or migration is
+involved.
+
+## Where this starts from
+
+The duplicates review screen exists and works. It currently lives at
+
+```
+frontend/src/routes/(root)/(library)/library/[libraryId]/duplicates/
+    +page.server.ts     loads the groups, plus the full Book records merge needs
+    +page.svelte        group cards, "Not duplicates", "Merge…"
+```
+
+and is reached from a **Duplicates entry in the main sidebar**
+(`frontend/src/lib/components/layout/nav-main.svelte`), which is what this change
+removes.
+
+Settings today is a flat, entirely global four-item nav
+(`frontend/src/routes/(root)/settings/+layout.svelte`): Account, Appearance, Libraries,
+Devices. `settings/libraries/+page.svelte` is a single table of every library whose rows
+link *out* to the library itself. **There is nowhere that means "settings for this
+library"** — that is the gap this change fills.
+
+## What to build — option B
+
+Libraries expands in the settings nav. Every library is a sub-item; selecting one swaps
+the pane; Duplicates is a section inside that pane. All libraries and all their sections
+end up one click apart.
+
+```
+/settings/libraries                        the existing table (leave it as the index)
+/settings/libraries/[libraryId]            redirects to the first section
+/settings/libraries/[libraryId]/duplicates the review screen, moved
+```
+
+Suggested files:
+
+| Path | What |
+| --- | --- |
+| `settings/libraries/[libraryId]/+layout.svelte` | Library name, and the section tabs |
+| `settings/libraries/[libraryId]/+page.ts` | `redirect(303, …/duplicates)` |
+| `settings/libraries/[libraryId]/duplicates/+page.server.ts` | Moved verbatim |
+| `settings/libraries/[libraryId]/duplicates/+page.svelte` | Moved verbatim |
+
+Duplicates is the **only** real section today. Build the tab strip so General and Danger
+zone have somewhere obvious to land, but do not invent them now — an empty tab is worse
+than no tab.
+
+## The nav
+
+In `settings/+layout.svelte`, `items` is a flat `as const` array matched on
+`page.route.id`. Libraries needs to render its children beneath it:
+
+```svelte
+{#each libraryState.libraries as library (library.id)}
+  <a href={resolve('/(root)/settings/libraries/[libraryId]/duplicates', {
+        libraryId: String(library.id) })}> … </a>
+{/each}
+```
+
+`getLibraryState()` **is** available under `/settings` — it is set in
+`(root)/+layout.svelte`, above the settings group, and `settings/libraries/+page.svelte`
+already uses it. No new load function is needed to list the libraries.
+
+**Active state is matched on route id, not pathname.** There is a comment in
+`settings/+layout.svelte` explaining why: `resolve()` returns an absolute path on the
+client and a relative one during SSR, so a pathname comparison is false on the server and
+true after hydration, and the highlight flashes in. A nested library item is active when
+the route id matches **and** `page.params.libraryId === String(library.id)` — both, or
+every library lights up at once.
+
+## Things that will bite
+
+1. **Remove the sidebar entry in the same change.** `nav-main.svelte` gained a
+   `Duplicates` item and a `CopyCheck` import when the screen was built. Delete both, and
+   delete the old route directory. Doing the removal and the move together is the point —
+   split across two commits the screen is unreachable in between.
+
+2. **Delete the old route, do not leave it.** Two live copies of a screen that both write
+   is how they drift.
+
+3. **`setBookSelectionState` is not available under `/settings`.** It is set in
+   `(root)/(library)/+layout.svelte`, which the settings group is not inside. This is
+   fine — the duplicates page uses `BookImage` directly, not `book-thumbnail.svelte`, and
+   `MergeBooks` takes its `libraryId` as a prop. **Verify this stays true** if you touch
+   either component; a `getBookSelectionState()` under settings returns `undefined` and
+   fails at the first access, not at import.
+
+4. **Keep `depends('app:duplicate-books')`.** Both the dismiss action and `MergeBooks`
+   call `invalidate('app:duplicate-books')` to make a resolved group leave the screen.
+   Drop it and the page silently stops refreshing. `MergeBooks` also invalidates
+   `app:books`, which is a no-op under settings and should stay that way.
+
+5. **The settings shell is height-constrained.** `settings/+layout.svelte` is
+   `h-[calc(100vh-var(--header-height)-2rem)]` with `overflow-auto` on the content pane.
+   The review screen is a long list of cards — it must scroll *inside* that pane. Its
+   current `mx-auto max-w-5xl` wrapper will want revisiting.
+
+6. **Three levels of nav is option B's known cost.** Nav → library → section, and the
+   pane is narrower than the full-width route the screen was designed against. The group
+   cards are `w-36` covers in a wrapping flex row, so they reflow, but check a group of
+   four at a narrow window before calling it done.
+
+7. **`resolve()` must be a direct call in markup** for `svelte/no-navigation-without-resolve`.
+   Where `nav-main.svelte` computes a url through a variable it carries an
+   `eslint-disable-next-line`; prefer the direct call over inheriting that.
+
+8. **The loader depends on the `?ids=` fix.** `+page.server.ts` fetches full `Book`
+   records with `listBooks({ ids, pageSize })` because the merge workbench needs
+   identifiers, description and publisher, which `DuplicateBookRead` does not carry.
+   advanced_alchemy's stock id filter types that parameter as `list[str]` regardless of
+   config, which made Postgres refuse `bigint = character varying`; the override lives in
+   `backend/src/chitai/services/dependencies.py` (`create_book_filter_dependencies`).
+   If `GET /books?ids=1&ids=2` 500s, that override is missing — do not work around it in
+   the loader.
+
+## Out of scope
+
+The **General** and **Danger zone** sections (rename, path template, read-only, consume
+directory, delete), and any change to the merge workbench itself. The toolbar entry point
+for merge — select 2+ books in the library view — is unrelated and stays where it is.
+
+## Verification
+
+```bash
+cd frontend
+pnpm check     # baseline: 30 errors, 1 warning, 8 files — none of them yours
+pnpm lint      # not clean either; check the files you touched, not the tree
+pnpm build
+```
+
+By hand, with a library that has a duplicate group:
+
+- Settings → Libraries lists every library beneath it; clicking one opens its pane.
+- Duplicates shows the same groups the old route did, and scrolls inside the settings pane.
+- **Not duplicates** removes the group and it stays gone after a reload.
+- **Merge…** opens the workbench, merges, and the group leaves the screen.
+- The main sidebar no longer has a Duplicates entry, and
+  `/library/<id>/duplicates` no longer resolves.
+- A library with no duplicates shows the empty state, not a blank pane.

frontend/.prettierignore

@@ -11,3 +11,6 @@ coverage
 
 # Miscellaneous
 /static/
+
+# Vendored third-party source, copied verbatim by scripts/vendor-foliate.sh
+/src/lib/vendor/

frontend/AGENTS.md

@@ -4,13 +4,16 @@ SvelteKit web app for the eBook library. See the repo-root `AGENTS.md` for the o
 dev-environment setup.
 
 **Stack:** SvelteKit 2 with `adapter-node` · Svelte 5 (runes) · Tailwind v4 · Zod v4 ·
-`epubjs` · `mode-watcher` (dark mode) · `svelte-sonner` (toasts) · pnpm.
+vendored `foliate-js` (EPUB) · vendored `pdf.js` (PDF) · `mode-watcher` (dark mode) ·
+`svelte-sonner` (toasts) · pnpm.
 
 Two experimental flags are on in `svelte.config.js` and the codebase depends on both:
 `kit.experimental.remoteFunctions` and `compilerOptions.experimental.async` (`await` in components).
 
-Tailwind v4 has **no config file** — the theme, oklch colour tokens and `@custom-variant dark` all
-live in `src/app.css`.
+Tailwind v4 has **no config file** — the theme, colour tokens (hex, not oklch) and
+`@custom-variant dark` all live in `src/app.css`. `dark` is the _only_ custom variant defined, so
+generated components that assume others — shadcn's slider ships `data-horizontal:` / `data-vertical:`
+classes — silently produce no styles. Use the `data-[orientation=…]` form instead.
 
 ## Talking to the backend
 
@@ -65,8 +68,12 @@ Svelte context with a module-level `Symbol` key and a `setXState` / `getXState`
 
 ```ts
 const LIBRARY_KEY = Symbol('LIBRARY');
-export function setLibraryState(libraries: Library[]) { return setContext(LIBRARY_KEY, new LibraryState(libraries)); }
-export function getLibraryState() { return getContext<ReturnType<typeof setLibraryState>>(LIBRARY_KEY); }
+export function setLibraryState(libraries: Library[]) {
+	return setContext(LIBRARY_KEY, new LibraryState(libraries));
+}
+export function getLibraryState() {
+	return getContext<ReturnType<typeof setLibraryState>>(LIBRARY_KEY);
+}
 ```
 
 Follow that pattern rather than introducing stores. `library.svelte.ts` is the reference — including
@@ -79,10 +86,56 @@ its optimistic-delete-with-rollback and toast handling. `bookCollection` / `book
   `@ieedan/shadcn-svelte-extras` (`jsrepo.json`). Treat as generated: add components with the CLIs
   rather than hand-writing them, and prefer wrapping over editing.
 - App components live in `forms/`, `layout/`, `view/` (browser, grid/list/table, filters, sort) and
-  `reader/` (epub reader + chapter sidebar).
+  `reader/` (see [The readers](#the-readers)).
 - `cn()` from `$lib/utils` merges Tailwind classes; the `WithElementRef` / `WithoutChild` helpers
   there are the shadcn prop-typing conventions.
 
+## The readers
+
+**PDF** is the pdf.js viewer vendored under `static/pdfjs/`, pointed at by an iframe. Untouched by
+the EPUB work; leave it alone unless the task is about PDFs.
+
+**EPUB** is built on `foliate-js`, copied verbatim into `src/lib/vendor/foliate-js/` by
+`scripts/vendor-foliate.sh` (pinned commit; see `src/lib/vendor/foliate-js/README.chitai.md`).
+Upstream has no npm release and recommends a submodule; this repo has none and already vendors
+pdf.js the same way, so it is copied instead. Only the import closure reachable from `view.js` is
+vendored, and **`pdf.js` in that directory is our stub, not upstream's** — the real one imports a
+bare `@pdfjs/pdf.min.mjs` that Rollup resolves at build time even though the path never runs.
+
+Layout:
+
+| Path                                       | What                                                                         |
+| ------------------------------------------ | ---------------------------------------------------------------------------- |
+| `lib/vendor/foliate-js/`                   | The engine. Do not edit — `vendor-foliate.sh` overwrites it.                 |
+| `lib/reader/foliate.ts`                    | Lazy loader for the custom elements. The only thing that imports `$foliate`. |
+| `lib/reader/settings.ts` · `stylesheet.ts` | Defaults/bounds, and the CSS injected into the book.                         |
+| `lib/reader/progress.ts`                   | Debounced progress writer with a `sendBeacon` flush.                         |
+| `lib/state/reader-settings.svelte.ts`      | Settings state, persisted to `localStorage`.                                 |
+| `components/reader/foliate-view.svelte`    | Wraps `<foliate-view>`; owns the imperative lifecycle.                       |
+| `components/reader/epub-reader.svelte`     | The shell: chrome, TOC, errors, progress.                                    |
+
+Things that will bite:
+
+- **`$foliate` is a Vite-only alias.** It is deliberately absent from `kit.alias` and tsconfig
+  `paths` so TypeScript cannot resolve it and falls back to the ambient declaration in
+  `lib/reader/foliate-js.d.ts`; `src/lib/vendor` is also in tsconfig `exclude`. Without both,
+  `checkJs` walks ~11k lines of untyped JS. The declaration file must **not** be named `foliate.d.ts`
+  — beside `foliate.ts`, TypeScript takes it for that file's emitted declaration and drops it.
+- **Never import the vendored code at module scope.** `view.js` calls `customElements.define` and
+  subclasses `HTMLElement` on import, so it must stay behind `loadFoliate()` inside `onMount`. SSR is
+  otherwise on for the reader route.
+- **Sections render in iframes, which swallow key events.** Keyboard handlers are bound per section
+  document on the `load` event, and modifier combinations are replayed onto the host window so app
+  shortcuts (the sidebar's ctrl+B) still work while reading.
+- **Renderer settings split two ways.** Flow, gap, margins, column count and line width are
+  _attributes_ set with `setAttribute` (there is no JS property API, no `margin` shorthand and no
+  `spread` — a spread is `max-column-count: 2`). Typography is CSS passed to `renderer.setStyles`,
+  which takes a `[before, after]` pair: the first is prepended to the section head so the book
+  overrides it, the second appended so it wins. User settings belong in the second, with
+  `!important`, or the book's own CSS beats them.
+- **Progress needs no locations pre-pass.** `relocate` carries both a CFI and an overall `fraction`,
+  which map straight onto `epub_cfi` and `percentage`.
+
 ## Routing
 
 Route groups carry the layout structure:
@@ -95,21 +148,31 @@ Route groups carry the layout structure:
 ## Conventions
 
 Prettier (`.prettierrc`): tabs, single quotes, no trailing commas, 100 columns, with the Svelte and
-Tailwind plugins. Run `pnpm check` (svelte-check) and `pnpm lint` before considering work done.
+Tailwind plugins. Run `pnpm check` (svelte-check) and `pnpm lint` before considering work done —
+but take a baseline first, because neither is clean (see below).
+
+`src/lib/vendor/` is excluded from Prettier, ESLint and svelte-check. Don't reformat vendored code.
 
 ## Known rough edges
 
 Observed in the current tree — don't mistake these for intentional patterns to copy:
 
-- `src/lib/schema/openapi/schema.d.ts` is **stale** — it predates the `BookProgress` rework and has
-  no `BookProgressRead`, so `book.progress` types as `{}`. This is the source of most of the ~104
-  errors `pnpm check` reports on a clean tree; regenerating it should clear them. Get a baseline
-  before assuming an error is yours.
+- `pnpm check` is not clean. **Baseline as of 2026-08-13: 30 errors, 1 warning, 8 files**, most of
+  them in `src/routes/api/[...path]/+server.ts` (see below). Get your own baseline before assuming
+  an error is yours. `src/lib/schema/openapi/schema.d.ts` was regenerated on that date and is
+  current; regenerate it again after any backend API change, with
+  `pnpm exec openapi-typescript http://localhost:8000/schema/openapi.json -o src/lib/schema/openapi/schema.d.ts`
+  against a backend running **your** branch — a stale server silently writes a stale file.
+- `pnpm lint` does not pass either — 131 pre-existing ESLint errors, 35 of them
+  `svelte/no-navigation-without-resolve` on plain `href`s, plus ~59 files Prettier would rewrite
+  (mostly vendored shadcn components). Check the files you touched, not the whole tree.
 - `src/routes/api/[...path]/+server.ts` — all four handlers are annotated `RequestHandler` while the
-  import of that type is commented out at line 4.
-- `src/app.d.ts` — `App.Locals["user"]` is typed from `lucide-svelte`'s `User` *icon* component
+  import of that type is commented out at line 4. It also buffers whole **responses** with
+  `arrayBuffer()` and forwards no `Range` header, so book downloads are not streamed. **Requests**
+  are streamed — POST and PATCH pass `request.body` through with `duplex: 'half'` (see `bodyOf`),
+  because a zipped Calibre library upload cannot be held in this process. The response side is
+  still buffered; see `TODO.md`.
+- `src/app.d.ts` — `App.Locals["user"]` is typed from `lucide-svelte`'s `User` _icon_ component
   rather than the `User` interface in `$lib/server/auth`.
-- Uncommitted work in progress (as of 2026-08-10): library icons, spanning
-  `components/ui/icon-picker/`, the newly vendored `components/ui/popover/`,
-  `forms/library-create-form.svelte`, `layout/library-switcher.svelte` and `schema/library.ts`.
-  Prefer not to refactor those files mid-flight.
+- No CSP, which foliate's README asks for because EPUBs can carry scripts. See `TODO.md` for why it
+  is not enabled yet.

frontend/eslint.config.js

@@ -12,6 +12,8 @@ const gitignorePath = fileURLToPath(new URL('./.gitignore', import.meta.url));
 
 export default defineConfig(
 	includeIgnoreFile(gitignorePath),
+	// Vendored third-party source. Tracked, so .gitignore does not cover it.
+	{ ignores: ['src/lib/vendor/**'] },
 	js.configs.recommended,
 	...ts.configs.recommended,
 	...svelte.configs.recommended,
@@ -27,6 +29,12 @@ export default defineConfig(
 			'no-undef': 'off'
 		}
 	},
+	{
+		// Generated shadcn components take href as a prop and cannot resolve it —
+		// that is the caller's job. Editing them here would be lost on regeneration.
+		files: ['src/lib/components/ui/**'],
+		rules: { 'svelte/no-navigation-without-resolve': 'off' }
+	},
 	{
 		files: ['**/*.svelte', '**/*.svelte.ts', '**/*.svelte.js'],
 		languageOptions: {

frontend/package.json

@@ -33,8 +33,8 @@
 		"jsrepo": "^2.5.2",
 		"openapi-typescript": "^7.13.0",
 		"prettier": "^3.8.1",
-		"prettier-plugin-svelte": "^3.5.1",
-		"prettier-plugin-tailwindcss": "^0.6.14",
+		"prettier-plugin-svelte": "^3.5.2",
+		"prettier-plugin-tailwindcss": "^0.8.1",
 		"svelte": "^5.53.7",
 		"svelte-check": "^4.4.5",
 		"tailwind-merge": "^3.5.0",
@@ -47,7 +47,7 @@
 		"vite": "^7.3.1"
 	},
 	"dependencies": {
-		"epubjs": "^0.3.93",
+		"construct-style-sheets-polyfill": "^3.1.0",
 		"mode-watcher": "^1.1.0",
 		"svelte-sonner": "^1.0.8",
 		"zod": "^4.3.6"

frontend/scripts/vendor-foliate.sh

@@ -0,0 +1,129 @@
+#!/usr/bin/env bash
+#
+# Vendors foliate-js into src/lib/vendor/foliate-js/.
+#
+# foliate-js has no npm release and no build step; upstream recommends a git
+# submodule. We copy instead, because this repo has no submodules (pdf.js is
+# vendored the same way under static/pdfjs) and because a submodule would drag
+# in 231 files / 13 MB, of which 191 files / 12 MB is a bundled pdf.js build we
+# deliberately do not use — Chitai serves PDFs through static/pdfjs/web/viewer.html.
+#
+# Only the files reachable from view.js are copied: 15 upstream files, ~656 KB.
+# pdf.js is NOT copied; a stub is written in its place (see below).
+#
+# To update: bump FOLIATE_SHA, re-run, review the diff, then smoke-test the
+# reader — paginator.js is ~3800 lines of gesture and animation code and this
+# fork is pushed to frequently.
+#
+# Usage:  ./scripts/vendor-foliate.sh
+set -euo pipefail
+
+FOLIATE_REPO="https://github.com/readest/foliate-js.git"
+FOLIATE_SHA="63a2eb1fc1e4813c4e849ccdb3d4be2c54a35869"
+
+DEST="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)/src/lib/vendor/foliate-js"
+TMP="$(mktemp -d)"
+trap 'rm -rf "$TMP"' EXIT
+
+# The reachable closure from view.js. Everything else upstream ships is either
+# unreachable (dict.js, opds.js, footnotes.js, quote-image.js, uri-template.js),
+# a demo (reader.js), or build tooling (rollup.config.js, eslint.config.js).
+FILES=(
+	view.js
+	epub.js
+	epubcfi.js
+	paginator.js
+	fixed-layout.js
+	overlayer.js
+	progress.js
+	search.js
+	text-walker.js
+	tts.js
+	mobi.js
+	comic-book.js
+	fb2.js
+	vendor/zip.js
+	vendor/fflate.js
+)
+
+echo "Cloning $FOLIATE_REPO @ ${FOLIATE_SHA:0:7} ..."
+git clone --quiet --filter=blob:none --no-checkout "$FOLIATE_REPO" "$TMP/foliate"
+git -C "$TMP/foliate" checkout --quiet "$FOLIATE_SHA"
+
+rm -rf "$DEST"
+mkdir -p "$DEST/vendor"
+
+for f in "${FILES[@]}"; do
+	if [ ! -f "$TMP/foliate/$f" ]; then
+		echo "ERROR: $f is missing upstream at ${FOLIATE_SHA:0:7}." >&2
+		echo "The file list in this script is stale; re-check the import graph." >&2
+		exit 1
+	fi
+	cp "$TMP/foliate/$f" "$DEST/$f"
+done
+
+cp "$TMP/foliate/LICENSE" "$DEST/LICENSE"
+
+# view.js does `await import('./pdf.js')` inside makeBook. That is a static-string
+# dynamic import, so Rollup resolves it at build time whether or not the code path
+# ever runs — and upstream's pdf.js opens with `import '@pdfjs/pdf.min.mjs'`, a bare
+# specifier that does not resolve here. Shipping this stub at that path keeps the
+# build working without a Vite alias, and without vendoring 12 MB of pdf.js.
+cat > "$DEST/pdf.js" <<'STUB'
+// NOT upstream foliate-js. See README.chitai.md.
+//
+// Chitai renders PDFs with the pdf.js viewer vendored at static/pdfjs/, so
+// foliate's PDF backend is not vendored. view.js still references this module
+// from makeBook via a static-string dynamic import, which Rollup resolves at
+// build time regardless of whether it executes — so the file has to exist.
+//
+// Throwing at module scope surfaces a legible message in the reader's error
+// card if a PDF is ever routed to the EPUB reader by mistake, rather than a
+// TypeError from `globalThis.pdfjsLib` being undefined.
+throw new Error('foliate-js PDF rendering is not enabled in Chitai');
+STUB
+
+cat > "$DEST/README.chitai.md" <<EOF
+# Vendored foliate-js
+
+Do not edit these files. They are copied verbatim from upstream by
+\`frontend/scripts/vendor-foliate.sh\`; local changes are lost on the next run.
+
+| | |
+| --- | --- |
+| Upstream | <https://github.com/readest/foliate-js> (Readest's fork of johnfactotum/foliate-js) |
+| Pinned commit | \`$FOLIATE_SHA\` |
+| Licence | MIT — see \`LICENSE\` |
+
+Readest's fork is used rather than upstream for its paginator work: touch/swipe
+turn handling, fixed-layout spread centring, and a malformed-XHTML fallback in
+\`loadDocument\`.
+
+## What is here
+
+Only the import closure reachable from \`view.js\`. Not vendored, because nothing
+reaches them: \`dict.js\`, \`opds.js\`, \`footnotes.js\`, \`quote-image.js\`,
+\`uri-template.js\`, \`reader.js\` (upstream's demo), and the build configs.
+
+## pdf.js is ours, not upstream's
+
+\`pdf.js\` in this directory is a **stub that throws**. Upstream's version imports
+\`@pdfjs/pdf.min.mjs\` — a bare specifier backed by a 12 MB vendored pdf.js build —
+and \`view.js\` reaches it through \`await import('./pdf.js')\`, which Rollup resolves
+at build time even though Chitai never takes that path. Chitai serves PDFs from
+\`static/pdfjs/web/viewer.html\` instead.
+
+To enable foliate's PDF backend, add \`pdf.js\` and \`vendor/pdfjs/\` to the file list
+in the vendor script and drop the stub.
+
+## Updating
+
+Bump \`FOLIATE_SHA\` in \`frontend/scripts/vendor-foliate.sh\`, re-run it, review the
+diff, and smoke-test the reader — \`paginator.js\` is ~3800 lines of gesture and
+animation code and this fork is pushed to frequently.
+EOF
+
+echo
+echo "Vendored ${#FILES[@]} files + LICENSE + pdf.js stub + README.chitai.md to:"
+echo "  $DEST"
+du -sh "$DEST" | sed 's/^/  /'

frontend/src/app.css

@@ -8,74 +8,100 @@
 
 :root {
 	--radius: 0.625rem;
-	--background: oklch(1 0 0);
-	--foreground: oklch(0.129 0.042 264.695);
-	--card: oklch(1 0 0);
-	--card-foreground: oklch(0.129 0.042 264.695);
-	--popover: oklch(1 0 0);
-	--popover-foreground: oklch(0.129 0.042 264.695);
-	--primary: oklch(0.208 0.042 265.755);
-	--primary-foreground: oklch(0.984 0.003 247.858);
-	--secondary: oklch(0.968 0.007 247.896);
-	--secondary-foreground: oklch(0.208 0.042 265.755);
-	--muted: oklch(0.968 0.007 247.896);
-	--muted-foreground: oklch(0.554 0.046 257.417);
-	--accent: oklch(0.968 0.007 247.896);
-	--accent-foreground: oklch(0.208 0.042 265.755);
-	--destructive: oklch(0.577 0.245 27.325);
-	--border: oklch(0.929 0.013 255.508);
-	--input: oklch(0.929 0.013 255.508);
-	--ring: oklch(0.704 0.04 256.788);
-	--chart-1: oklch(0.646 0.222 41.116);
-	--chart-2: oklch(0.6 0.118 184.704);
-	--chart-3: oklch(0.398 0.07 227.392);
-	--chart-4: oklch(0.828 0.189 84.429);
-	--chart-5: oklch(0.769 0.188 70.08);
-	--sidebar: oklch(0.984 0.003 247.858);
-	--sidebar-foreground: oklch(0.129 0.042 264.695);
-	--sidebar-primary: oklch(0.208 0.042 265.755);
-	--sidebar-primary-foreground: oklch(0.984 0.003 247.858);
-	--sidebar-accent: oklch(0.968 0.007 247.896);
-	--sidebar-accent-foreground: oklch(0.208 0.042 265.755);
-	--sidebar-border: oklch(0.929 0.013 255.508);
-	--sidebar-ring: oklch(0.704 0.04 256.788);
+
+	/* Typography — system stacks, so nothing depends on a CDN or a webfont build. */
+	--app-font-sans: system-ui, -apple-system, 'Segoe UI', Roboto, 'Helvetica Neue', Arial, sans-serif;
+	--app-font-serif: Georgia, 'Iowan Old Style', 'Times New Roman', serif;
+	--app-font-mono: ui-monospace, 'SF Mono', 'Cascadia Mono', Menlo, Consolas, monospace;
+
+	/* Reading Room — light */
+	--background: #e9ebee;
+	--foreground: #1b1f24;
+	--card: #fbfbfc;
+	--card-foreground: #1b1f24;
+	--popover: #fbfbfc;
+	--popover-foreground: #1b1f24;
+	--primary: #1f5f5b;
+	--primary-foreground: #f2f7f6;
+	--secondary: #dfe3e8;
+	--secondary-foreground: #1b1f24;
+	--muted: #e2e5e9;
+	--muted-foreground: #7d858f;
+	--accent: #d7e5e3;
+	--accent-foreground: #164743;
+	--destructive: #a6402f;
+	--border: #d2d6dc;
+	--input: #d2d6dc;
+	--ring: #1f5f5b;
+
+	/* Semantic — deliberately not the accent, so state never reads as branding. */
+	--success: #2f7d4f;
+	--success-foreground: #f2f7f6;
+	--flag: #b08a1e;
+	--star: #c79a25;
+
+	--chart-1: #1f5f5b;
+	--chart-2: #2f7d4f;
+	--chart-3: #b08a1e;
+	--chart-4: #4c6b8a;
+	--chart-5: #a6402f;
+
+	--sidebar: #e2e5e9;
+	--sidebar-foreground: #1b1f24;
+	--sidebar-primary: #1f5f5b;
+	--sidebar-primary-foreground: #f2f7f6;
+	--sidebar-accent: #d7e5e3;
+	--sidebar-accent-foreground: #164743;
+	--sidebar-border: #d2d6dc;
+	--sidebar-ring: #1f5f5b;
 }
 
 .dark {
-	--background: oklch(0.129 0.042 264.695);
-	--foreground: oklch(0.984 0.003 247.858);
-	--card: oklch(0.208 0.042 265.755);
-	--card-foreground: oklch(0.984 0.003 247.858);
-	--popover: oklch(0.208 0.042 265.755);
-	--popover-foreground: oklch(0.984 0.003 247.858);
-	--primary: oklch(0.929 0.013 255.508);
-	--primary-foreground: oklch(0.208 0.042 265.755);
-	--secondary: oklch(0.279 0.041 260.031);
-	--secondary-foreground: oklch(0.984 0.003 247.858);
-	--muted: oklch(0.279 0.041 260.031);
-	--muted-foreground: oklch(0.704 0.04 256.788);
-	--accent: oklch(0.279 0.041 260.031);
-	--accent-foreground: oklch(0.984 0.003 247.858);
-	--destructive: oklch(0.704 0.191 22.216);
-	--border: oklch(1 0 0 / 10%);
-	--input: oklch(1 0 0 / 15%);
-	--ring: oklch(0.551 0.027 264.364);
-	--chart-1: oklch(0.488 0.243 264.376);
-	--chart-2: oklch(0.696 0.17 162.48);
-	--chart-3: oklch(0.769 0.188 70.08);
-	--chart-4: oklch(0.627 0.265 303.9);
-	--chart-5: oklch(0.645 0.246 16.439);
-	--sidebar: oklch(0.208 0.042 265.755);
-	--sidebar-foreground: oklch(0.984 0.003 247.858);
-	--sidebar-primary: oklch(0.488 0.243 264.376);
-	--sidebar-primary-foreground: oklch(0.984 0.003 247.858);
-	--sidebar-accent: oklch(0.279 0.041 260.031);
-	--sidebar-accent-foreground: oklch(0.984 0.003 247.858);
-	--sidebar-border: oklch(1 0 0 / 10%);
-	--sidebar-ring: oklch(0.551 0.027 264.364);
+	/* Reading Room — dark */
+	--background: #15191b;
+	--foreground: #e6eae9;
+	--card: #1d2226;
+	--card-foreground: #e6eae9;
+	--popover: #1d2226;
+	--popover-foreground: #e6eae9;
+	--primary: #6fbab0;
+	--primary-foreground: #0e1a19;
+	--secondary: #232a2d;
+	--secondary-foreground: #e6eae9;
+	--muted: #232a2d;
+	--muted-foreground: #78868a;
+	--accent: #1b3330;
+	--accent-foreground: #9fd8d0;
+	--destructive: #e0796b;
+	--border: #2a3034;
+	--input: #2a3034;
+	--ring: #6fbab0;
+
+	--success: #4fa97a;
+	--success-foreground: #0e1a19;
+	--flag: #d4a63a;
+	--star: #e5b84b;
+
+	--chart-1: #6fbab0;
+	--chart-2: #4fa97a;
+	--chart-3: #d4a63a;
+	--chart-4: #7f9dc0;
+	--chart-5: #e0796b;
+
+	--sidebar: #111517;
+	--sidebar-foreground: #e6eae9;
+	--sidebar-primary: #6fbab0;
+	--sidebar-primary-foreground: #0e1a19;
+	--sidebar-accent: #1b3330;
+	--sidebar-accent-foreground: #9fd8d0;
+	--sidebar-border: #2a3034;
+	--sidebar-ring: #6fbab0;
 }
 
 @theme inline {
+	--font-sans: var(--app-font-sans);
+	--font-serif: var(--app-font-serif);
+	--font-mono: var(--app-font-mono);
 	--radius-sm: calc(var(--radius) - 4px);
 	--radius-md: calc(var(--radius) - 2px);
 	--radius-lg: var(--radius);
@@ -95,6 +121,10 @@
 	--color-accent: var(--accent);
 	--color-accent-foreground: var(--accent-foreground);
 	--color-destructive: var(--destructive);
+	--color-success: var(--success);
+	--color-success-foreground: var(--success-foreground);
+	--color-flag: var(--flag);
+	--color-star: var(--star);
 	--color-border: var(--border);
 	--color-input: var(--input);
 	--color-ring: var(--ring);

frontend/src/app.d.ts

@@ -2,7 +2,8 @@
 // for information about these interfaces
 
 import type { ApiClient } from '$lib/server/api';
-import type { User } from 'lucide-svelte';
+import type { User } from '$lib/server/auth';
+import type { ThemeConfig } from '$lib/theme/presets';
 
 declare global {
 	namespace App {
@@ -11,6 +12,7 @@ declare global {
 			authToken: string | null;
 			api: ApiClient;
 			user: User;
+			theme: ThemeConfig;
 		}
 		// interface PageData {}
 		// interface PageState {}

frontend/src/app.html

@@ -4,6 +4,7 @@
 		<meta charset="utf-8" />
 		<meta name="viewport" content="width=device-width, initial-scale=1" />
 		%sveltekit.head%
+		<!--theme-->
 	</head>
 	<body data-sveltekit-preload-data="hover" class="overflow-hidden">
 		<div style="display: contents">%sveltekit.body%</div>

frontend/src/hooks.server.ts

@@ -1,5 +1,6 @@
 import { ApiClient } from '$lib/server/api';
 import { validateToken } from '$lib/server/auth';
+import { THEME_COOKIE, parseThemeCookie, themeToCss } from '$lib/theme/presets';
 import { redirect, type Handle } from '@sveltejs/kit';
 import { sequence } from '@sveltejs/kit/hooks';
 
@@ -34,4 +35,28 @@ const protectedRoutesHandle: Handle = async ({ event, resolve }) => {
 	return resolve(event);
 };
 
-export const handle = sequence(authHandle, protectedRoutesHandle);
+/**
+ * Inline the stored theme into the document head.
+ *
+ * The palette has to be in the very first byte of HTML the browser paints,
+ * otherwise every page load flashes the default theme before hydration swaps
+ * it. The `<!--theme-->` placeholder in app.html is the injection point.
+ */
+const themeHandle: Handle = async ({ event, resolve }) => {
+	const config = parseThemeCookie(event.cookies.get(THEME_COOKIE));
+	event.locals.theme = config;
+
+	return resolve(event, {
+		// The placeholder comment is kept, not replaced. Svelte 5 uses HTML
+		// comments as hydration markers, so SvelteKit warns when a chunk comes
+		// back with fewer comments than it went in with — removing this one is
+		// enough to trip that check.
+		transformPageChunk: ({ html }) =>
+			html.replace(
+				'<!--theme-->',
+				`<!--theme--><style id="chitai-theme">${themeToCss(config)}</style>`
+			)
+	});
+};
+
+export const handle = sequence(themeHandle, authHandle, protectedRoutesHandle);

frontend/src/lib/api/book.remote.ts

@@ -8,12 +8,31 @@ import {
 	deleteBooksSchema,
 	editBookMetadataSchema,
 	updateBookProgressSchema,
+	duplicateDismissalSchema,
+	bookMergeSchema,
 	type Book,
+	type BooksUploadResult,
+	type DuplicateBookGroup,
 	bookFilesUpload
 } from '$lib/schema/index';
 import { stringCoerce, type PaginatedResponse } from '$lib/schema/common';
 import { createQueryParams } from '$lib/utils';
 
+/**
+ * The backend's own message for a failed response, rather than its JSON envelope.
+ *
+ * A refused duplicate answers 409 with a `detail` worth reading and the offending
+ * files in `extra`; passing the body through whole puts JSON in front of the reader.
+ */
+function detailOf(body: string): string {
+	try {
+		const parsed = JSON.parse(body);
+		return typeof parsed?.detail === 'string' ? parsed.detail : body;
+	} catch {
+		return body;
+	}
+}
+
 export const getBook = query(stringCoerce, async (id): Promise<Book> => {
 	const { locals } = getRequestEvent();
 
@@ -68,7 +87,9 @@ export const updateBookCover = form(bookCoverUpload, async ({ book_id, file }) =
 	return await response.json();
 });
 
-export const uploadBooks = form(booksUpload, async ({ library_id, files }) => {
+export const uploadBooks = form(
+	booksUpload,
+	async ({ library_id, files }): Promise<BooksUploadResult> => {
 		const { locals } = getRequestEvent();
 
 		const formData = new FormData();
@@ -87,7 +108,8 @@ export const uploadBooks = form(booksUpload, async ({ library_id, files }) => {
 		}
 
 		return await response.json();
-});
+	}
+);
 
 export const uploadBookFiles = form(bookFilesUpload, async ({ book_id, files }) => {
 	const { locals } = getRequestEvent();
@@ -100,8 +122,9 @@ export const uploadBookFiles = form(bookFilesUpload, async ({ book_id, files })
 	const response = await locals.api.postMultipart(`/books/${book_id}/files`, formData);
 
 	if (!response.ok) {
-		const message = await response.text();
-		error(response.status, message);
+		// 409 here means the file is already stored under a different book, which is
+		// something the reader can act on — so the message has to survive the trip.
+		error(response.status, detailOf(await response.text()));
 	}
 
 	return await response.json();
@@ -133,6 +156,65 @@ export const deleteBookFiles = command(deleteBookFilesSchema, async ({ book_id,
 	}
 });
 
+/**
+ * Books already in the library that look like copies of one another.
+ *
+ * Metadata only, so every group is a question rather than a verdict — which is why
+ * the screen it feeds offers a way to disagree.
+ */
+export const listDuplicateBooks = query(
+	stringCoerce,
+	async (libraryId): Promise<DuplicateBookGroup[]> => {
+		const { locals } = getRequestEvent();
+
+		const response = await locals.api.get(`/books/duplicate-books?library_id=${libraryId}`);
+
+		if (!response.ok) error(response.status, detailOf(await response.text()));
+
+		return await response.json();
+	}
+);
+
+/**
+ * Fold several books into one and delete the records folded in.
+ *
+ * Irreversible, so the caller is expected to have shown what is about to happen.
+ */
+export const mergeBooks = command(
+	bookMergeSchema,
+	async ({ library_id, ...data }): Promise<Book> => {
+		const { locals } = getRequestEvent();
+
+		const response = await locals.api.post(`/books/merge?library_id=${library_id}`, data);
+
+		if (!response.ok) error(response.status, detailOf(await response.text()));
+
+		return await response.json();
+	}
+);
+
+/** Record that two books are not the same book, so the pair stops being proposed. */
+export const dismissDuplicateBooks = command(duplicateDismissalSchema, async (data) => {
+	const { locals } = getRequestEvent();
+
+	const response = await locals.api.post('/books/duplicate-books/dismissals', data);
+
+	if (!response.ok) error(response.status, detailOf(await response.text()));
+});
+
+/** Undo a dismissal, so the pair is proposed again. */
+export const restoreDuplicateBooks = command(duplicateDismissalSchema, async (data) => {
+	const { locals } = getRequestEvent();
+
+	const params = createQueryParams(data);
+
+	const response = await locals.api.delete(
+		`/books/duplicate-books/dismissals?${params.toString()}`
+	);
+
+	if (!response.ok) error(response.status, detailOf(await response.text()));
+});
+
 export const updateBookProgress = command(
 	updateBookProgressSchema,
 	async ({ book_ids, ...data }) => {

frontend/src/lib/api/calibre-import.remote.ts

@@ -0,0 +1,51 @@
+import { command, getRequestEvent, query } from '$app/server';
+import { error } from '@sveltejs/kit';
+
+import { stringCoerce } from '$lib/schema/common';
+import type { CalibreImport } from '$lib/schema/library';
+
+/** The backend's own message for a failed response, rather than its JSON envelope. */
+function detailOf(body: string): string {
+	try {
+		const parsed = JSON.parse(body);
+		return typeof parsed?.detail === 'string' ? parsed.detail : body;
+	} catch {
+		return body;
+	}
+}
+
+/**
+ * Where an import has got to.
+ *
+ * A `query` rather than a `command` so it can be refreshed, but it is polled on a timer
+ * rather than cached — the answer changes on its own.
+ *
+ * Starting an import is deliberately **not** here: the archive goes straight to the
+ * backend through the proxy, so it never passes through this process. See the import
+ * screen's `upload`.
+ */
+export const getCalibreImport = query(stringCoerce, async (jobId): Promise<CalibreImport> => {
+	const { locals } = getRequestEvent();
+
+	const response = await locals.api.get(`/libraries/imports/${jobId}`);
+
+	if (!response.ok) error(response.status, detailOf(await response.text()));
+
+	return await response.json();
+});
+
+/**
+ * Ask an import to stop after the book it is on.
+ *
+ * Not an abort: a book abandoned mid-copy would leave files on disk with no row
+ * describing them. Whatever it has imported stays imported.
+ */
+export const cancelCalibreImport = command(stringCoerce, async (jobId): Promise<CalibreImport> => {
+	const { locals } = getRequestEvent();
+
+	const response = await locals.api.delete(`/libraries/imports/${jobId}`);
+
+	if (!response.ok) error(response.status, detailOf(await response.text()));
+
+	return await response.json();
+});

frontend/src/lib/api/index.ts

@@ -2,6 +2,7 @@ export * from './auth.remote';
 export * from './author.remote';
 export * from './book.remote';
 export * from './bookshelf.remote';
+export * from './calibre-import.remote';
 export * from './library.remote';
 export * from './publisher.remote';
 export * from './tag.remote';

frontend/src/lib/components/forms/book-drop-zone.svelte

@@ -0,0 +1,223 @@
+<script lang="ts">
+	import { FileUp, Folder, FileText } from '@lucide/svelte';
+
+	import { Button } from '$lib/components/ui/button/index.js';
+	import type { FileRejectedReason } from '$lib/components/ui/file-drop-zone';
+	import { cn } from '$lib/utils';
+
+	let {
+		onUpload,
+		onFileRejected,
+		accept,
+		maxFileSize,
+		disabled = false,
+		class: className
+	}: {
+		onUpload: (files: File[]) => Promise<void> | void;
+		onFileRejected?: (opts: { reason: FileRejectedReason; file: File }) => void;
+		/** Comma separated extensions and/or MIME types, as the `accept` attribute takes. */
+		accept?: string;
+		/** Bytes. */
+		maxFileSize?: number;
+		disabled?: boolean;
+		class?: string;
+	} = $props();
+
+	/**
+	 * Two inputs rather than one.
+	 *
+	 * `webkitdirectory` is not a filter — it switches the picker into folder mode,
+	 * so a single input can offer files or folders but never both. The drop target
+	 * has no such constraint and stays one area.
+	 */
+	let fileInput = $state<HTMLInputElement>();
+	let folderInput = $state<HTMLInputElement>();
+
+	let dragging = $state(false);
+	let busy = $state(false);
+
+	const active = $derived(!disabled && !busy);
+
+	function accepts(file: File): FileRejectedReason | undefined {
+		if (maxFileSize !== undefined && file.size > maxFileSize) return 'Maximum file size exceeded';
+		if (!accept) return undefined;
+
+		const name = file.name.toLowerCase();
+		const type = file.type.toLowerCase();
+
+		const ok = accept
+			.split(',')
+			.map((pattern) => pattern.trim().toLowerCase())
+			.some((pattern) => {
+				// Match on the pattern, not the file's type. Testing `type` here is
+				// what makes MOBI fail: browsers report no MIME type for it, so a
+				// ".mobi" rule never gets compared against the filename.
+				if (pattern.startsWith('.')) return name.endsWith(pattern);
+				if (pattern.endsWith('/*')) return type.startsWith(pattern.slice(0, -1));
+				return type === pattern;
+			});
+
+		return ok ? undefined : 'File type not allowed';
+	}
+
+	/** readEntries hands back at most 100 at a time and signals the end with an empty batch. */
+	function readAll(reader: FileSystemDirectoryReader): Promise<FileSystemEntry[]> {
+		return new Promise((resolve, reject) => {
+			const entries: FileSystemEntry[] = [];
+
+			const next = () =>
+				reader.readEntries((batch) => {
+					if (batch.length === 0) return resolve(entries);
+					entries.push(...batch);
+					next();
+				}, reject);
+
+			next();
+		});
+	}
+
+	/**
+	 * Flattens a dropped entry into files, naming each one with its path inside the
+	 * dropped folder so it matches what the folder picker puts in
+	 * `webkitRelativePath` — which is what the upload form reads to keep structure.
+	 */
+	async function walk(entry: FileSystemEntry, prefix = ''): Promise<File[]> {
+		if (entry.isFile) {
+			const file = await new Promise<File>((resolve, reject) =>
+				(entry as FileSystemFileEntry).file(resolve, reject)
+			);
+			return [new File([file], `${prefix}${file.name}`, { type: file.type })];
+		}
+
+		if (entry.isDirectory) {
+			const entries = await readAll((entry as FileSystemDirectoryEntry).createReader());
+			const nested = await Promise.all(entries.map((e) => walk(e, `${prefix}${entry.name}/`)));
+			return nested.flat();
+		}
+
+		return [];
+	}
+
+	async function handleDrop(event: DragEvent) {
+		event.preventDefault();
+		dragging = false;
+		if (!active) return;
+
+		// Read the entries synchronously: DataTransfer is emptied as soon as this
+		// handler yields, so awaiting first loses everything that was dropped.
+		const entries = Array.from(event.dataTransfer?.items ?? [])
+			.filter((item) => item.kind === 'file')
+			.map((item) => item.webkitGetAsEntry())
+			.filter((entry): entry is FileSystemEntry => entry !== null);
+
+		// Older engines expose no entries; fall back to the flat list, which cannot
+		// contain folders anyway.
+		if (entries.length === 0) {
+			await submit(Array.from(event.dataTransfer?.files ?? []));
+			return;
+		}
+
+		busy = true;
+		try {
+			const nested = await Promise.all(entries.map((entry) => walk(entry)));
+			await submit(nested.flat());
+		} finally {
+			busy = false;
+		}
+	}
+
+	async function handleChange(event: Event) {
+		const input = event.currentTarget as HTMLInputElement;
+		const chosen = Array.from(input.files ?? []);
+		// Reset first so picking the same file twice still fires a change event.
+		input.value = '';
+		await submit(chosen);
+	}
+
+	async function submit(candidates: File[]) {
+		const accepted: File[] = [];
+
+		for (const file of candidates) {
+			const reason = accepts(file);
+			if (reason) onFileRejected?.({ file, reason });
+			else accepted.push(file);
+		}
+
+		if (accepted.length > 0) await onUpload(accepted);
+	}
+</script>
+
+<div
+	role="group"
+	aria-label="Add books"
+	aria-disabled={!active}
+	ondragover={(e) => {
+		e.preventDefault();
+		if (active) dragging = true;
+	}}
+	ondragleave={() => (dragging = false)}
+	ondrop={handleDrop}
+	class={cn(
+		'flex flex-col items-center gap-3 rounded-lg border-2 border-dashed border-border bg-accent/20 p-6 text-center transition-colors',
+		dragging && 'border-primary bg-accent/50',
+		!active && 'opacity-50',
+		className
+	)}
+>
+	<div
+		class="flex size-12 place-items-center justify-center rounded-full border border-dashed border-border text-muted-foreground"
+	>
+		<FileUp class="size-5" />
+	</div>
+
+	<div class="flex flex-col gap-0.5">
+		<span class="font-medium">
+			{busy ? 'Reading folder…' : 'Drop books here'}
+		</span>
+		<span class="text-sm text-muted-foreground">A folder keeps its structure</span>
+	</div>
+
+	<span class="font-mono text-[10px] tracking-widest text-muted-foreground uppercase">or</span>
+
+	<div class="flex flex-wrap justify-center gap-2">
+		<Button
+			type="button"
+			variant="outline"
+			size="sm"
+			disabled={!active}
+			onclick={() => fileInput?.click()}
+		>
+			<FileText class="size-4" />
+			Choose files
+		</Button>
+		<Button
+			type="button"
+			variant="outline"
+			size="sm"
+			disabled={!active}
+			onclick={() => folderInput?.click()}
+		>
+			<Folder class="size-4" />
+			Choose folder
+		</Button>
+	</div>
+
+	<input
+		bind:this={fileInput}
+		type="file"
+		multiple
+		{accept}
+		class="hidden"
+		onchange={handleChange}
+	/>
+	<!-- webkitdirectory is why this needs to be a second input: it turns the
+	     picker into a folder chooser rather than filtering what it accepts. -->
+	<input
+		bind:this={folderInput}
+		type="file"
+		multiple
+		webkitdirectory
+		class="hidden"
+		onchange={handleChange}
+	/>
+</div>

frontend/src/lib/components/forms/books-upload.svelte

@@ -1,165 +1,231 @@
 <script lang="ts">
+	import { untrack } from 'svelte';
+	import { goto } from '$app/navigation';
+	import { resolve } from '$app/paths';
+	import { X } from '@lucide/svelte';
+
 	import * as Dialog from '$lib/components/ui/dialog/index.js';
 	import * as Field from '$lib/components/ui/field/index.js';
-	import { Button } from '$lib/components/ui/button';
 	import * as NativeSelect from '$lib/components/ui/native-select/index.js';
-	import {
-		displaySize,
-		FileDropZone,
-		type FileDropZoneProps
-	} from '$lib/components/ui/file-drop-zone';
-	import { X } from '@lucide/svelte';
-	import { toast } from 'svelte-sonner';
-
+	import { Button } from '$lib/components/ui/button';
 	import { Switch } from '$lib/components/ui/switch/index';
+	import { displaySize, type FileRejectedReason } from '$lib/components/ui/file-drop-zone';
+	import BookDropZone from './book-drop-zone.svelte';
 
-	import { tick } from 'svelte';
-	import { Spinner } from '$lib/components/ui/spinner/index';
 	import { getLibraryState } from '$lib/state/library.svelte';
-	import { uploadBooks } from '$lib/api';
-	import type { Book, PaginatedResponse } from '$lib/schema';
-	import { goto } from '$app/navigation';
+	import { getUploadQueueState } from '$lib/state/upload-queue.svelte';
 
 	let { open = $bindable() }: { open?: boolean } = $props();
 
-	let libraryState = getLibraryState();
-
-	$effect(() => {
-		uploadBooks.fields.library_id.set(libraryState.activeLibrary!.id);
-	});
-
-	let files = $derived(uploadBooks.fields.files.value() ?? []);
+	const libraryState = getLibraryState();
+	const queue = getUploadQueueState();
 
+	let libraryId = $state<number>(untrack(() => libraryState.activeLibrary!.id));
+	let files = $state<File[]>([]);
 	let autoUploadOnDrop = $state(true);
 	let navigateOnUpload = $state(true);
-	let formEl = $state<HTMLFormElement>();
 
-	const onUpload: FileDropZoneProps['onUpload'] = async (uploadedFiles) => {
-		// Rename files to use webkitRelativePath so directory structure is preserved through form submission
-		const renamedFiles = uploadedFiles.map(f =>
-			new File([f], f.webkitRelativePath || f.name, { type: f.type })
+	const totalSize = $derived(files.reduce((sum, file) => sum + file.size, 0));
+
+	/**
+	 * Collected rather than raised one at a time. A folder of a few hundred books
+	 * carries covers and notes alongside them, and a toast per rejected file
+	 * buries the screen.
+	 */
+	let rejected = $state<{ name: string; reason: FileRejectedReason }[]>([]);
+	let showRejected = $state(false);
+
+	// Start each visit clean — a list left over from last time reads as though it
+	// applies to what was just chosen.
+	$effect(() => {
+		if (open) {
+			untrack(() => {
+				files = [];
+				rejected = [];
+				showRejected = false;
+				libraryId = libraryState.activeLibrary!.id;
+			});
+		}
+	});
+
+	const onUpload = async (uploadedFiles: File[]) => {
+		// Rename to the path relative to the chosen folder, which is what decides
+		// how the API groups files into books. Dropped folders already arrive named
+		// this way, since the drop zone builds the path while walking them.
+		const named = uploadedFiles.map(
+			(file) => new File([file], file.webkitRelativePath || file.name, { type: file.type })
 		);
-		uploadBooks.fields.files.set([...Array.from(files), ...renamedFiles]);
-		if (autoUploadOnDrop && files.length > 0) {
-			await tick();
-			formEl?.requestSubmit();
+
+		// Same relative path twice is the same file — dropping a folder a second
+		// time should not queue everything again.
+		const seen = new Set(files.map((file) => file.name));
+		files = [...files, ...named.filter((file) => !seen.has(file.name))];
+
+		if (autoUploadOnDrop) await startUpload();
+	};
+
+	const onFileRejected = ({ reason, file }: { reason: FileRejectedReason; file: File }) => {
+		rejected = [...rejected, { name: file.webkitRelativePath || file.name, reason }];
+	};
+
+	/**
+	 * Files picked from a folder carry their whole relative path as the name, so
+	 * truncating the end would cut off the filename — the only part worth reading.
+	 * Split it and let the folder sit on its own, quieter line.
+	 */
+	function splitPath(path: string) {
+		const cut = path.lastIndexOf('/');
+		return cut === -1
+			? { dir: '', name: path }
+			: { dir: path.slice(0, cut), name: path.slice(cut + 1) };
 	}
-	};
 
-	const onFileRejected: FileDropZoneProps['onFileRejected'] = async ({ reason, file }) => {
-		toast.error(`${file.name} failed to upload!`, { description: reason });
-	};
+	/**
+	 * Hands the books to the queue and closes.
+	 *
+	 * Nothing is awaited here: the queue lives in the root layout and reports
+	 * through the tray, so the import carries on while the library stays usable.
+	 */
+	function startUpload() {
+		if (files.length === 0) return;
 
-	function navigateToBooks(books: PaginatedResponse<Book>) {
+		const queued = files;
+		const target = libraryId;
+
+		files = [];
+		rejected = [];
 		open = false;
-		let libraryId = books.items[0].library_id;
-		libraryState.setActive(libraryId);
-		if (books.items.length === 1) {
-			goto(`/book/${books.items[0].id}`);
-		} else {
-			goto(`/library/${libraryId}/view?orderBy=created_at&sortOrder=desc`);
+
+		queue.enqueue(target, queued, ({ created, firstBook }) => {
+			if (created === 0) return;
+
+			const library = libraryState.libraries.find((lib) => lib.id === target);
+			if (library) library.total = (library.total ?? 0) + created;
+
+			// Only for a single book. Jumping somewhere after a bulk import would
+			// land minutes after the reader moved on.
+			if (navigateOnUpload && created === 1 && firstBook) {
+				libraryState.setActive(target);
+				void goto(resolve('/(root)/(library)/book/[bookId]', { bookId: String(firstBook.id) }));
 			}
+		});
 	}
 </script>
 
 <Dialog.Root bind:open>
-	<Dialog.Content>
-		{#if uploadBooks.pending}
-			<div class="flex flex-col items-center gap-4">
-				<span class="text-lg font-semibold"
-					>Uploading {uploadBooks.fields.files.value().length} files...</span
-				>
-				<Spinner class="scale-150" />
-			</div>
-		{:else}
+	<!--
+		Wider than the default lg: a folder's worth of rows needs the room.
+
+		overflow-hidden and the min-w-0 on the body below are what keep a long
+		filename inside the dialog. Dialog.Content is a grid, and grid and flex
+		items default to min-width:auto — they refuse to shrink below their
+		content's intrinsic width, so one long name widened the body and pushed it
+		straight through the dialog's edge regardless of any truncate further down.
+	-->
+	<Dialog.Content class="overflow-hidden sm:max-w-2xl">
 		<Dialog.Header>
-				<Dialog.Title>Upload Books</Dialog.Title>
+			<Dialog.Title>Add books</Dialog.Title>
+			<Dialog.Description>
+				Files or a folder. A folder becomes one book per directory.
+			</Dialog.Description>
 		</Dialog.Header>
 
-			<form
-				{...uploadBooks.enhance(async ({ submit, form }) => {
-					try {
-						await submit();
-
-						// Check if there are any validation issues
-						const issues = uploadBooks.fields.allIssues();
-						if (issues && issues.length > 0) {
-							return;
-						}
-
-						// Update library book count
-						const count = uploadBooks.result.total
-						libraryState.libraries.find(lib => uploadBooks.fields.library_id.value() == lib.id.toString())!.total += count
-
-						// Reset the files field
-						uploadBooks.fields.files.set([]);
-						toast.success('Books successfully uploaded!');
-
-						if (navigateOnUpload) {
-							navigateToBooks(uploadBooks.result);
-						}
-					} catch (error) {
-						console.error('Failed to upload book: ', error);
-						toast.error('Failed to upload books');
-					}
-				})}
-				bind:this={formEl}
-				enctype="multipart/form-data"
-				class="flex w-full flex-col gap-2 p-4"
-			>
-				<!-- Library select field  -->
-				<Field.Label for="library_id">Select Library</Field.Label>
-				<NativeSelect.Root {...uploadBooks.fields.library_id.as('select')} class="w-36">
-					{#each libraryState.libraries as library}
-						<NativeSelect.Option value={library.id}>
-							{library.name}
-						</NativeSelect.Option>
+		<div class="flex w-full min-w-0 flex-col gap-3">
+			<div class="flex flex-col gap-1.5">
+				<Field.Label for="library_id">Library</Field.Label>
+				<NativeSelect.Root id="library_id" bind:value={libraryId} class="w-48">
+					{#each libraryState.libraries as library (library.id)}
+						<NativeSelect.Option value={library.id}>{library.name}</NativeSelect.Option>
 					{/each}
 				</NativeSelect.Root>
+			</div>
 
-				<FileDropZone
+			<BookDropZone
 				{onUpload}
 				{onFileRejected}
-					directory={true}
 				accept=".pdf,.epub,.mobi,application/pdf,application/epub+zip,application/x-mobipocket-ebook"
-					sublabel="Only PDF, EPUB, and MOBI files supported"
 			/>
-				<input class="hidden" {...uploadBooks.fields.files.as('file multiple')} />
-				<div class="flex max-h-[300px] flex-col gap-2 overflow-y-auto">
-					{#each files as file, idx}
-						<div class="flex place-items-center justify-between gap-2">
-							<div class="flex flex-col">
-								<span>{file.name}</span>
-								<span class="text-xs text-muted-foreground">{displaySize(file.size)}</span>
+
+			{#if files.length > 0}
+				<div class="flex items-baseline justify-between border-b pb-1 text-sm">
+					<span><strong class="tabular-nums">{files.length}</strong> ready to upload</span>
+					<span class="font-mono text-xs text-muted-foreground tabular-nums">
+						{displaySize(totalSize)}
+					</span>
+				</div>
+			{/if}
+
+			<div class="flex max-h-[300px] min-w-0 flex-col gap-2 overflow-y-auto">
+				{#each files as file, idx (file.name)}
+					{@const location = splitPath(file.name)}
+					<div class="flex min-w-0 items-center justify-between gap-2">
+						<!-- flex-1 as well as min-w-0: without a constrained width there is
+						     nothing for truncate to act against and the row pushes the dialog wide -->
+						<div class="flex min-w-0 flex-1 flex-col">
+							<span class="truncate text-sm" title={file.name}>{location.name}</span>
+							<span class="flex min-w-0 items-baseline gap-2 text-xs text-muted-foreground">
+								{#if location.dir}
+									<span class="truncate font-mono" title={location.dir}>{location.dir}</span>
+								{/if}
+								<span class="shrink-0 tabular-nums">{displaySize(file.size)}</span>
+							</span>
 						</div>
 						<Button
 							variant="outline"
 							size="icon"
-								onclick={() => {
-									uploadBooks.fields.files.set([
-										...Array.from(files).slice(0, idx),
-										...Array.from(files).slice(idx + 1)
-									]);
-								}}
+							class="shrink-0"
+							onclick={() => (files = files.filter((_, i) => i !== idx))}
 						>
 							<X />
+							<span class="sr-only">Remove {file.name}</span>
 						</Button>
 					</div>
 				{/each}
 			</div>
 
-				<div class="flex flex-col gap-2">
-					<div class="flex items-center gap-2">
-						<Switch bind:checked={autoUploadOnDrop} />
-						<Field.Label for="auto-upload-on-drop">Auto upload on file drop</Field.Label>
-						<Button type="submit" class="ml-auto w-fit">Upload</Button>
+			{#if rejected.length > 0}
+				<div class="rounded-md border border-star/50 bg-star/10 p-2 text-sm">
+					<div class="flex items-center justify-between gap-2">
+						<span>
+							{rejected.length}
+							{rejected.length === 1 ? 'file was' : 'files were'} skipped
+						</span>
+						<Button variant="ghost" size="sm" onclick={() => (showRejected = !showRejected)}>
+							{showRejected ? 'Hide' : 'Show'}
+						</Button>
 					</div>
-					<div class="flex items-center gap-2">
-						<Switch bind:checked={navigateOnUpload} />
-						<Field.Label for="navigate-to-book">Navigate to book on upload</Field.Label>
-					</div>
-				</div>
-			</form>
+					{#if showRejected}
+						<ul class="mt-2 flex max-h-32 min-w-0 flex-col gap-1 overflow-y-auto">
+							{#each rejected as entry (entry.name)}
+								<li class="flex min-w-0 justify-between gap-2 text-xs text-muted-foreground">
+									<span class="min-w-0 flex-1 truncate" title={entry.name}>
+										{splitPath(entry.name).name}
+									</span>
+									<span class="shrink-0">{entry.reason}</span>
+								</li>
+							{/each}
+						</ul>
 					{/if}
+				</div>
+			{/if}
+
+			<div class="flex flex-col gap-2 border-t pt-3">
+				<div class="flex items-center gap-2">
+					<Switch id="auto-upload-on-drop" bind:checked={autoUploadOnDrop} />
+					<Field.Label for="auto-upload-on-drop" class="font-normal">
+						Start as soon as books are added
+					</Field.Label>
+					<Button class="ml-auto w-fit" disabled={files.length === 0} onclick={startUpload}>
+						Upload
+					</Button>
+				</div>
+				<div class="flex items-center gap-2">
+					<Switch id="navigate-to-book" bind:checked={navigateOnUpload} />
+					<Field.Label for="navigate-to-book" class="font-normal">
+						Open the book when a single one is added
+					</Field.Label>
+				</div>
+			</div>
+		</div>
 	</Dialog.Content>
 </Dialog.Root>

frontend/src/lib/components/forms/edit-book/edit-book.svelte

@@ -1,39 +1,68 @@
 <script lang="ts">
 	import * as Dialog from '$lib/components/ui/dialog/index.js';
-	import * as Tabs from '$lib/components/ui/tabs/index.js';
+	import { Button } from '$lib/components/ui/button/index.js';
 	import { type Book } from '$lib/schema';
 	import EditCover from './edit-cover.svelte';
 	import EditFiles from './edit-files.svelte';
 	import EditMetadata from './edit-metadata.svelte';
 
 	let { book, open = $bindable() }: { book?: Book; open: boolean } = $props();
+
+	/**
+	 * The metadata form lives in a child, but Save belongs in the dialog footer —
+	 * a submit button reaches it by id rather than the footer having to duplicate
+	 * the submit logic.
+	 */
+	const METADATA_FORM_ID = 'edit-book-metadata';
 </script>
 
+<!--
+	This dialog is mounted once, in (root)/(library)/+layout.svelte, and `book`
+	changes underneath it as different books are edited. bookToEdit is never
+	cleared, so {#if book} stays true and the forms never unmount — they seed
+	local state from `book` on mount, so without this key you would open Edit on a
+	second book and see the first book's authors, tags and cover, then save them
+	onto the wrong record. Keying on the id remounts the forms per book.
+-->
 <Dialog.Root bind:open>
 	{#if book}
-		<Dialog.Content class="sm:max-w-xl">
-			<Tabs.Root value="metadata" class="h-[500px] max-w-xl py-4 md:h-[700px]">
-				<Tabs.List class="grid w-full grid-cols-3">
-					<Tabs.Trigger value="metadata">Metadata</Tabs.Trigger>
-					<Tabs.Trigger value="cover">Cover</Tabs.Trigger>
-					<Tabs.Trigger value="files">Files</Tabs.Trigger>
-				</Tabs.List>
+		{#key book.id}
+			<Dialog.Content
+				class="flex h-[85vh] max-w-[min(72rem,95vw)] flex-col gap-0 overflow-hidden p-0 sm:max-w-[min(72rem,95vw)]"
+			>
+				<Dialog.Header class="shrink-0 space-y-0 border-b px-5 py-3 text-left">
+					<Dialog.Title class="truncate font-serif text-base font-normal">{book.title}</Dialog.Title
+					>
+					<Dialog.Description class="truncate text-xs">
+						{book.authors.map((author) => author.name).join(', ') || 'Unknown author'}
+					</Dialog.Description>
+				</Dialog.Header>
 
-				<!-- Metadata form -->
-				<Tabs.Content value="metadata" class="h-full overflow-y-auto pb-1">
-					<EditMetadata {book} {open} />
-				</Tabs.Content>
-
-				<!-- Cover form -->
-				<Tabs.Content value="cover">
-					<EditCover {book} {open} />
-				</Tabs.Content>
-
-				<!-- Add files form -->
-				<Tabs.Content value="files">
+				<!-- Rail and form scroll independently so the footer never moves -->
+				<div class="grid min-h-0 flex-1 grid-cols-1 md:grid-cols-[16rem_1fr]">
+					<aside
+						class="flex min-w-0 flex-col gap-5 overflow-y-auto border-b bg-sidebar p-4 md:border-r md:border-b-0"
+					>
+						<EditCover {book} />
 						<EditFiles {book} />
-				</Tabs.Content>
-			</Tabs.Root>
+					</aside>
+
+					<div class="min-h-0 overflow-y-auto p-5">
+						<EditMetadata {book} bind:open formId={METADATA_FORM_ID} />
+					</div>
+				</div>
+
+				<Dialog.Footer class="shrink-0 items-center gap-2 border-t px-5 py-3 sm:justify-between">
+					<!-- Cover and file changes hit the server as they happen, while the
+					     fields wait for Save. Saying so is the cheapest way to stop
+					     Cancel reading as "undo everything". -->
+					<p class="text-xs text-muted-foreground">Cover and file changes apply immediately</p>
+					<div class="flex gap-2">
+						<Button variant="outline" onclick={() => (open = false)}>Cancel</Button>
+						<Button type="submit" form={METADATA_FORM_ID}>Save changes</Button>
+					</div>
+				</Dialog.Footer>
 			</Dialog.Content>
+		{/key}
 	{/if}
 </Dialog.Root>

frontend/src/lib/components/forms/edit-book/edit-cover.svelte

@@ -1,33 +1,23 @@
 <script lang="ts">
-	import {
-		FileDropZone,
-		type FileDropZoneProps,
-		displaySize
-	} from '$lib/components/ui/file-drop-zone/index.js';
-	import { Button } from '$lib/components/ui/button/index.js';
-	import * as Field from '$lib/components/ui/field/index.js';
-	import { Switch } from '$lib/components/ui/switch/index';
-	import BookImage from '$lib/components/view/book-image.svelte';
+	import { tick, untrack } from 'svelte';
 	import { toast } from 'svelte-sonner';
 
 	import { updateBookCover } from '$lib/api';
 	import type { Book } from '$lib/schema';
-	import { tick, untrack } from 'svelte';
-	import { X } from '@lucide/svelte';
+	import BookImage from '$lib/components/view/book-image.svelte';
+	import { FileDropZone, type FileDropZoneProps } from '$lib/components/ui/file-drop-zone/index.js';
 
-	let { book, open = $bindable() }: { book: Book; open: boolean } = $props();
+	let { book }: { book: Book } = $props();
 
 	let formEl = $state<HTMLFormElement>();
-	let coverImagePreview = $state(`/api/${book.cover_image}`);
-	let autoUploadOnDrop = $state(true);
+	// Seeded once per mount — see the key in edit-book.svelte
+	let coverImagePreview = $state<string>(untrack(() => `/api/${book.cover_image}`));
 
 	const onUpload: FileDropZoneProps['onUpload'] = async (uploadedFiles) => {
 		updateBookCover.fields.file.set(uploadedFiles[0]);
 		updateCoverPreview();
-		if (autoUploadOnDrop && updateBookCover.fields.file.value()) {
 		await tick();
 		formEl?.requestSubmit();
-		}
 	};
 
 	function updateCoverPreview() {
@@ -35,14 +25,14 @@
 		if (file && file.type.startsWith('image/')) {
 			const reader = new FileReader();
 			reader.onloadend = () => {
-				coverImagePreview = reader.result;
+				coverImagePreview = reader.result as string;
 			};
 			reader.readAsDataURL(file);
 		}
 	}
 
 	const onFileRejected: FileDropZoneProps['onFileRejected'] = async ({ reason, file }) => {
-		toast.error(`${file.name} failed to upload!`, { description: reason });
+		toast.error(`${file.name} was not used`, { description: reason });
 	};
 
 	$effect(() => {
@@ -53,65 +43,40 @@
 	});
 </script>
 
-<form
+<section class="flex min-w-0 flex-col gap-2">
+	<h3 class="font-mono text-[10px] tracking-widest text-muted-foreground uppercase">Cover</h3>
+
+	<BookImage src={coverImagePreview} class="w-full rounded-md border" />
+
+	<form
 		bind:this={formEl}
 		{...updateBookCover.enhance(async ({ submit, form }) => {
 			try {
 				await submit();
 				form.reset();
-			open = false;
-			toast.success('Updated book cover!');
+				// Deliberately does not close the dialog. The cover is one panel of a
+				// larger form now, and closing here would throw away metadata edits
+				// the reader has not saved yet.
+				toast.success('Cover updated');
 			} catch (error) {
 				console.error('Failed to update book cover: ', error);
-			toast.error('Failed to update cover.');
+				toast.error('Failed to update the cover');
 			}
 		})}
 		enctype="multipart/form-data"
-	class="grid grid-cols-[1fr_2fr] gap-4 p-6"
->
+		class="flex flex-col gap-2"
+	>
 		<input class="hidden" {...updateBookCover.fields.book_id.as('text')} />
+		<input class="hidden" {...updateBookCover.fields.file.as('file')} />
 
-	<div class="flex flex-col gap-2">
-		<BookImage src={coverImagePreview} class="w-64 rounded" />
-	</div>
-
-	<div class="flex flex-col gap-2">
 		<FileDropZone
 			{onUpload}
 			{onFileRejected}
 			accept=".jpeg,.jpg,.png,.webp,image/*"
-			label="Only JPEG, PNG, and WEBP images supported"
+			label="Replace cover"
+			sublabel="JPEG, PNG or WEBP"
 			maxFiles={1}
 			fileCount={updateBookCover.fields.file.value() ? 1 : 0}
 		/>
-		<input class="hidden" {...updateBookCover.fields.file.as('file')} />
-		<div class="flex flex-col gap-2">
-			{#if updateBookCover.fields.file.value()}
-				<div class="flex place-items-center justify-between gap-2">
-					<div class="flex flex-col">
-						<span>{updateBookCover.fields.file.value().name}</span>
-						<span class="text-xs text-muted-foreground"
-							>{displaySize(updateBookCover.fields.file.value().size)}</span
-						>
-					</div>
-					<Button
-						variant="outline"
-						size="icon"
-						onclick={() => {
-							updateBookCover.fields.file.set(undefined);
-							coverImagePreview = `/api/${book.cover_image}`;
-						}}
-					>
-						<X />
-					</Button>
-				</div>
-			{/if}
-		</div>
-
-		<div class="flex flex-row items-center space-x-2">
-			<Switch bind:checked={autoUploadOnDrop} />
-			<Field.Label for="auto-upload-on-drop">Auto upload on file drop</Field.Label>
-			<Button type="submit" class="ml-auto w-fit" disabled={!updateBookCover.fields.file.value()}>Upload</Button>
-		</div>
-	</div>
-</form>
+	</form>
+</section>

frontend/src/lib/components/forms/edit-book/edit-files.svelte

@@ -1,98 +1,213 @@
 <script lang="ts">
-	import { uploadBookFiles } from '$lib/api';
-	import {
-		displaySize,
-		FileDropZone,
-		type FileDropZoneProps
-	} from '$lib/components/ui/file-drop-zone';
-	import { Button } from '$lib/components/ui/button';
-	import * as Field from '$lib/components/ui/field/index.js';
-
-	import { Switch } from '$lib/components/ui/switch/index';
-	import { X } from '@lucide/svelte';
-
-	import type { Book } from '$lib/schema';
-	import { tick } from 'svelte';
+	import { tick, untrack } from 'svelte';
 	import { toast } from 'svelte-sonner';
+	import { Download, Trash2 } from '@lucide/svelte';
+
+	import { uploadBookFiles } from '$lib/api';
+	import type { Book, BookFile } from '$lib/schema';
+	import { formatFileSize, getFileType } from '$lib/utils';
+	import { getBookOperationsState } from '$lib/state/bookOperations.svelte';
+
+	import * as AlertDialog from '$lib/components/ui/alert-dialog/index.js';
+	import * as Field from '$lib/components/ui/field/index.js';
+	import { Button, buttonVariants } from '$lib/components/ui/button/index.js';
+	import { Checkbox } from '$lib/components/ui/checkbox/index.js';
+	import { FileDropZone, type FileDropZoneProps } from '$lib/components/ui/file-drop-zone';
+	import { Spinner } from '$lib/components/ui/spinner/index';
 
 	let { book }: { book: Book } = $props();
 
-	let files = $derived(uploadBookFiles.fields.files.value() ?? []);
+	const bookOps = getBookOperationsState();
 
-	let autoUploadOnDrop = $state(true);
+	/**
+	 * The book's files, owned locally.
+	 *
+	 * `book` is a snapshot handed down from bookOperations rather than a live
+	 * query, so invalidating after a delete does not reach it. Keeping the list
+	 * here lets the rail reflect an add or a remove straight away. Seeded once per
+	 * mount — edit-book.svelte keys the dialog on book.id.
+	 */
+	let files = $state<BookFile[]>(untrack(() => [...book.files]));
+
+	// The field's value is a sparse-ish list until the form settles, so narrow it
+	// before rendering rather than asserting at each use.
+	let pending = $derived(
+		(uploadBookFiles.fields.files.value() ?? []).filter((file): file is File => Boolean(file))
+	);
+
+	let fileToDelete = $state<BookFile>();
+	let deleteFromDisk = $state(true);
+	let confirmOpen = $state(false);
 	let formEl = $state<HTMLFormElement>();
 
-	const onUpload: FileDropZoneProps['onUpload'] = async (uploadedFiles) => {
-		uploadBookFiles.fields.files.set([...Array.from(files), ...uploadedFiles]);
-		if (autoUploadOnDrop && files.length > 0) {
+	const onUpload: FileDropZoneProps['onUpload'] = async (uploaded) => {
+		uploadBookFiles.fields.files.set([...Array.from(pending), ...uploaded]);
 		await tick();
 		formEl?.requestSubmit();
-		}
 	};
 
 	const onFileRejected: FileDropZoneProps['onFileRejected'] = async ({ reason, file }) => {
-		toast.error(`${file.name} failed to upload!`, { description: reason });
+		toast.error(`${file.name} was not added`, { description: reason });
 	};
+
+	/**
+	 * The API's own words, when it has any.
+	 *
+	 * Adding a file the library already holds under another book is refused with a
+	 * 409 naming it — far more use than "failed to add files". SvelteKit hands an
+	 * `error()` back as an HttpError on the client, so the message sits on `body`.
+	 */
+	function apiMessage(error: unknown): string | undefined {
+		if (typeof error !== 'object' || error === null) return undefined;
+
+		const body = (error as { body?: { message?: string } }).body;
+		if (typeof body?.message === 'string') return body.message;
+
+		return error instanceof Error ? error.message : undefined;
+	}
+
+	function confirmDelete(file: BookFile) {
+		fileToDelete = file;
+		confirmOpen = true;
+	}
+
+	async function removeFile() {
+		if (!fileToDelete) return;
+
+		const target = fileToDelete;
+		confirmOpen = false;
+
+		// Drop it from the list first: the request invalidates the books query, but
+		// this dialog holds its own copy of the book and would not see that.
+		files = files.filter((file) => file.id !== target.id);
+
+		await bookOps.deleteBookFiles(book.id, [target.id], deleteFromDisk);
+		fileToDelete = undefined;
+	}
 </script>
 
-<form
-	{...uploadBookFiles.enhance(async ({ submit, form }) => {
+<section class="flex min-w-0 flex-col gap-2">
+	<h3 class="font-mono text-[10px] tracking-widest text-muted-foreground uppercase">Files</h3>
+
+	{#if files.length === 0}
+		<p class="text-xs text-muted-foreground">
+			No files yet. Add one below so this book can be read or downloaded.
+		</p>
+	{/if}
+
+	<ul class="flex flex-col gap-1.5">
+		{#each files as file (file.id)}
+			<li class="flex items-center gap-2 rounded-md border bg-background p-2">
+				<span
+					class="rounded-sm bg-accent px-1.5 py-0.5 font-mono text-[9px] font-semibold text-accent-foreground"
+				>
+					{getFileType(file.filename)}
+				</span>
+
+				<span class="min-w-0 flex-1">
+					<span class="block truncate text-xs" title={file.filename}>{file.filename}</span>
+					<span class="block font-mono text-[10px] text-muted-foreground tabular-nums">
+						{formatFileSize(file.size)}
+					</span>
+				</span>
+
+				<Button
+					variant="ghost"
+					size="icon"
+					class="size-7 shrink-0"
+					title="Download {file.filename}"
+					onclick={() => bookOps.downloadBookFile(book.id, file.id, file.filename)}
+				>
+					<Download class="size-3.5" />
+					<span class="sr-only">Download {file.filename}</span>
+				</Button>
+
+				<Button
+					variant="ghost"
+					size="icon"
+					class="size-7 shrink-0 text-destructive hover:text-destructive"
+					title="Remove {file.filename}"
+					onclick={() => confirmDelete(file)}
+				>
+					<Trash2 class="size-3.5" />
+					<span class="sr-only">Remove {file.filename}</span>
+				</Button>
+			</li>
+		{/each}
+
+		{#each pending as file (file.name)}
+			<li
+				class="flex items-center gap-2 rounded-md border border-dashed bg-background p-2 text-muted-foreground"
+			>
+				<Spinner class="size-3.5 shrink-0" />
+				<span class="min-w-0 flex-1">
+					<span class="block truncate text-xs">{file.name}</span>
+					<span class="block font-mono text-[10px] tabular-nums">Uploading…</span>
+				</span>
+			</li>
+		{/each}
+	</ul>
+
+	<form
+		bind:this={formEl}
+		{...uploadBookFiles.enhance(async ({ submit }) => {
 			try {
 				await submit();
 
-			// Check if there are any validation issues
 				const issues = uploadBookFiles.fields.allIssues();
-			if (issues && issues.length > 0) {
-				return;
-			}
+				if (issues && issues.length > 0) return;
 
-			// Reset the files field
+				// The endpoint answers with the updated book, so the new files come
+				// back with their ids rather than having to be guessed at.
+				files = uploadBookFiles.result?.files ?? files;
 				uploadBookFiles.fields.files.set([]);
-			toast.success('Files successfully added!');
+				toast.success('Files added');
 			} catch (error) {
-			console.error('Failed to upload files: ', error);
-			toast.error('Failed to upload files');
+				console.error('Failed to add files: ', error);
+				toast.error(apiMessage(error) ?? 'Failed to add files');
+				uploadBookFiles.fields.files.set([]);
 			}
 		})}
-	bind:this={formEl}
 		enctype="multipart/form-data"
-	class="flex w-full flex-col gap-2 p-4"
->
+		class="flex flex-col gap-2"
+	>
 		<input {...uploadBookFiles.fields.book_id.as('hidden', book.id)} />
+		<input class="hidden" {...uploadBookFiles.fields.files.as('file multiple')} />
 
 		<FileDropZone
 			{onUpload}
 			{onFileRejected}
 			accept=".pdf,.epub,.mobi,application/pdf,application/epub+zip,application/x-mobipocket-ebook"
-		sublabel="Only PDF, EPUB, and MOBI files supported"
+			label="Add a file"
+			sublabel="EPUB, PDF or MOBI"
 		/>
-	<input class="hidden" {...uploadBookFiles.fields.files.as('file multiple')} />
-	<div class="flex flex-col gap-2">
-		{#each files as file, idx}
-			<div class="flex place-items-center justify-between gap-2">
-				<div class="flex flex-col">
-					<span>{file.name}</span>
-					<span class="text-xs text-muted-foreground">{displaySize(file.size)}</span>
-				</div>
-				<Button
-					variant="outline"
-					size="icon"
-					onclick={() => {
-						uploadBookFiles.fields.files.set([
-							...Array.from(files).slice(0, idx),
-							...Array.from(files).slice(idx + 1)
-						]);
-					}}
-				>
-					<X />
-				</Button>
-			</div>
-		{/each}
+	</form>
+</section>
+
+<AlertDialog.Root bind:open={confirmOpen}>
+	<AlertDialog.Content>
+		<AlertDialog.Header>
+			<AlertDialog.Title>Remove {fileToDelete?.filename}?</AlertDialog.Title>
+			<AlertDialog.Description>
+				This cannot be undone. The other files on this book are not affected.
+			</AlertDialog.Description>
+		</AlertDialog.Header>
+
+		<!-- The API takes these as separate outcomes: drop the record, or drop the
+		     record and the file on disk. Leaving it implicit would mean deleting
+		     someone's only copy without saying so. -->
+		<div class="flex items-center gap-2">
+			<Checkbox id="delete-from-disk" bind:checked={deleteFromDisk} />
+			<Field.Label for="delete-from-disk" class="font-normal">
+				Also delete the file from the filesystem
+			</Field.Label>
 		</div>
 
-	<div class="flex flex-row items-center space-x-2">
-		<Switch bind:checked={autoUploadOnDrop} />
-		<Field.Label for="auto-upload-on-drop">Auto upload on file drop</Field.Label>
-		<Button type="submit" class="ml-auto w-fit">Upload</Button>
-	</div>
-</form>
+		<AlertDialog.Footer>
+			<AlertDialog.Cancel>Cancel</AlertDialog.Cancel>
+			<AlertDialog.Action class={buttonVariants({ variant: 'destructive' })} onclick={removeFile}>
+				Remove
+			</AlertDialog.Action>
+		</AlertDialog.Footer>
+	</AlertDialog.Content>
+</AlertDialog.Root>

frontend/src/lib/components/forms/edit-book/edit-metadata.svelte

@@ -1,26 +1,32 @@
 <script lang="ts">
-	import * as Card from '$lib/components/ui/card/index.js';
-	import * as Field from '$lib/components/ui/field/index.js';
-	import { TagsInput, type TagsInputProps } from '$lib/components/ui/tags-input/index.js';
-	import { Input } from '$lib/components/ui/input/index.js';
-	import { Textarea } from '$lib/components/ui/textarea/index.js';
-	import { Button } from '$lib/components/ui/button/index.js';
+	import { untrack } from 'svelte';
 	import { toast } from 'svelte-sonner';
+	import { Minus, Plus } from '@lucide/svelte';
 
 	import { updateBookMetadata } from '$lib/api';
 	import type { Book } from '$lib/schema';
-	import { untrack } from 'svelte';
-	import { Minus, Plus } from '@lucide/svelte';
 
-	let { book, open = $bindable() }: { book: Book; open: boolean } = $props();
+	import * as Field from '$lib/components/ui/field/index.js';
+	import { Button } from '$lib/components/ui/button/index.js';
+	import { Input } from '$lib/components/ui/input/index.js';
+	import { TagsInput, type TagsInputProps } from '$lib/components/ui/tags-input/index.js';
+	import { Textarea } from '$lib/components/ui/textarea/index.js';
 
-	let authors = $state(book.authors.map((author) => author.name) || []);
-	let tags = $state(book.tags.map((tag) => tag.name) || []);
-	let identifierKeys = $state(Object.keys(book.identifiers));
-	let identifierValues = $state(Object.values(book.identifiers));
+	let {
+		book,
+		open = $bindable(),
+		/** Lets the dialog footer own the submit button via the `form` attribute. */
+		formId
+	}: { book: Book; open: boolean; formId: string } = $props();
+
+	// Seeded once per mount; edit-book.svelte keys this form on book.id so a
+	// different book gets a fresh form rather than the previous book's values.
+	let authors = $state(untrack(() => book.authors.map((author) => author.name) || []));
+	let tags = $state(untrack(() => book.tags.map((tag) => tag.name) || []));
+	let identifierKeys = $state(untrack(() => Object.keys(book.identifiers)));
+	let identifierValues = $state(untrack(() => Object.values(book.identifiers)));
 
 	function handleAddIdentifier() {
-		// Add empty strings to both arrays
 		identifierKeys = [...identifierKeys, ''];
 		identifierValues = [...identifierValues, ''];
 	}
@@ -86,9 +92,17 @@
 	});
 </script>
 
-<Card.Root class="w-full ">
-	<form
-		{...updateBookMetadata.enhance(async ({ submit, form }) => {
+{#snippet groupHeading(label: string)}
+	<h3
+		class="col-span-full mt-2 border-b pb-1 font-mono text-[10px] tracking-widest text-primary uppercase first:mt-0"
+	>
+		{label}
+	</h3>
+{/snippet}
+
+<form
+	id={formId}
+	{...updateBookMetadata.enhance(async ({ submit }) => {
 		try {
 			await submit();
 
@@ -99,28 +113,19 @@
 			}
 
 			open = false;
-				book = book;
 			toast.success('Updated book metadata!');
 		} catch (error) {
 			console.error('Error occurred updating book metadata: ', error);
 			toast.error('Failed to update book metadata.');
 		}
 	})}
-	>
-		<Card.Content>
-			<Field.Set>
-				<Field.Group class="flex flex-col">
-					<!-- Book ID field -->
-					<Field.Field class="hidden">
-						<Field.Label for="book_id">Book ID</Field.Label>
-						<Input {...updateBookMetadata.fields.book_id.as('text')} />
-						{#each updateBookMetadata.fields.book_id.issues() ?? [] as issue}
-							<Field.Error>{issue.message}</Field.Error>
-						{/each}
-					</Field.Field>
+	class="grid grid-cols-1 items-start gap-x-4 gap-y-3 sm:grid-cols-2"
+>
+	<input class="hidden" {...updateBookMetadata.fields.book_id.as('text')} />
 
-					<!-- Title field -->
-					<Field.Field>
+	{@render groupHeading('Identity')}
+
+	<Field.Field class="col-span-full">
 		<Field.Label for="title">Title</Field.Label>
 		<Input {...updateBookMetadata.fields.title.as('text')} />
 		{#each updateBookMetadata.fields.title.issues() ?? [] as issue}
@@ -128,7 +133,6 @@
 		{/each}
 	</Field.Field>
 
-					<!-- Subtitle field -->
 	<Field.Field>
 		<Field.Label for="subtitle">Subtitle</Field.Label>
 		<Input {...updateBookMetadata.fields.subtitle.as('text')} />
@@ -137,8 +141,14 @@
 		{/each}
 	</Field.Field>
 
-					<div class="grid grid-cols-[3fr_1fr] gap-2">
-						<!-- Series field -->
+	<Field.Field>
+		<Field.Label for="edition">Edition</Field.Label>
+		<Input {...updateBookMetadata.fields.edition.as('number')} />
+		{#each updateBookMetadata.fields.edition.issues() ?? [] as issue}
+			<Field.Error>{issue.message}</Field.Error>
+		{/each}
+	</Field.Field>
+
 	<Field.Field>
 		<Field.Label for="series">Series</Field.Label>
 		<Input {...updateBookMetadata.fields.series.as('text')} />
@@ -147,18 +157,27 @@
 		{/each}
 	</Field.Field>
 
-						<!-- Series position field -->
+	<div class="grid grid-cols-2 gap-2">
 		<Field.Field>
-							<Field.Label for="series_position">Series position</Field.Label>
+			<Field.Label for="series_position">No.</Field.Label>
 			<Input {...updateBookMetadata.fields.series_position.as('text')} />
 			{#each updateBookMetadata.fields.series_position.issues() ?? [] as issue}
 				<Field.Error>{issue.message}</Field.Error>
 			{/each}
 		</Field.Field>
+
+		<Field.Field>
+			<Field.Label for="language">Language</Field.Label>
+			<Input {...updateBookMetadata.fields.language.as('text')} />
+			{#each updateBookMetadata.fields.language.issues() ?? [] as issue}
+				<Field.Error>{issue.message}</Field.Error>
+			{/each}
+		</Field.Field>
 	</div>
 
-					<!-- Authors field -->
-					<Field.Field>
+	{@render groupHeading('People and subjects')}
+
+	<Field.Field class="col-span-full">
 		<Field.Label for="authors">Authors</Field.Label>
 		<TagsInput
 			bind:value={authors}
@@ -174,8 +193,7 @@
 		{/each}
 	</Field.Field>
 
-					<!-- Tags field -->
-					<Field.Field>
+	<Field.Field class="col-span-full">
 		<Field.Label for="tags">Tags</Field.Label>
 		<TagsInput
 			bind:value={tags}
@@ -191,39 +209,8 @@
 		{/each}
 	</Field.Field>
 
-					<!-- Description field -->
-					<Field.Field>
-						<Field.Label for="description">Description</Field.Label>
-						<Textarea {...updateBookMetadata.fields.description.as('text')} />
-						{#each updateBookMetadata.fields.description.issues() ?? [] as issue}
-							<Field.Error>{issue.message}</Field.Error>
-						{/each}
-					</Field.Field>
+	{@render groupHeading('Publication')}
 
-					<!-- Identifier fields -->
-					<Field.Field>
-						<Field.Label for="identifiers">Identifiers</Field.Label>
-						<div class="grid grid-cols-[5fr_10fr_0.5fr] gap-2">
-							{#each identifierKeys as _, idx}
-								<Input bind:value={identifierKeys[idx]} placeholder="Identifier..." />
-								<Input bind:value={identifierValues[idx]} placeholder="Value..." />
-
-								<Button variant="outline" size="icon" onclick={() => handleRemoveIdentifier(idx)}>
-									<Minus />
-								</Button>
-							{/each}
-
-							<Button variant="outline" onclick={() => handleAddIdentifier()}>
-								<Plus />
-								Add Identifier
-							</Button>
-
-							<input {...updateBookMetadata.fields.identifiers.as('text')} class="hidden" />
-						</div>
-					</Field.Field>
-
-					<!-- Publisher field -->
-					<div class="grid grid-cols-[2fr_1fr] gap-2">
 	<Field.Field>
 		<Field.Label for="publisher">Publisher</Field.Label>
 		<Input {...updateBookMetadata.fields.publisher.as('text')} />
@@ -232,18 +219,15 @@
 		{/each}
 	</Field.Field>
 
-						<!-- Published date field -->
+	<div class="grid grid-cols-2 gap-2">
 		<Field.Field>
-							<Field.Label for="published_date">Date published</Field.Label>
+			<Field.Label for="published_date">Published</Field.Label>
 			<Input {...updateBookMetadata.fields.published_date.as('date')} />
 			{#each updateBookMetadata.fields.published_date.issues() ?? [] as issue}
 				<Field.Error>{issue.message}</Field.Error>
 			{/each}
 		</Field.Field>
-					</div>
 
-					<div class="grid grid-cols-[2fr_1fr_1fr] gap-2">
-						<!-- Pages field -->
 		<Field.Field>
 			<Field.Label for="pages">Pages</Field.Label>
 			<Input {...updateBookMetadata.fields.pages.as('number')} />
@@ -251,32 +235,44 @@
 				<Field.Error>{issue.message}</Field.Error>
 			{/each}
 		</Field.Field>
-
-						<!-- Language field -->
-						<Field.Field>
-							<Field.Label for="language">Language</Field.Label>
-							<Input {...updateBookMetadata.fields.language.as('text')} />
-							{#each updateBookMetadata.fields.language.issues() ?? [] as issue}
-								<Field.Error>{issue.message}</Field.Error>
-							{/each}
-						</Field.Field>
-
-						<!-- Edition field -->
-						<Field.Field>
-							<Field.Label for="edition">Edition</Field.Label>
-							<Input {...updateBookMetadata.fields.edition.as('number')} />
-							{#each updateBookMetadata.fields.edition.issues() ?? [] as issue}
-								<Field.Error>{issue.message}</Field.Error>
-							{/each}
-						</Field.Field>
 	</div>
-				</Field.Group>
-			</Field.Set>
-		</Card.Content>
 
-		<!-- Submit button -->
-		<Card.Footer class="flex-col gap-2 pt-6">
-			<Button type="submit" class="w-full">Save</Button>
-		</Card.Footer>
-	</form>
-</Card.Root>
+	<Field.Field class="col-span-full">
+		<Field.Label for="identifiers">Identifiers</Field.Label>
+		<div class="flex flex-col gap-2">
+			{#each identifierKeys as _, idx}
+				<div class="grid grid-cols-[1fr_1.6fr_auto] gap-2">
+					<Input bind:value={identifierKeys[idx]} placeholder="ISBN, DOI…" />
+					<Input bind:value={identifierValues[idx]} placeholder="Value" />
+					<Button
+						type="button"
+						variant="outline"
+						size="icon"
+						title="Remove identifier"
+						onclick={() => handleRemoveIdentifier(idx)}
+					>
+						<Minus />
+						<span class="sr-only">Remove identifier</span>
+					</Button>
+				</div>
+			{/each}
+
+			<Button type="button" variant="outline" class="w-fit" onclick={handleAddIdentifier}>
+				<Plus />
+				Add identifier
+			</Button>
+
+			<input {...updateBookMetadata.fields.identifiers.as('text')} class="hidden" />
+		</div>
+	</Field.Field>
+
+	{@render groupHeading('Description')}
+
+	<Field.Field class="col-span-full">
+		<Field.Label for="description" class="sr-only">Description</Field.Label>
+		<Textarea {...updateBookMetadata.fields.description.as('text')} rows={6} />
+		{#each updateBookMetadata.fields.description.issues() ?? [] as issue}
+			<Field.Error>{issue.message}</Field.Error>
+		{/each}
+	</Field.Field>
+</form>

frontend/src/lib/components/forms/merge-books/field-spec.ts

@@ -0,0 +1,149 @@
+import type { Book } from '$lib/schema';
+
+/**
+ * What kind of thing a metadata field holds, which is the only thing that decides
+ * what you can do with it when two records disagree.
+ */
+export type FieldKind = 'text' | 'number' | 'date' | 'list' | 'keyed' | 'longtext';
+
+/** An action offered on a field, beyond replacing it outright. */
+export type FieldAction = 'replace' | 'merge' | 'append';
+
+export interface FieldSpec {
+	/** The key sent in `BookMetadataUpdate`. */
+	key: string;
+	label: string;
+	kind: FieldKind;
+	group: string;
+	/** Extra actions past `replace`, which every field has. */
+	extra: FieldAction[];
+}
+
+/**
+ * The fields a merge can resolve, in the order and grouping the edit form uses.
+ *
+ * Deliberately a spec rather than markup: the merge workbench and, later, the
+ * provider review screen both render from this, so a field cannot exist in one and
+ * not the other. `cover` and `files` are absent because they are not choices —
+ * files always come across and the cover has its own endpoint.
+ */
+export const MERGE_FIELDS: FieldSpec[] = [
+	{ key: 'title', label: 'Title', kind: 'text', group: 'Identity', extra: [] },
+	{ key: 'subtitle', label: 'Subtitle', kind: 'text', group: 'Identity', extra: [] },
+	{ key: 'edition', label: 'Edition', kind: 'number', group: 'Identity', extra: [] },
+	{ key: 'series', label: 'Series', kind: 'text', group: 'Identity', extra: [] },
+	{ key: 'series_position', label: 'No.', kind: 'text', group: 'Identity', extra: [] },
+	{ key: 'language', label: 'Language', kind: 'text', group: 'Identity', extra: [] },
+
+	// Order is meaningful for authors, so the second list is appended rather than
+	// interleaved; tags are a set, so they merge.
+	{
+		key: 'authors',
+		label: 'Authors',
+		kind: 'list',
+		group: 'People and subjects',
+		extra: ['append']
+	},
+	{ key: 'tags', label: 'Tags', kind: 'list', group: 'People and subjects', extra: ['merge'] },
+
+	{ key: 'publisher', label: 'Publisher', kind: 'text', group: 'Publication', extra: [] },
+	{ key: 'published_date', label: 'Published', kind: 'date', group: 'Publication', extra: [] },
+	{ key: 'pages', label: 'Pages', kind: 'number', group: 'Publication', extra: [] },
+	{
+		key: 'identifiers',
+		label: 'Identifiers',
+		kind: 'keyed',
+		group: 'Publication',
+		extra: ['merge']
+	},
+
+	{
+		key: 'description',
+		label: 'Description',
+		kind: 'longtext',
+		group: 'Description',
+		extra: ['append']
+	}
+];
+
+/** A field's value, in the shape `BookMetadataUpdate` expects to receive it. */
+export type FieldValue = string | number | string[] | Record<string, string> | null;
+
+/** Read one field off a book, flattening the relations the API returns as objects. */
+export function readField(book: Book, key: string): FieldValue {
+	switch (key) {
+		case 'authors':
+			return book.authors.map((author) => author.name);
+		case 'tags':
+			return book.tags.map((tag) => tag.name);
+		case 'publisher':
+			return book.publisher?.name ?? null;
+		case 'series':
+			return book.series?.title ?? null;
+		case 'identifiers':
+			return book.identifiers ?? {};
+		default:
+			return (book as unknown as Record<string, FieldValue>)[key] ?? null;
+	}
+}
+
+/** Whether a field holds nothing, and so has no decision attached to it. */
+export function isEmpty(value: FieldValue): boolean {
+	if (value === null || value === undefined || value === '') return true;
+	if (Array.isArray(value)) return value.length === 0;
+	if (typeof value === 'object') return Object.keys(value).length === 0;
+	return false;
+}
+
+/** Whether two field values say the same thing, order included for lists. */
+export function isSame(left: FieldValue, right: FieldValue): boolean {
+	if (isEmpty(left) && isEmpty(right)) return true;
+	return JSON.stringify(left ?? null) === JSON.stringify(right ?? null);
+}
+
+/**
+ * Apply an action to a pair of values and return what the target becomes.
+ *
+ * `merge` on a keyed collection is per name and the target wins a clash, because
+ * `Identifier` is unique on `(name, book_id)` — a book cannot hold both its print
+ * and its ebook ISBN, so the second one has nowhere to go.
+ */
+export function applyAction(
+	action: FieldAction,
+	kind: FieldKind,
+	target: FieldValue,
+	incoming: FieldValue
+): FieldValue {
+	if (action === 'replace') return incoming;
+
+	if (kind === 'list') {
+		const current = Array.isArray(target) ? target : [];
+		const extra = Array.isArray(incoming) ? incoming : [];
+		// Order preserved, duplicates dropped — works for both append and merge.
+		return [...new Set([...current, ...extra])];
+	}
+
+	if (kind === 'keyed') {
+		return { ...(incoming as Record<string, string>), ...(target as Record<string, string>) };
+	}
+
+	if (kind === 'longtext') {
+		const current = typeof target === 'string' ? target.trim() : '';
+		const extra = typeof incoming === 'string' ? incoming.trim() : '';
+		return [current, extra].filter(Boolean).join('\n\n');
+	}
+
+	return incoming;
+}
+
+/** How a value reads in the inert reference column. */
+export function displayValue(value: FieldValue): string {
+	if (isEmpty(value)) return '';
+	if (Array.isArray(value)) return value.join(', ');
+	if (typeof value === 'object') {
+		return Object.entries(value as Record<string, string>)
+			.map(([name, id]) => `${name}: ${id}`)
+			.join('  ·  ');
+	}
+	return String(value);
+}

frontend/src/lib/components/forms/merge-books/merge-books.svelte

@@ -0,0 +1,394 @@
+<script lang="ts">
+	import { untrack } from 'svelte';
+	import { ArrowRight, GitMerge, Plus, Undo2 } from '@lucide/svelte';
+
+	import * as Dialog from '$lib/components/ui/dialog/index.js';
+	import { Button } from '$lib/components/ui/button/index.js';
+	import { Input } from '$lib/components/ui/input/index.js';
+	import { Textarea } from '$lib/components/ui/textarea/index.js';
+	import { Badge } from '$lib/components/ui/badge/index.js';
+	import BookImage from '$lib/components/view/book-image.svelte';
+	import GeneratedCover from '$lib/components/view/generated-cover.svelte';
+	import type { Book } from '$lib/schema';
+
+	import { mergeInto } from './merge';
+
+	import {
+		MERGE_FIELDS,
+		applyAction,
+		displayValue,
+		isEmpty,
+		isSame,
+		readField,
+		type FieldAction,
+		type FieldSpec,
+		type FieldValue
+	} from './field-spec';
+
+	let {
+		books,
+		libraryId,
+		open = $bindable(),
+		onmerged
+	}: {
+		books: Book[];
+		libraryId: number | string;
+		open: boolean;
+		/** Given the record that survived and the ones folded into it and deleted. */
+		onmerged?: (survivor: Book, folded: Book[]) => void;
+	} = $props();
+
+	// Seeded once per mount. The dialog is keyed on the group upstream, so a
+	// different group gets a fresh workbench rather than the previous one's draft.
+	let survivorId = $state(untrack(() => books[0]?.id));
+	let candidateId = $state(untrack(() => books[1]?.id));
+	let draft = $state<Record<string, FieldValue>>({});
+	let busy = $state(false);
+
+	const survivor = $derived(books.find((book) => book.id === survivorId) ?? books[0]);
+	const candidates = $derived(books.filter((book) => book.id !== survivorId));
+	/** The records that will be deleted — the same set, named for what happens to them. */
+	const folded = $derived(candidates);
+	const candidate = $derived(candidates.find((book) => book.id === candidateId) ?? candidates[0]);
+
+	/** The survivor's stored value for a field, or the draft if it has been touched. */
+	function current(field: FieldSpec): FieldValue {
+		return field.key in draft ? draft[field.key] : readField(survivor, field.key);
+	}
+
+	function take(field: FieldSpec, action: FieldAction) {
+		draft[field.key] = applyAction(
+			action,
+			field.kind,
+			current(field),
+			readField(candidate, field.key)
+		);
+	}
+
+	function undo(field: FieldSpec) {
+		delete draft[field.key];
+		draft = { ...draft };
+	}
+
+	/**
+	 * Fill only the fields the survivor has nothing in.
+	 *
+	 * The safe bulk action, and the one worth reaching for: it cannot overwrite a
+	 * value, so it needs no per-field protection to be pressed without reading.
+	 */
+	function fillEmpty() {
+		for (const field of MERGE_FIELDS) {
+			const incoming = readField(candidate, field.key);
+			if (isEmpty(current(field)) && !isEmpty(incoming)) draft[field.key] = incoming;
+		}
+	}
+
+	const changed = $derived(Object.keys(draft));
+
+	const differing = $derived(
+		candidate
+			? MERGE_FIELDS.filter((field) => !isSame(current(field), readField(candidate, field.key)))
+			: []
+	);
+
+	/** Fields shown as a row: anything the two disagree on, plus anything edited. */
+	const shown = $derived(
+		MERGE_FIELDS.filter((field) => differing.includes(field) || field.key in draft)
+	);
+
+	const agreed = $derived(MERGE_FIELDS.filter((field) => !shown.includes(field)));
+
+	const groups = $derived([...new Set(shown.map((field) => field.group))]);
+
+	async function submit() {
+		if (busy) return;
+		busy = true;
+
+		const merged = await mergeInto(libraryId, survivor, folded, changed.length ? draft : undefined);
+
+		busy = false;
+
+		if (merged) {
+			open = false;
+			onmerged?.(survivor, folded);
+		}
+	}
+</script>
+
+{#snippet cover(book: Book, size: string)}
+	<span class="{size} shrink-0 overflow-hidden rounded-sm border bg-muted">
+		{#if book.cover_image}
+			<BookImage src="/api/{book.cover_image}" class="h-full w-full object-cover" />
+		{:else}
+			<!-- Drawn rather than left blank, so the rail tells two coverless books
+			     apart the same way the shelves do. -->
+			<GeneratedCover {book} />
+		{/if}
+	</span>
+{/snippet}
+
+<!-- The inert reference column: the same shape as the control opposite it, with
+     nothing that invites a click. -->
+{#snippet reference(field: FieldSpec, book: Book)}
+	{@const value = readField(book, field.key)}
+	<div class="flex min-w-0 flex-col gap-1">
+		<span class="text-[11px] text-muted-foreground">{field.label}</span>
+		{#if isEmpty(value)}
+			<span class="min-h-8 py-1 text-sm text-muted-foreground italic">empty</span>
+		{:else if field.kind === 'list'}
+			<span class="flex min-h-8 flex-wrap items-center gap-1 py-0.5">
+				{#each value as string[] as item (item)}
+					<Badge variant="secondary" class="font-normal">{item}</Badge>
+				{/each}
+			</span>
+		{:else}
+			<span class="min-h-8 py-1 text-sm break-words">{displayValue(value)}</span>
+		{/if}
+	</div>
+{/snippet}
+
+<Dialog.Root bind:open>
+	<Dialog.Content
+		class="flex h-[85vh] max-w-[min(72rem,95vw)] flex-col gap-0 overflow-hidden p-0 sm:max-w-[min(72rem,95vw)]"
+	>
+		<Dialog.Header class="shrink-0 space-y-0 border-b px-5 py-3 text-left">
+			<Dialog.Title class="font-serif text-base font-normal">
+				Merge {books.length} books
+			</Dialog.Title>
+			<Dialog.Description class="text-xs">
+				{candidates.length}
+				{candidates.length === 1 ? 'record is' : 'records are'} deleted. Their files move onto the book
+				you keep — nothing is removed from disk.
+			</Dialog.Description>
+		</Dialog.Header>
+
+		<div class="grid min-h-0 flex-1 grid-cols-1 md:grid-cols-[15rem_1fr]">
+			<!-- Rail: the books being folded in, one open at a time -->
+			<aside
+				class="flex min-w-0 flex-col overflow-y-auto border-b bg-sidebar md:border-r md:border-b-0"
+			>
+				<p
+					class="px-3 pt-3 pb-1 font-mono text-[10px] tracking-widest text-muted-foreground uppercase"
+				>
+					Taking from · {candidates.length}
+				</p>
+				{#each candidates as book (book.id)}
+					{@const count = MERGE_FIELDS.filter(
+						(field) => !isSame(readField(survivor, field.key), readField(book, field.key))
+					).length}
+					<button
+						type="button"
+						onclick={() => (candidateId = book.id)}
+						class="flex items-start gap-2 border-l-2 px-3 py-2 text-left transition-colors hover:bg-muted/50 {book.id ===
+						candidate?.id
+							? 'border-l-primary bg-background'
+							: 'border-l-transparent'}"
+					>
+						{@render cover(book, 'h-10 w-7')}
+						<span class="min-w-0 flex-1">
+							<span class="line-clamp-2 font-serif text-[13px]">{book.title}</span>
+							<span class="block text-[10px] text-muted-foreground">
+								#{book.id} · {book.files.length}
+								{book.files.length === 1 ? 'file' : 'files'}
+							</span>
+							{#if count === 0}
+								<Badge variant="secondary" class="mt-1 text-[10px] font-normal">
+									nothing to take
+								</Badge>
+							{/if}
+						</span>
+					</button>
+				{/each}
+			</aside>
+
+			<div class="flex min-h-0 flex-col">
+				<!-- Which record survives -->
+				<div class="flex shrink-0 flex-wrap items-center gap-2 border-b px-5 py-2.5">
+					<span class="text-[11px] text-muted-foreground">Keeping</span>
+					{#each books as book (book.id)}
+						<button
+							type="button"
+							onclick={() => (survivorId = book.id)}
+							class="flex items-center gap-1.5 rounded-md border px-2 py-1 text-xs transition-colors {book.id ===
+							survivorId
+								? 'border-primary bg-accent text-accent-foreground'
+								: 'text-muted-foreground hover:bg-muted'}"
+						>
+							{@render cover(book, 'h-6 w-4')}
+							<span class="max-w-32 truncate">#{book.id}</span>
+						</button>
+					{/each}
+
+					<div class="ml-auto flex gap-2">
+						<Button variant="outline" size="sm" onclick={fillEmpty} disabled={!candidate}>
+							Fill empty fields
+						</Button>
+					</div>
+				</div>
+
+				<div class="min-h-0 flex-1 overflow-y-auto px-5 py-4">
+					{#if !candidate}
+						<p class="text-sm text-muted-foreground">Nothing left to fold in.</p>
+					{:else if shown.length === 0}
+						<p class="text-sm text-muted-foreground">
+							These records agree on every field. Merging keeps
+							<span class="font-medium text-foreground">#{survivor.id}</span> and moves the others' files
+							onto it.
+						</p>
+					{:else}
+						{#each groups as group (group)}
+							<h3
+								class="mt-5 border-b pb-1 font-mono text-[10px] tracking-widest text-primary uppercase first:mt-0"
+							>
+								{group}
+							</h3>
+
+							{#each shown.filter((field) => field.group === group) as field (field.key)}
+								{@const value = current(field)}
+								{@const incoming = readField(candidate, field.key)}
+								{@const edited = field.key in draft}
+								<div class="grid grid-cols-[1fr_auto_1fr] items-center gap-3 border-b py-2">
+									{@render reference(field, candidate)}
+
+									<!-- Actions, on the row rather than stacked beside it -->
+									<div class="flex items-center gap-1">
+										{#if edited}
+											<Button
+												variant="ghost"
+												size="icon"
+												class="size-7"
+												title="Undo"
+												onclick={() => undo(field)}
+											>
+												<Undo2 class="size-3.5" />
+												<span class="sr-only">Undo {field.label}</span>
+											</Button>
+										{/if}
+
+										{#if !isSame(value, incoming)}
+											{#if isEmpty(value)}
+												<!-- Nothing to weigh, so it is an offer rather than a choice -->
+												<Button
+													size="sm"
+													class="h-7 gap-1 px-2 text-[11px]"
+													onclick={() => take(field, 'replace')}
+												>
+													<ArrowRight class="size-3.5" />
+													Take
+												</Button>
+											{:else}
+												<Button
+													variant="outline"
+													size="icon"
+													class="size-7"
+													title="Replace"
+													onclick={() => take(field, 'replace')}
+												>
+													<ArrowRight class="size-3.5" />
+													<span class="sr-only">Replace {field.label}</span>
+												</Button>
+												{#each field.extra as action (action)}
+													<Button
+														variant="outline"
+														size="icon"
+														class="size-7"
+														title={action === 'merge' ? 'Merge' : 'Append'}
+														onclick={() => take(field, action)}
+													>
+														{#if action === 'merge'}
+															<GitMerge class="size-3.5" />
+														{:else}
+															<Plus class="size-3.5" />
+														{/if}
+														<span class="sr-only">{action} {field.label}</span>
+													</Button>
+												{/each}
+											{/if}
+										{/if}
+									</div>
+
+									<!-- The survivor's side: the edit form, live -->
+									<div class="flex min-w-0 flex-col gap-1">
+										<span class="flex items-center gap-1.5 text-[11px] text-muted-foreground">
+											{field.label}
+											{#if edited}
+												<Badge class="h-4 px-1.5 text-[9px] font-semibold">taken</Badge>
+											{/if}
+										</span>
+
+										{#if field.kind === 'longtext'}
+											<Textarea
+												rows={4}
+												class="text-sm"
+												value={(value as string) ?? ''}
+												oninput={(event) => (draft[field.key] = event.currentTarget.value)}
+											/>
+										{:else if field.kind === 'list' || field.kind === 'keyed'}
+											<!-- Edited through the take actions; typing here would need the
+											     tags and key/value editors, which belong to the edit form. -->
+											<div
+												class="flex min-h-8 flex-wrap items-center gap-1 rounded-md border bg-background px-2 py-1"
+											>
+												{#if isEmpty(value)}
+													<span class="text-sm text-muted-foreground">—</span>
+												{:else if field.kind === 'list'}
+													{#each value as string[] as item (item)}
+														<Badge variant="secondary" class="font-normal">{item}</Badge>
+													{/each}
+												{:else}
+													{#each Object.entries(value as Record<string, string>) as [name, id] (name)}
+														<Badge variant="secondary" class="font-mono text-[10px] font-normal">
+															{name}: {id}
+														</Badge>
+													{/each}
+												{/if}
+											</div>
+										{:else}
+											<Input
+												type={field.kind === 'number'
+													? 'number'
+													: field.kind === 'date'
+														? 'date'
+														: 'text'}
+												class="h-8 text-sm"
+												value={(value as string | number) ?? ''}
+												oninput={(event) =>
+													(draft[field.key] =
+														field.kind === 'number'
+															? Number(event.currentTarget.value) || null
+															: event.currentTarget.value)}
+											/>
+										{/if}
+									</div>
+								</div>
+							{/each}
+						{/each}
+
+						{#if agreed.length > 0}
+							<p class="pt-3 text-center text-[11px] text-muted-foreground italic">
+								{agreed.map((field) => field.label).join(', ')} — identical in both
+							</p>
+						{/if}
+					{/if}
+				</div>
+			</div>
+		</div>
+
+		<Dialog.Footer class="shrink-0 items-center gap-2 border-t px-5 py-3 sm:justify-between">
+			<p class="text-xs text-muted-foreground">
+				{#if changed.length}
+					{changed.length}
+					{changed.length === 1 ? 'change' : 'changes'} pending · this cannot be undone
+				{:else}
+					Metadata is left as #{survivor.id} has it · this cannot be undone
+				{/if}
+			</p>
+			<div class="flex gap-2">
+				<Button variant="outline" onclick={() => (open = false)}>Cancel</Button>
+				<Button onclick={submit} disabled={busy || candidates.length === 0}>
+					Merge into “{survivor.title}”
+				</Button>
+			</div>
+		</Dialog.Footer>
+	</Dialog.Content>
+</Dialog.Root>

frontend/src/lib/components/forms/merge-books/merge.ts

@@ -0,0 +1,44 @@
+import { invalidate } from '$app/navigation';
+import { toast } from 'svelte-sonner';
+
+import { mergeBooks } from '$lib/api/book.remote';
+import type { Book } from '$lib/schema';
+import type { FieldValue } from './field-spec';
+
+/**
+ * Fold `folded` into `survivor`, and tell the reader how it went.
+ *
+ * Shared because a merge is reachable two ways — the workbench, where the reader
+ * resolved the metadata field by field, and the quick action beside it, which
+ * takes the survivor's metadata as it stands. Only the `metadata` argument
+ * differs, and the invalidation and the wording should not.
+ *
+ * @returns whether the merge went through; the caller decides what to close or
+ * clear, and has no toast of its own to write either way.
+ */
+export async function mergeInto(
+	libraryId: number | string,
+	survivor: Book,
+	folded: Book[],
+	metadata?: Record<string, FieldValue>
+): Promise<boolean> {
+	try {
+		await mergeBooks({
+			library_id: libraryId,
+			survivor_id: survivor.id,
+			merged_ids: folded.map((book) => book.id),
+			metadata
+		});
+
+		// Both, because a merge is started from the duplicates review and from the
+		// library's selection toolbar, and each page depends on a different one.
+		await Promise.all([invalidate('app:books'), invalidate('app:duplicate-books')]);
+
+		toast.success(`Merged into “${survivor.title}”`);
+		return true;
+	} catch (error) {
+		console.error('Failed to merge books', error);
+		toast.error(error instanceof Error ? error.message : 'Could not merge these books');
+		return false;
+	}
+}

frontend/src/lib/components/icons/UnjsDb0.svelte

@@ -1,73 +0,0 @@
-<svg xmlns="http://www.w3.org/2000/svg" width="1em" height="1em" viewBox="0 0 32 32" {...$$props}
-	><g fill="none"
-		><path
-			fill="#01579B"
-			d="m29.508 25.58l-11.435 5.17c-.706.332-1.44.3-2.116-.09L1.524 23.33c-.413-.265-.536-.652-.51-.922c.025-.27.087-.563.812-.773l1.07-.395l14.48 7.893l10.29-4.206z"
-		/><path
-			fill="#F5F5F5"
-			d="M17.935 29.923a1.99 1.99 0 0 1-1.815-.065l-14.093-7.1a.395.395 0 0 1-.14-.558c.31-.512.88-2.133-.06-3.477l15.916 7.662z"
-		/><path
-			fill="#94C6D6"
-			d="m28.898 24.995l-10.963 4.927c-.362.158-1.085.438-1.918-.122c.658.047 1.12-.225 1.358-.668c.233-.43.163-1.135-.12-1.532c-.172-.24-.635-.623-.837-.838l11.857-4.637c1.05-.433 2.035.215 2.193 1.003c.174.89-.96 1.617-1.57 1.867"
-		/><path
-			fill="#01579B"
-			d="m29.445 21.74l-11.317 5.05a2.67 2.67 0 0 1-2.193-.1l-14.46-7.345a.94.94 0 0 1-.43-.525c-.135-.437.035-.988.548-1.163l15.67 7.988l10.73-4.593z"
-		/><path
-			fill="#0091EA"
-			d="m30.298 22.473l-1.233-.448l-2.54.148l-8.395 3.747a2.67 2.67 0 0 1-2.192-.1L1.475 18.477a.438.438 0 0 1 .117-.82l10.423-4.662c.297-.055.602-.022.88.095l14.872 7.09s2.363 1.617 2.53 2.293"
-		/><path
-			fill="#616161"
-			d="M26.383 22.245s1.565-.613 2.794-.558s1.658.918 1.658.918c-.233-1.058-1.325-1.598-1.325-1.598l-16.25-8.182c-.112-.047-.527-.145-1.165.118c-.515.212-2.198 1-2.198 1z"
-		/><path
-			fill="#424242"
-			d="M30.905 22.805c-.117-.467-.408-.967-.943-1.21c-.704-.317-1.71-.235-2.352.1l-1.227.545v.865l1.552-.69c1.51-.672 2.18.335 2.238.573c.24.967-.225 1.527-1.598 2.157l-2.23 1.005v.87l2.565-1.142c1.135-.455 2.43-1.31 1.995-3.073"
-		/><path
-			fill="#01579B"
-			d="M8.253 22.578L3.935 16.61l.677-.302l4.858 6.675zm5.537 2.75l-.77-.61l13.363-2.728v.438l-1.873.71z"
-		/><path
-			fill="#9CCC65"
-			d="m3.7 11.545l16.878-2.82l7.372 8.118a.689.689 0 0 1-.36 1.15l-17.425 3.575z"
-		/><path
-			fill="#689F38"
-			d="m27.59 17.293l-17.305 3.505l-.055.825l17.36-3.56a.686.686 0 0 0 .427-1.058a.67.67 0 0 1-.427.288m.933 3.782a.49.49 0 0 1-.318.74l-15.93 3.23c-.957.197-1.898-.43-1.982-1.405a1.635 1.635 0 0 1 1.297-1.742l15.32-3.44z"
-		/><path
-			fill="#616161"
-			d="m13.898 20.025l-6.345-9.08l-3.62.957c-.838.833-.525 2.2-.525 2.2s5.542 8.895 6.417 10.033s2.153.96 2.153.96l2.157-.435l-.225-4.025z"
-		/><path
-			fill="#424242"
-			d="m13.898 20.085l-3.048.63c-.832.188-.982.97-.982.97L2.51 11.143s-1.047 1.267-.352 2.345l7.667 10.647c.838 1.192 2.153.97 2.153.97l2.157-.435l-.223-3.945z"
-		/><path
-			fill="#B9E4EA"
-			d="M27.563 20.75a.29.29 0 0 1-.205.405L12.125 24.28c-.957.197-1.635-.438-1.6-1.303c.045-1.092.657-1.555 1.467-1.722L27.3 18.127s-.52.585-.123 1.68c.136.378.28.713.386.943"
-		/><path
-			stroke="#424242"
-			stroke-miterlimit="10"
-			stroke-width=".518"
-			d="M11.303 20.925L4.775 11.69"
-		/><path fill="#424242" d="m11.815 16.987l-8.395-4.23l-.34.67l9.7 4.943z" /><path
-			fill="#689F38"
-			d="m27.198 16.008l-.616-.675l-9.457 4.34l-5.31-2.628l.905 1.295l3.518 1.797l2.62-.402z"
-		/><path
-			fill="#C62828"
-			d="m29.505 14.338l-11.432 5.17c-.706.332-1.44.3-2.116-.09l-14.434-7.33c-.413-.265-.536-.653-.51-.923c.025-.27.087-.562.812-.772l.678-.25l14.83 7.277l12.042-4.983z"
-		/><path
-			fill="#F5F5F5"
-			d="M17.933 18.68a1.99 1.99 0 0 1-1.815-.065l-14.093-7.1a.395.395 0 0 1-.14-.558c.31-.512.88-2.132-.06-3.477l15.56 7.915z"
-		/><path
-			fill="#94C6D6"
-			d="M28.895 13.753L17.933 18.68c-.363.157-1.085.438-1.918-.122c.657.047 1.12-.226 1.357-.668c.233-.43.163-1.135-.12-1.532c-.172-.24-.634-.623-.837-.838l11.858-4.637c1.05-.433 2.035.215 2.192 1.002c.175.89-.96 1.617-1.57 1.867"
-		/><path
-			fill="#C62828"
-			d="m29.445 10.498l-11.317 5.05a2.67 2.67 0 0 1-2.193-.1L1.472 8.102a.9.9 0 0 1-.447-.54c-.108-.405.032-.938.565-1.148l13.253-2.807z"
-		/><path
-			fill="#F44336"
-			d="m30.295 11.23l-1.233-.448l-2.54.148l-8.394 3.747a2.67 2.67 0 0 1-2.193-.1L1.472 7.232c-.372-.19-.24-.692.118-.82l10.425-4.66c.297-.055.602-.022.88.095l14.872 7.09s2.36 1.615 2.528 2.293"
-		/><path
-			fill="#616161"
-			d="M26.383 11s1.302-.457 2.532-.402s1.922.762 1.922.762c-.252-1.13-1.325-1.598-1.325-1.598L13.263 1.58c-.112-.048-.527-.145-1.165.117a99 99 0 0 0-2.197 1z"
-		/><path fill="#424242" d="M27.87 10.465L11.243 2.077l.55-.247l16.91 8.475z" /><path
-			fill="#424242"
-			d="M30.903 11.563c-.118-.468-.316-.92-.873-1.155c-.713-.3-1.363-.363-2.422.045l-1.228.545v.865l1.553-.69c.787-.37 1.947-.29 2.237.572c.317.945-.225 1.528-1.598 2.158l-2.23 1.005v.87l2.566-1.143c1.137-.455 2.432-1.31 1.994-3.072"
-		/></g
-	></svg
->

frontend/src/lib/components/icons/chitai-mark.svelte

@@ -0,0 +1,29 @@
+<script lang="ts">
+	import type { SVGAttributes } from 'svelte/elements';
+
+	let { ...rest }: SVGAttributes<SVGElement> = $props();
+</script>
+
+<!--
+	Drawn in currentColor and stroked at lucide's weight so the mark sits with
+	the rest of the sidebar icons and picks up the theme's sidebar foreground.
+	Intrinsic size matches its rendered size — see app-sidebar.svelte.
+-->
+<svg
+	xmlns="http://www.w3.org/2000/svg"
+	width="28"
+	height="28"
+	viewBox="0 0 24 24"
+	fill="none"
+	stroke="currentColor"
+	stroke-width="1.6"
+	stroke-linecap="round"
+	stroke-linejoin="round"
+	aria-hidden="true"
+	{...rest}
+>
+	<path d="M4 5.4A2.4 2.4 0 0 1 6.4 3H20v15.2H6.4A2.4 2.4 0 0 0 4 20.6z" />
+	<path d="M8.6 3v15.2" />
+	<path d="M12.4 7.6h4" />
+	<path d="M12.4 11.2h4" />
+</svg>

frontend/src/lib/components/layout/app-sidebar.svelte

@@ -1,45 +1,51 @@
 <script lang="ts">
+	import { resolve } from '$app/paths';
 	import { Separator } from '$lib/components/ui/separator/index.js';
 
-	import SettingsIcon from '@lucide/svelte/icons/settings';
-	import UnjsDb0 from '$lib/components/icons/UnjsDb0.svelte';
+	import ChitaiMark from '$lib/components/icons/chitai-mark.svelte';
 
 	import NavMain from './nav-main.svelte';
+	import NavUser from './nav-user.svelte';
 	import LibrarySwitcher from './library-switcher.svelte';
 	import * as Sidebar from '$lib/components/ui/sidebar/index.js';
 	import type { ComponentProps } from 'svelte';
 	import { getLibraryState } from '$lib/state/library.svelte';
-	import { page } from '$app/state';
 
 	let {
 		ref = $bindable(null),
 		collapsible = 'icon',
+		user,
 		...restProps
-	}: ComponentProps<typeof Sidebar.Root> = $props();
+	}: ComponentProps<typeof Sidebar.Root> & { user: { email: string } } = $props();
 
 	const libraryState = getLibraryState();
 
+	// Keep components out of these objects. `<header.icon />` is a *dynamic*
+	// component: every time the $derived object is recomputed — on hydration
+	// when LibraryState picks up previousLibraryId, and on every navigation for
+	// the footer — Svelte re-evaluates the expression and can remount the SVG,
+	// which reads as a flash and shifts layout. Referencing the component
+	// directly in the markup keeps it static.
 	const header = $derived({
 		title: 'chitai',
-		icon: UnjsDb0,
-		url: `/library/${libraryState.activeLibrary!.id}`
-	});
-
-	const footer = $derived({
-		title: 'Settings',
-		url: '/settings',
-		icon: SettingsIcon,
-		isActive: page.url.pathname.startsWith('/settings')
+		url: resolve('/(root)/(library)/library/[libraryId]', { libraryId: String(libraryState.activeLibrary!.id) })
 	});
 </script>
 
 <Sidebar.Root variant="floating" {collapsible} {...restProps} class="ml-1 py-3">
 	<Sidebar.Header class="h-(--header-height)">
-		<Sidebar.MenuButton>
+		<!--
+			The mark is 28px where the nav icons are 16px, and both start at the same
+			8px padding — so its centre sits 6px right of theirs. -ml-1.5 is that 6px,
+			which lines the logo up with the nav in both expanded and collapsed states.
+		-->
+		<Sidebar.MenuButton class="mt-1 -ml-1.5">
 			{#snippet child({ props })}
+				<!-- header.url is built with resolve(); the rule cannot trace the variable. -->
+				<!-- eslint-disable-next-line svelte/no-navigation-without-resolve -->
 				<a href={header.url} {...props}>
-					<header.icon class="mr-3 scale-175" />
-					<span class="text-xl font-semibold">{header.title}</span>
+					<ChitaiMark class="mr-3 size-7!" />
+					<span class="font-serif text-xl tracking-tight">{header.title}</span>
 				</a>
 			{/snippet}
 		</Sidebar.MenuButton>
@@ -50,15 +56,6 @@
 		<NavMain />
 	</Sidebar.Content>
 	<Sidebar.Footer>
-		<Sidebar.MenuButton class="mb-2" isActive={footer.isActive}>
-			{#snippet child({ props })}
-				<a href={footer.url} {...props}>
-					{#if footer.icon}
-						<footer.icon class="scale-125" />
-					{/if}
-					<span class="text-md pl-2">{footer.title}</span>
-				</a>
-			{/snippet}
-		</Sidebar.MenuButton>
+		<NavUser {user} />
 	</Sidebar.Footer>
 </Sidebar.Root>

frontend/src/lib/components/layout/library-switcher.svelte

@@ -35,6 +35,9 @@
 							<span class="ml-1 truncate font-semibold">
 								{libraryState.activeLibrary!.name}
 							</span>
+							<span class="ml-1 truncate font-mono text-xs tabular-nums text-muted-foreground">
+								{libraryState.activeLibrary!.total ?? 0} books
+							</span>
 						</div>
 						<ChevronsUpDownIcon class="mr-2 ml-auto" />
 					</Sidebar.MenuButton>

frontend/src/lib/components/layout/nav-main.svelte

@@ -1,60 +1,94 @@
 <script lang="ts">
+	import { resolve } from '$app/paths';
 	import { page } from '$app/state';
 	import * as Collapsible from '$lib/components/ui/collapsible/index.js';
 	import * as Sidebar from '$lib/components/ui/sidebar/index.js';
-    import { Badge } from "$lib/components/ui/badge/index.js";
+	import { Badge } from '$lib/components/ui/badge/index.js';
+	import { useSidebar } from '$lib/components/ui/sidebar/index.js';
 	import { getBookshelfState } from '$lib/state/bookshelf.svelte';
 	import { getLibraryState } from '$lib/state/library.svelte';
 	import { House, LibraryBig, Rows3, ChevronRightIcon } from '@lucide/svelte';
 
 	const libraryState = getLibraryState();
 	const bookshelfState = getBookshelfState();
+	const sidebar = useSidebar();
 
-    let items = $derived(
-        [
+	// Owned here so the collapsed rail can force it open when it expands the
+	// sidebar. Previously open={isActive}, which was always false — the Shelves
+	// item has no route of its own, so pathname never matched.
+	let shelvesOpen = $state(false);
+
+	// Deliberately a plain const, not $derived. These objects hold component
+	// references, and `<item.icon />` is a dynamic component — if the array were
+	// rebuilt on every navigation the icons would remount, flashing and shifting
+	// layout. Only the url and active state need to be reactive, so they are
+	// computed per-item in the markup instead.
+	//
+	// Active state is matched on route id rather than pathname. resolve() returns
+	// an absolute path on the client but a relative one during SSR, so comparing
+	// it to page.url.pathname would be false on the server and true after
+	// hydration — the highlight would flash in.
+	const items = [
 		{
 			title: 'Home',
-                url: `/library/${libraryState.activeLibrary?.id}`,
-                icon: House
+			icon: House,
+			routeId: '/(root)/(library)/library/[libraryId]',
+			path: (id?: number) =>
+				resolve('/(root)/(library)/library/[libraryId]', { libraryId: String(id ?? '') })
 		},
 		{
 			title: 'Library',
-                url: `/library/${libraryState.activeLibrary?.id}/view`,
-                icon: LibraryBig
+			icon: LibraryBig,
+			routeId: '/(root)/(library)/library/[libraryId]/view',
+			path: (id?: number) =>
+				resolve('/(root)/(library)/library/[libraryId]/view', { libraryId: String(id ?? '') })
 		},
-            {
-                title: 'Shelves',
-                url: '#',
-                icon: Rows3,
-                shelves: []
-            }
-        ].map((item) => ({
-            ...item,
-            isActive: page.url.pathname === item.url
-        }))
-    );
+		{ title: 'Shelves', icon: Rows3, routeId: null, path: () => '#', shelves: [] }
+	];
 </script>
 
-<Sidebar.Group>
-    <Sidebar.Menu>
-        {#each items as item (item.title)}
-            {#if 'shelves' in item}
-                <Collapsible.Root open={item.isActive} class="group/collapsible">
-                    {#snippet child({ props })}
-                        <Sidebar.MenuItem {...props}>
-                            <Collapsible.Trigger>
-                                {#snippet child({ props })}
-                                    <Sidebar.MenuButton {...props} tooltipContent={item.title} class="h-10">
-                                        {#if item.icon}
+{#snippet shelvesTrigger(item: { title: string; icon: typeof House })}
 	<item.icon class="scale-125" />
-                                        {/if}
 	<span class="text-md ml-2">{item.title}</span>
 	<ChevronRightIcon
 		class="ml-auto transition-transform duration-200 group-data-[state=open]/collapsible:rotate-90"
 	/>
+{/snippet}
+
+<Sidebar.Group>
+	<Sidebar.Menu>
+		{#each items as item (item.title)}
+			{@const url = item.path(libraryState.activeLibrary?.id)}
+			{@const isActive = item.routeId !== null && page.route.id === item.routeId}
+			{#if 'shelves' in item}
+				<Collapsible.Root bind:open={shelvesOpen} class="group/collapsible">
+					{#snippet child({ props })}
+						<Sidebar.MenuItem {...props}>
+							{#if sidebar.state === 'collapsed'}
+								<!--
+                                    Sidebar.MenuSub is group-data-[collapsible=icon]:hidden, so
+                                    toggling in the rail opens a list nobody can see. Expand the
+                                    sidebar and open the section instead of toggling.
+                                -->
+								<Sidebar.MenuButton
+									tooltipContent={item.title}
+									class="h-10"
+									onclick={() => {
+										sidebar.setOpen(true);
+										shelvesOpen = true;
+									}}
+								>
+									{@render shelvesTrigger(item)}
+								</Sidebar.MenuButton>
+							{:else}
+								<Collapsible.Trigger>
+									{#snippet child({ props })}
+										<Sidebar.MenuButton {...props} tooltipContent={item.title} class="h-10">
+											{@render shelvesTrigger(item)}
 										</Sidebar.MenuButton>
 									{/snippet}
 								</Collapsible.Trigger>
+							{/if}
 							<Collapsible.Content>
 								<Sidebar.MenuSub class="w-full">
 									{#each bookshelfState.getBookshelves(libraryState.activeLibrary!.id) ?? [] as shelf (shelf.id)}
@@ -62,17 +96,18 @@
 											<Sidebar.MenuSubButton>
 												{#snippet child({ props })}
 													<a
-                                                            href={`/library/${libraryState.activeLibrary!.id}/view?shelves=${shelf.id}`}
+														href="{resolve('/(root)/(library)/library/[libraryId]/view', {
+															libraryId: String(libraryState.activeLibrary!.id)
+														})}?shelves={shelf.id}"
 														{...props}
 													>
 														<Badge
 															variant="outline"
-                                                            class="scale-90 font-semibold bg-sidebar-primary text-sidebar-primary-foreground mr-1">
+															class="mr-1 scale-90 bg-sidebar-primary font-semibold text-sidebar-primary-foreground"
+														>
 															{shelf.total}
 														</Badge>
 														<span>{shelf.title}</span>
-                                                            
-                                                            
 													</a>
 												{/snippet}
 											</Sidebar.MenuSubButton>
@@ -85,9 +120,11 @@
 				</Collapsible.Root>
 			{:else}
 				<Sidebar.MenuItem>
-                    <Sidebar.MenuButton isActive={item.isActive} tooltipContent={item.title} class="h-10">
+					<Sidebar.MenuButton {isActive} tooltipContent={item.title} class="h-10">
 						{#snippet child({ props })}
-                            <a href={item.url} {...props}>
+							<!-- item.path() calls resolve(); the rule cannot trace the variable. -->
+							<!-- eslint-disable-next-line svelte/no-navigation-without-resolve -->
+							<a href={url} {...props}>
 								{#if item.icon}
 									<item.icon class="scale-125" />
 								{/if}

frontend/src/lib/components/layout/nav-user.svelte

@@ -0,0 +1,101 @@
+<script lang="ts">
+	import { resolve } from '$app/paths';
+	import * as Avatar from '$lib/components/ui/avatar/index.js';
+	import * as DropdownMenu from '$lib/components/ui/dropdown-menu/index.js';
+	import * as Sidebar from '$lib/components/ui/sidebar/index.js';
+	import { useSidebar } from '$lib/components/ui/sidebar/index.js';
+	import { logout } from '$lib/api';
+	import { goto } from '$app/navigation';
+	import ChevronsUpDownIcon from '@lucide/svelte/icons/chevrons-up-down';
+	import SettingsIcon from '@lucide/svelte/icons/settings';
+	import PaletteIcon from '@lucide/svelte/icons/palette';
+	import LogOutIcon from '@lucide/svelte/icons/log-out';
+
+	let { user }: { user: { email: string } } = $props();
+
+	const sidebar = useSidebar();
+
+	// No display name on the account yet, so the local part of the address
+	// stands in for one and its first two letters make the avatar.
+	const handle = $derived(user?.email?.split('@')[0] ?? 'Account');
+	const initials = $derived(handle.slice(0, 2).toUpperCase());
+</script>
+
+<Sidebar.Menu>
+	<Sidebar.MenuItem>
+		<DropdownMenu.Root>
+			<DropdownMenu.Trigger>
+				{#snippet child({ props })}
+					<Sidebar.MenuButton
+						{...props}
+						size="lg"
+						class="data-[state=open]:bg-sidebar-accent data-[state=open]:text-sidebar-accent-foreground"
+					>
+						<Avatar.Root class="size-8 rounded-lg">
+							<Avatar.Fallback
+								class="rounded-lg bg-sidebar-primary text-sidebar-primary-foreground"
+							>
+								{initials}
+							</Avatar.Fallback>
+						</Avatar.Root>
+						<div class="grid flex-1 text-left text-sm leading-tight">
+							<span class="truncate font-semibold">{handle}</span>
+							<span class="truncate font-mono text-[11px] text-muted-foreground">{user?.email}</span
+							>
+						</div>
+						<ChevronsUpDownIcon class="ml-auto size-4" />
+					</Sidebar.MenuButton>
+				{/snippet}
+			</DropdownMenu.Trigger>
+
+			<DropdownMenu.Content
+				class="w-(--bits-dropdown-menu-anchor-width) min-w-56 rounded-lg"
+				side={sidebar.isMobile ? 'bottom' : 'right'}
+				align="end"
+				sideOffset={4}
+			>
+				<DropdownMenu.Label class="p-0 font-normal">
+					<div class="flex items-center gap-2 px-1 py-1.5 text-left text-sm">
+						<Avatar.Root class="size-8 rounded-lg">
+							<Avatar.Fallback
+								class="rounded-lg bg-sidebar-primary text-sidebar-primary-foreground"
+							>
+								{initials}
+							</Avatar.Fallback>
+						</Avatar.Root>
+						<div class="grid flex-1 text-left text-sm leading-tight">
+							<span class="truncate font-semibold">{handle}</span>
+							<span class="truncate font-mono text-[11px] text-muted-foreground">{user?.email}</span
+							>
+						</div>
+					</div>
+				</DropdownMenu.Label>
+
+				<DropdownMenu.Separator />
+
+				<DropdownMenu.Group>
+					<DropdownMenu.Item onSelect={() => goto(resolve('/settings/account'))}>
+						<SettingsIcon class="size-4" />
+						Settings
+					</DropdownMenu.Item>
+					<DropdownMenu.Item onSelect={() => goto(resolve('/settings/appearance'))}>
+						<PaletteIcon class="size-4" />
+						Appearance
+					</DropdownMenu.Item>
+				</DropdownMenu.Group>
+
+				<DropdownMenu.Separator />
+
+				<DropdownMenu.Item
+					onSelect={async () => {
+						await logout();
+						await goto(resolve('/login'));
+					}}
+				>
+					<LogOutIcon class="size-4" />
+					Log out
+				</DropdownMenu.Item>
+			</DropdownMenu.Content>
+		</DropdownMenu.Root>
+	</Sidebar.MenuItem>
+</Sidebar.Menu>

frontend/src/lib/components/layout/search-form.svelte

@@ -1,4 +1,5 @@
 <script lang="ts">
+	import { resolve } from '$app/paths';
 	import * as Command from '$lib/components/ui/command/index';
 	import * as Kbd from '$lib/components/ui/kbd/index.js';
 	import * as InputGroup from '$lib/components/ui/input-group/index.js';
@@ -9,6 +10,7 @@
 	import { getLibraryState } from '$lib/state/library.svelte';
 	import type { PaginatedResponse, Book } from '$lib/schema';
 	import BookImage from '../view/book-image.svelte';
+	import GeneratedCover from '../view/generated-cover.svelte';
 	import { goto } from '$app/navigation';
 
 	const libraryState = getLibraryState();
@@ -59,14 +61,14 @@
 
 <svelte:document onkeydown={handleKeydown} />
 
-<form>
-	<div class="flex w-full max-w-xs flex-col gap-6">
+<form class="min-w-0 flex-1">
+	<div class="flex w-full max-w-xl flex-col">
 		<InputGroup.Root>
-			<InputGroup.Input placeholder="Search..." onclick={handleClick} />
+			<InputGroup.Input placeholder="Search books..." onclick={handleClick} />
 			<InputGroup.Addon>
 				<SearchIcon />
 			</InputGroup.Addon>
-			<InputGroup.Addon align="inline-end">
+			<InputGroup.Addon align="inline-end" class="hidden sm:flex">
 				<Kbd.Root>Ctrl</Kbd.Root>+
 				<Kbd.Root>k</Kbd.Root>
 			</InputGroup.Addon>
@@ -97,25 +99,33 @@
 						<Command.Item
 							value={String(book.id)}
 							onSelect={() => {
-								goto(`/book/${book.id}`);
+								goto(resolve('/(root)/(library)/book/[bookId]', { bookId: String(book.id) }));
 								open = false;
 							}}
 							class="cursor-pointer"
 						>
 							<div class="hover:bg-base-200 flex gap-4 p-3">
+								{#if book.cover_image}
 									<BookImage
 										src="/api/{book.cover_image}"
 										class="w-24 rounded object-cover shadow-lg"
 									/>
+								{:else}
+									<div class="aspect-9/12 w-24 shrink-0 overflow-hidden rounded shadow-lg">
+										<GeneratedCover {book} />
+									</div>
+								{/if}
 								<div class="flex-top flex flex-col">
-									<span class="text-lg font-medium">{book.title}</span>
+									<span class="font-serif text-lg">{book.title}</span>
 									<span class=" text-md">{book.subtitle}</span>
 									{#if book.authors.length > 0}
 										<span class="line-clamp-1 w-full text-sm text-muted-foreground">
 											by
 											{#each book.authors as author}
 												<a
-													href="/library/{libraryState.activeLibrary!.id}/view?authors={author.id}"
+													href="{resolve('/(root)/(library)/library/[libraryId]/view', {
+														libraryId: String(libraryState.activeLibrary!.id)
+													})}?authors={author.id}"
 													class="cs-list hover:underline">{author.name}</a
 												> &thinsp;
 											{/each}

frontend/src/lib/components/layout/site-header.svelte

@@ -11,21 +11,18 @@
 </script>
 
 <header class="sticky top-0 z-50 flex w-full items-center border-b bg-background">
-	<div class="my-2 flex h-(--header-height) w-full items-center gap-2 px-4">
-		<div class="flex w-full justify-between">
-			<div class="flex gap-3">
-				<Button class="size-8" variant="ghost" size="icon" onclick={sidebar.toggle}>
+	<div class="my-2 flex h-(--header-height) w-full items-center gap-3 px-4">
+		<Button class="size-8 shrink-0" variant="ghost" size="icon" onclick={sidebar.toggle}>
 			<SidebarIcon />
 		</Button>
-				<Separator orientation="vertical" class="mr-2 h-(--header-height) border" />
-			</div>
+		<Separator orientation="vertical" class="mr-1 h-6" />
 
+		<!-- Search takes the free space rather than sitting at a fixed width -->
 		<SearchForm />
 
-			<div class="flex w-24 items-center gap-4">
+		<div class="ml-auto flex shrink-0 items-center gap-2">
 			<UploadButton />
 			<ThemeToggle />
 		</div>
 	</div>
-	</div>
 </header>

frontend/src/lib/components/layout/theme-toggle.svelte

@@ -1,5 +1,5 @@
 <script lang="ts">
-	import { toggleMode, mode } from 'mode-watcher';
+	import { toggleMode } from 'mode-watcher';
 
 	import * as Tooltip from '$lib/components/ui/tooltip/index.js';
 	import { buttonVariants } from '$lib/components/ui/button/button.svelte';
@@ -9,25 +9,26 @@
 	let { class: className = '' }: { class?: string } = $props();
 </script>
 
+<!--
+	Which icon shows is decided by CSS off the `dark` class, not by reading
+	`mode.current` in the markup. The server cannot know the mode, so a
+	`{#if mode.current === 'light'}` renders the wrong branch during SSR and
+	visibly swaps on hydration.
+-->
 <Tooltip.Provider>
 	<Tooltip.Root>
 		<Tooltip.Trigger
 			onclick={toggleMode}
 			class="{buttonVariants({ variant: 'ghost', size: 'icon' })} scale-110 {className}"
 		>
-			{#if mode.current === 'light'}
-				<Moon class="scale-110" />
-			{:else}
-				<Sun class="scale-110" />
-			{/if}
+			<Moon class="scale-110 dark:hidden" />
+			<Sun class="hidden scale-110 dark:block" />
+			<span class="sr-only">Toggle light and dark mode</span>
 		</Tooltip.Trigger>
 
 		<Tooltip.Content>
-			{#if mode.current === 'light'}
-				<p>Dark Mode</p>
-			{:else}
-				<p>Light Mode</p>
-			{/if}
+			<p class="dark:hidden">Dark Mode</p>
+			<p class="hidden dark:block">Light Mode</p>
 		</Tooltip.Content>
 	</Tooltip.Root>
 </Tooltip.Provider>

frontend/src/lib/components/layout/upload-button.svelte

@@ -1,29 +1,16 @@
 <script>
 	import Upload from '@lucide/svelte/icons/upload';
-	import { buttonVariants } from '$lib/components/ui/button/button.svelte';
-	import * as Tooltip from '$lib/components/ui/tooltip/index';
+	import { Button } from '$lib/components/ui/button/index.js';
 	import BooksUpload from '$lib/components/forms/books-upload.svelte';
-	import { goto } from '$app/navigation';
 	import { getBookOperationsState } from '$lib/state/bookOperations.svelte';
 
 	const bookOps = getBookOperationsState();
 </script>
 
-<Tooltip.Provider>
-	<Tooltip.Root ignoreNonKeyboardFocus>
-		<Tooltip.Trigger
-			onclick={async () => {
-				bookOps.uploadDialogOpen = true;
-			}}
-			class="{buttonVariants({ variant: 'ghost', size: 'icon' })} scale-110"
-		>
-			<Upload class="scale-110" />
-		</Tooltip.Trigger>
-
-		<Tooltip.Content>
-			<p>Upload Book</p>
-		</Tooltip.Content>
-	</Tooltip.Root>
-</Tooltip.Provider>
+<!-- Labelled, so no tooltip: the button already says what it does. -->
+<Button onclick={() => (bookOps.uploadDialogOpen = true)}>
+	<Upload class="size-4" />
+	Upload
+</Button>
 
 <BooksUpload bind:open={bookOps.uploadDialogOpen} />

frontend/src/lib/components/layout/upload-tray.svelte

@@ -0,0 +1,213 @@
+<script lang="ts">
+	import { fly } from 'svelte/transition';
+	import { prefersReducedMotion } from 'svelte/motion';
+	import { resolve } from '$app/paths';
+	import { ChevronDown, CircleAlert, Copy, RotateCcw, X } from '@lucide/svelte';
+
+	import { Button } from '$lib/components/ui/button/index.js';
+	import { Spinner } from '$lib/components/ui/spinner/index';
+	import { getUploadQueueState } from '$lib/state/upload-queue.svelte';
+	import type { DuplicateBook, DuplicateFile } from '$lib/schema';
+	import { formatFileSize } from '$lib/utils';
+
+	const queue = getUploadQueueState();
+
+	// A clean run clears itself after a few seconds, so it needs to leave rather
+	// than blink out. Nothing to animate for anyone who asked not to see it.
+	const motion = $derived(prefersReducedMotion.current ? 0 : 200);
+
+	const percent = $derived(queue.total === 0 ? 0 : Math.round((queue.settled / queue.total) * 100));
+
+	const heading = $derived.by(() => {
+		const noun = queue.total === 1 ? 'book' : 'books';
+		if (queue.active) return `Adding ${queue.settled} of ${queue.total} ${noun}`;
+		if (queue.failed === 0 && queue.skipped === 0) return `Added ${queue.done} ${noun}`;
+		if (queue.done === 0 && queue.skipped === 0) return `Couldn't add ${queue.failed} ${noun}`;
+
+		const parts = [];
+		if (queue.done > 0) parts.push(`Added ${queue.done}`);
+		if (queue.skipped > 0) parts.push(`${queue.skipped} already here`);
+		if (queue.failed > 0) parts.push(`${queue.failed} failed`);
+		return parts.join(', ');
+	});
+
+	/**
+	 * What the reader needs to know about files that were not stored.
+	 *
+	 * One duplicate can name where its bytes already live; several would be a list,
+	 * and the row has no room for one, so they collapse to a count.
+	 */
+	function duplicateNote(duplicates: DuplicateFile[]) {
+		if (duplicates.length > 1) return `${duplicates.length} files are already in your library`;
+
+		const [only] = duplicates;
+		// No book to name: it matched another file in this same upload.
+		return only.book_title ? `Already in ${only.book_title}` : 'Already added by this upload';
+	}
+
+	/** The book a note can link to, when a single duplicate points at exactly one. */
+	function noteTarget(duplicates: DuplicateFile[]) {
+		return duplicates.length === 1 ? duplicates[0].book_id : null;
+	}
+
+	/**
+	 * What the reader needs to know about a book that went in and may already be here.
+	 *
+	 * Worded weaker than the file-level "Already in …" on purpose. That one means the
+	 * library holds these exact bytes; this one means the metadata agrees, which a
+	 * second edition, a translation and a re-scan all do. Nothing was refused.
+	 */
+	function possibleNote(candidates: DuplicateBook[]) {
+		if (candidates.length > 1)
+			return `Might already be in your library, ${candidates.length} times`;
+
+		return `Might already be in your library — ${candidates[0].title}`;
+	}
+</script>
+
+{#if queue.total > 0}
+	<!--
+		Docked rather than a toast. This runs for minutes and carries a row per
+		book, which a notification surface is not built to hold still for.
+	-->
+	<aside
+		aria-label="Uploads"
+		transition:fly={{ y: 12, duration: motion }}
+		onmouseenter={() => queue.hold()}
+		onmouseleave={() => queue.release()}
+		onfocusin={() => queue.hold()}
+		onfocusout={() => queue.release()}
+		class="fixed right-4 bottom-4 z-50 w-80 max-w-[calc(100vw-2rem)] overflow-hidden rounded-lg border bg-card shadow-lg"
+	>
+		<div class="flex items-center gap-2 border-b px-3 py-2">
+			{#if queue.active}
+				<Spinner class="size-4 shrink-0" />
+			{:else if queue.failed > 0}
+				<CircleAlert class="size-4 shrink-0 text-destructive" />
+			{:else if queue.skipped > 0}
+				<!-- Muted, not alarming: nothing went wrong, the books were already here. -->
+				<Copy class="size-4 shrink-0 text-muted-foreground" />
+			{/if}
+
+			<span class="min-w-0 flex-1 truncate text-sm font-medium">{heading}</span>
+
+			<Button
+				variant="ghost"
+				size="icon"
+				class="size-7 shrink-0"
+				onclick={() => (queue.collapsed = !queue.collapsed)}
+			>
+				<ChevronDown class="size-4 transition-transform {queue.collapsed ? '' : 'rotate-180'}" />
+				<span class="sr-only">{queue.collapsed ? 'Show' : 'Hide'} the list</span>
+			</Button>
+
+			<!-- Only once nothing is in flight: dismissing mid-run would suggest it
+			     stopped the upload, which it does not. -->
+			{#if !queue.active}
+				<Button variant="ghost" size="icon" class="size-7 shrink-0" onclick={() => queue.dismiss()}>
+					<X class="size-4" />
+					<span class="sr-only">Dismiss</span>
+				</Button>
+			{/if}
+		</div>
+
+		<div class="h-1 bg-muted">
+			<div
+				class="h-full bg-primary transition-[width] duration-300"
+				style="width: {percent}%"
+			></div>
+		</div>
+
+		{#if !queue.collapsed}
+			<ul class="flex max-h-64 flex-col divide-y overflow-y-auto">
+				{#each queue.jobs as job (job.id)}
+					{@const duplicates = job.duplicates ?? []}
+					{@const target = noteTarget(duplicates)}
+					{@const possible = job.possibleDuplicates ?? []}
+					<li class="flex min-w-0 items-center gap-2 px-3 py-2">
+						<span class="min-w-0 flex-1">
+							<span class="block truncate text-sm" title={job.label}>{job.label}</span>
+
+							{#if job.error}
+								<span class="block font-mono text-[10px] text-muted-foreground tabular-nums">
+									{job.error}
+								</span>
+							{:else if duplicates.length > 0}
+								<!--
+									Reported, with no offer to override. Storing the same bytes twice
+									splits reading progress and shelves across two records that can
+									never converge, and there is no version of that the reader wants.
+									`allow_duplicates` stays on the API for when the match is wrong.
+								-->
+								<span class="block min-w-0 text-[10px] text-muted-foreground">
+									{#if target}
+										<a
+											href={resolve('/(root)/(library)/book/[bookId]', { bookId: String(target) })}
+											class="truncate underline underline-offset-2"
+										>
+											{duplicateNote(duplicates)}
+										</a>
+									{:else}
+										<span class="truncate">{duplicateNote(duplicates)}</span>
+									{/if}
+								</span>
+							{:else if possible.length > 0}
+								<!--
+									A guess, not a fact, so it says so and stops there: the book was
+									stored, and the library's duplicates screen is where a reader
+									decides what to do about it.
+								-->
+								<span class="block min-w-0 text-[10px] text-muted-foreground">
+									{#if possible.length === 1}
+										<a
+											href={resolve('/(root)/(library)/book/[bookId]', {
+												bookId: String(possible[0].book_id)
+											})}
+											class="truncate underline underline-offset-2"
+										>
+											{possibleNote(possible)}
+										</a>
+									{:else}
+										<span class="truncate">{possibleNote(possible)}</span>
+									{/if}
+								</span>
+							{:else}
+								<span class="block font-mono text-[10px] text-muted-foreground tabular-nums">
+									{formatFileSize(job.size)}
+								</span>
+							{/if}
+						</span>
+
+						<span
+							class="shrink-0 font-mono text-[10px] tracking-wider uppercase
+								{job.status === 'done' ? 'text-primary' : ''}
+								{job.status === 'failed' ? 'text-destructive' : ''}
+								{job.status !== 'done' && job.status !== 'failed' ? 'text-muted-foreground' : ''}"
+						>
+							{#if job.status === 'uploading'}
+								Adding
+							{:else if job.status === 'done'}
+								Done
+							{:else if job.status === 'skipped'}
+								Skipped
+							{:else if job.status === 'failed'}
+								Failed
+							{:else}
+								Queued
+							{/if}
+						</span>
+					</li>
+				{/each}
+			</ul>
+
+			{#if !queue.active && queue.failed > 0}
+				<div class="border-t p-2">
+					<Button variant="outline" size="sm" class="w-full" onclick={() => queue.retryFailed()}>
+						<RotateCcw class="size-4" />
+						Retry {queue.failed} failed
+					</Button>
+				</div>
+			{/if}
+		{/if}
+	</aside>
+{/if}

frontend/src/lib/components/reader/chapter-sidebar.svelte

@@ -1,27 +0,0 @@
-<script>
-	import * as Sidebar from '$lib/components/ui/sidebar/index';
-
-	let { chapters } = $props();
-</script>
-
-<Sidebar.Root>
-	<Sidebar.Header />
-	<Sidebar.Content>
-		<Sidebar.GroupLabel>Chapters</Sidebar.GroupLabel>
-		<Sidebar.GroupContent>
-			<Sidebar.Menu>
-				{#each chapters as chapter}
-					<Sidebar.MenuItem>
-						<Sidebar.MenuButton>
-							{#snippet child({ props })}
-								<span>{chapter.label}</span>
-							{/snippet}
-						</Sidebar.MenuButton>
-					</Sidebar.MenuItem>
-				{/each}
-			</Sidebar.Menu>
-		</Sidebar.GroupContent>
-		<Sidebar.Group />
-	</Sidebar.Content>
-	<Sidebar.Footer />
-</Sidebar.Root>

frontend/src/lib/components/reader/epub-reader.svelte

@@ -1,354 +1,247 @@
 <script lang="ts">
-	//  TODO: Add type hints to the rest of this file
+	import { resolve } from '$app/paths';
+	import { onMount, untrack } from 'svelte';
 
-	import { browser } from '$app/environment';
-	import { onDestroy, onMount } from 'svelte';
+	import {
+		ChevronLeft,
+		ChevronRight,
+		PanelLeft,
+		RotateCcw,
+		Settings2,
+		TriangleAlert
+	} from '@lucide/svelte';
 
-	import { Book, Rendition } from 'epubjs';
-
-	import '../../../app.css';
-	import * as Sidebar from '$lib/components/ui/sidebar/index';
-	import * as Collapsible from '$lib/components/ui/collapsible/index';
-
-	import { ChevronDown, ChevronRight, ChevronLeft } from '@lucide/svelte';
+	import type { FoliateRelocateDetail, FoliateTocItem } from '$foliate/view.js';
+	import { ProgressReporter } from '$lib/reader/progress';
+	import { purgeLegacyLocationCache } from '$lib/reader/legacy-cache';
+	import { setReaderSettingsState } from '$lib/state/reader-settings.svelte';
 
+	import { Button, buttonVariants } from '$lib/components/ui/button/index.js';
 	import { Spinner } from '$lib/components/ui/spinner/index';
-	import type { DisplayedLocation } from 'epubjs/types/rendition';
+	import * as Sidebar from '$lib/components/ui/sidebar/index';
+	import * as Tooltip from '$lib/components/ui/tooltip/index';
+	import ThemeToggle from '$lib/components/layout/theme-toggle.svelte';
+	import FoliateView from './foliate-view.svelte';
+	import ReaderSettings from './reader-settings.svelte';
+	import ReaderToc from './reader-toc.svelte';
 
-	let { bookUrl, bookId, initialProgress = 0, initialEpubLoc = null } = $props();
+	let {
+		fileUrl,
+		bookId,
+		filename,
+		title = '',
+		initialProgress = 0,
+		initialEpubLoc = null
+	}: {
+		fileUrl: string;
+		bookId: string | number;
+		filename: string;
+		title?: string;
+		initialProgress?: number;
+		initialEpubLoc?: string | null;
+	} = $props();
 
-	let epubViewer = $state<HTMLElement>();
-	let containerWidth = $state(0);
-	let containerHeight = $state(0);
-	let isReaderVisible = $state(false);
+	const settingsState = setReaderSettingsState();
 
-	let hasNextPage = $state(true);
-	let hasPrevPage = $state(false);
+	// Read once: the reader is remounted per file, so these are fixed for its
+	// lifetime, and tracking them would restart the load mid-read.
+	const reporter = new ProgressReporter(untrack(() => bookId));
+	const initialCfi = untrack(() => initialEpubLoc);
+	const initialFraction = untrack(() => initialProgress);
 
-	let book: Book | undefined = $state();
-	let rendition = $state<Rendition>();
-	let chapters = $state([]);
-
-	let isMounted = $state(false);
-
-	let currentLocation = $state(initialEpubLoc);
-	let currentProgress = $state(initialProgress);
+	let file = $state<File>();
+	let loadError = $state<string | null>(null);
+	let isReady = $state(false);
 
+	let toc = $state<FoliateTocItem[]>([]);
+	let activeTocId = $state<number | null>(null);
 	let isSidebarOpen = $state(false);
-	let debounceTimeout = $state<NodeJS.Timeout>();
+	let isSettingsOpen = $state(false);
+	let isFixedLayout = $state(false);
 
-	// Function to update dimensions
-	function updateDimensions() {
-		if (epubViewer) {
-			// Get parent element dimensions
-			const parent = epubViewer.parentElement;
+	let progress = $state(initialFraction);
+	let viewer = $state<ReturnType<typeof FoliateView>>();
 
-			containerWidth = parent?.clientWidth! * 0.9 - 240;
-			containerHeight = window.innerHeight * 0.8;
-		}
-	}
+	const percent = $derived(Math.round(progress * 100));
 
-	// Handle window resize
-	function handleResize() {
-		updateDimensions();
-		if (rendition) {
-			rendition.resize(containerWidth, containerHeight);
-		}
-	}
-
-	async function handleKeydown(e: KeyboardEvent) {
-		if (e.key === 'ArrowLeft') {
-			await prevPage();
-		} else if (e.key === 'ArrowRight') {
-			await nextPage();
-		}
-	}
-
-	async function nextPage() {
-		if (hasNextPage) await rendition!.next();
-	}
-
-	async function prevPage() {
-		if (hasPrevPage) await rendition!.prev();
-	}
-
-	async function setUserBookProgress() {
-		clearTimeout(debounceTimeout);
-
-		debounceTimeout = setTimeout(async () => {
-			await fetch(`/api/books/progress/${bookId}`, {
-				method: 'POST',
-				body: JSON.stringify({
-					percentage: currentProgress,
-					epub_cfi: currentLocation,
-					completed: currentProgress === 1
-				})
-			});
-		}, 3000);
-	}
-
-	const getChapters = async (book: Book) => {
-		await book.ready;
-		// Get spine items (basic chapter structure)
-		const spineItems = book.spine.items.map((item, index) => ({
-			id: item.idref,
-			href: item.href,
-			index: index,
-			label: item.label || `Chapter ${index + 1}`,
-			cfi: book.spine.get(index).cfiBase // Get CFI from spine
-		}));
-
-		// Get table of contents for better labels
-		const toc = await book.loaded.navigation;
-
-		// Combine spine items with TOC information
-		const chapters = toc.toc.map((chapter) => {
-			const spineItem = spineItems.find((item: any) => item.href === chapter.href);
-			return {
-				...spineItem,
-				label: chapter.label || spineItem?.label,
-				subitems: chapter.subitems,
-				href: chapter.href,
-				cfi: spineItem?.cfi || book.spine.get(chapter.href)?.cfiBase
-			};
-		});
-
-		return chapters;
-	};
-	async function navigateToChapter(chapter: any) {
-		try {
-			if (!rendition || !book) return;
-
-			// Ensure book is ready
-			await book.ready;
-
-			// Try different navigation methods
-			if (chapter.href) {
-				await rendition.display(chapter.href);
-			} else {
-				console.error('No valid navigation target found for chapter:', chapter);
-			}
-		} catch (error) {
-			console.error('Error navigating to chapter:', error);
-		}
-	}
-
-	onMount(async () => {
-		if (browser) {
-			// Initial setup
-			updateDimensions();
+	/**
+	 * Fetches the book ourselves rather than handing foliate the URL.
+	 *
+	 * view.open(url) would route through foliate's fetchFile, which names the File
+	 * after the URL path — "34", with no extension — and makeBook's CBZ/FB2 checks
+	 * are filename-based. Doing it here also keeps the response.ok check: the proxy
+	 * answers a failure with a SvelteKit error page, and fetch resolves on a 404,
+	 * so without it a missing file surfaced only as an opaque parse error.
+	 */
+	async function loadFile() {
+		loadError = null;
+		isReady = false;
+		file = undefined;
 
 		try {
-				// Fetch the EPUB file
-				const response = await fetch(bookUrl);
-				const arrayBuffer = await response.arrayBuffer();
+			const response = await fetch(fileUrl);
+			if (!response.ok) throw new Error(`The server returned ${response.status} for this file.`);
 
-				// Create book from array buffer
-				book = new Book();
-				await book.open(arrayBuffer, 'binary');
-
-				// Generate locations if they do not exist in localStorage
-				let existingLocations = localStorage.getItem(`${bookId}-locations`);
-				let locations;
-				if (existingLocations) {
-					locations = JSON.parse(existingLocations);
-					book.locations.load(locations);
-				} else {
-					locations = await book.locations.generate(1600);
-					// Save locations to localStorage
-					localStorage.setItem(`${bookId}-locations`, JSON.stringify(locations));
-				}
-
-				await book.ready;
-
-				// Render the book to the viewer element
-				rendition = book.renderTo('epub-viewer', {
-					width: containerWidth,
-					height: containerHeight,
-					spread: 'auto',
-					snap: true,
-					manager: 'continuous',
-					flow: 'paginated'
-				});
-
-				// Set the key listener on the iframe element
-				let keyListener = async function (e: any) {
-					// Left Key
-					if ((e.keyCode || e.which) == 37) {
-						await prevPage();
-					}
-					// Right Key
-					if ((e.keyCode || e.which) == 39) {
-						await nextPage();
-					}
-				};
-
-				// Add resize listener
-				window.addEventListener('resize', handleResize);
-
-				// Add Key listener
-				rendition.on('keydown', keyListener);
-
-				// Listen to location changes
-				rendition.on('locationChanged', async (location: DisplayedLocation) => {
-					if (!location?.start) return;
-
-					currentLocation = rendition!.currentLocation().start.cfi;
-					currentProgress = book?.locations.percentageFromCfi(currentLocation);
-
-					await setUserBookProgress();
-
-					hasNextPage = !rendition!.location.atEnd;
-					hasPrevPage = !rendition!.location.atStart;
-				});
-
-				chapters = await getChapters(book);
-
-				let initialLocationCfi =
-					currentLocation || book.locations.cfiFromPercentage(currentProgress);
-
-				if (initialLocationCfi) {
-					await rendition.display(initialLocationCfi);
-				} else {
-					await rendition.display();
-				}
-
-				isMounted = true;
-				isReaderVisible = true;
+			file = new File([await response.blob()], filename, { type: 'application/epub+zip' });
 		} catch (error) {
-				console.error('Error loading EPUB', error);
+			console.error('Could not download the book', error);
+			loadError = error instanceof Error ? error.message : 'The file could not be downloaded.';
 		}
 	}
-	});
 
-	onDestroy(() => {
-		if (browser) {
-			window.removeEventListener('resize', handleResize);
-			book.destroy();
+	function handleRelocate(detail: FoliateRelocateDetail) {
+		progress = detail.fraction;
+		activeTocId = detail.tocItem?.id ?? null;
+		reporter.record(detail.fraction, detail.cfi);
 	}
+
+	function navigateToChapter(href: string) {
+		// Deliberately leaves the sidebar open: picking a chapter is usually part of
+		// browsing several, and closing it each time made that tedious.
+		void viewer?.goTo(href);
+	}
+
+	onMount(() => {
+		purgeLegacyLocationCache();
+		reporter.listen();
+		void loadFile();
+
+		return () => reporter.dispose();
 	});
 </script>
 
-<svelte:document onkeydown={handleKeydown} />
+<svelte:head>
+	<title>{title ? `${title} — Chitai` : 'Reader — Chitai'}</title>
+</svelte:head>
 
-<Sidebar.Provider bind:open={isSidebarOpen}>
-	<Sidebar.Root class={!isReaderVisible ? 'hidden' : ''}>
-		<Sidebar.Header />
-		<Sidebar.Content>
-			<Sidebar.GroupLabel>Chapters</Sidebar.GroupLabel>
-			<Sidebar.GroupContent>
-				<Sidebar.Menu>
-					{#each chapters as chapter}
-						{#if chapter.subitems.length > 0}
-							<Collapsible.Root class="group/collapsible">
-								<div class="flex w-full items-center gap-1">
-									<Sidebar.MenuItem class="min-w-0 flex-1">
-										<Sidebar.MenuButton
-											class="w-full"
-											onclick={async () => await navigateToChapter(chapter)}
+<Sidebar.Provider bind:open={isSidebarOpen} class="min-h-0">
+	{#if isReady}
+		<ReaderToc {toc} activeId={activeTocId} onnavigate={navigateToChapter} />
+	{/if}
+
+	<main class="flex h-screen w-full flex-col overflow-hidden bg-background">
+		<!-- Reader chrome: somewhere to go back to, what you are reading, how far in -->
+		<header class="flex h-12 shrink-0 items-center gap-2 border-b px-3">
+			<Tooltip.Provider>
+				<Tooltip.Root ignoreNonKeyboardFocus>
+					<Tooltip.Trigger
+						class="{buttonVariants({ variant: 'ghost', size: 'icon' })} size-8"
+						onclick={() => (isSidebarOpen = !isSidebarOpen)}
+						disabled={!isReady || toc.length === 0}
 					>
-											<span class="block truncate" title={chapter.label}>{chapter.label}</span>
-										</Sidebar.MenuButton>
-									</Sidebar.MenuItem>
-									<Collapsible.Trigger class="flex-shrink-0 p-2">
-										<ChevronDown
-											class="h-4 w-4 transition-transform group-data-[state=open]/collapsible:rotate-180"
-										/>
-									</Collapsible.Trigger>
+						<PanelLeft class="size-4" />
+					</Tooltip.Trigger>
+					<Tooltip.Content side="bottom"><p>Chapters</p></Tooltip.Content>
+				</Tooltip.Root>
+			</Tooltip.Provider>
+
+			<a
+				href={resolve('/(root)/(library)/book/[bookId]', { bookId: String(bookId) })}
+				class="min-w-0 flex-1 truncate font-serif text-sm hover:underline"
+				{title}
+			>
+				{title || 'Reader'}
+			</a>
+
+			{#if isReady}
+				<span class="hidden items-center gap-2 sm:flex">
+					<span class="h-1 w-24 overflow-hidden rounded-full bg-muted-foreground/25">
+						<span class="block h-full bg-flag" style="width: {percent}%;"></span>
+					</span>
+					<span class="font-mono text-xs text-muted-foreground tabular-nums">{percent}%</span>
+				</span>
+			{/if}
+
+			<Tooltip.Provider>
+				<Tooltip.Root ignoreNonKeyboardFocus>
+					<Tooltip.Trigger
+						class="{buttonVariants({ variant: 'ghost', size: 'icon' })} size-8"
+						onclick={() => (isSettingsOpen = true)}
+						disabled={!isReady}
+					>
+						<Settings2 class="size-4" />
+					</Tooltip.Trigger>
+					<Tooltip.Content side="bottom"><p>Reading settings</p></Tooltip.Content>
+				</Tooltip.Root>
+			</Tooltip.Provider>
+
+			<ThemeToggle />
+		</header>
+
+		{#if loadError}
+			<div class="flex flex-1 items-center justify-center p-6">
+				<div class="flex max-w-sm flex-col items-center gap-3 text-center">
+					<TriangleAlert class="size-8 text-destructive" />
+					<h2 class="font-serif text-lg">This book wouldn't open</h2>
+					<p class="text-sm text-muted-foreground">{loadError}</p>
+					<div class="mt-2 flex gap-2">
+						<Button onclick={loadFile}>
+							<RotateCcw class="size-4" />
+							Try again
+						</Button>
+						<a
+							href={resolve('/(root)/(library)/book/[bookId]', { bookId: String(bookId) })}
+							class={buttonVariants({ variant: 'outline' })}
+						>
+							Back to book
+						</a>
+					</div>
+				</div>
 			</div>
-								<Collapsible.Content>
-									<Sidebar.MenuSub>
-										{#each chapter.subitems as subchapter}
-											<Sidebar.MenuSubItem class="min-w-0">
-												<Sidebar.MenuButton
-													class="w-full"
-													onclick={async () => await navigateToChapter(subchapter)}
-												>
-													<span class="block truncate" title={subchapter.label}
-														>{subchapter.label}</span
-													>
-												</Sidebar.MenuButton>
-											</Sidebar.MenuSubItem>
-										{/each}
-									</Sidebar.MenuSub>
-								</Collapsible.Content>
-							</Collapsible.Root>
 		{:else}
-							<Sidebar.MenuItem class="min-w-0">
-								<Sidebar.MenuButton
-									class="w-full"
-									onclick={async () => await navigateToChapter(chapter)}
-								>
-									<span class="block truncate">{chapter.label}</span>
-								</Sidebar.MenuButton>
-							</Sidebar.MenuItem>
-						{/if}
-					{/each}
-				</Sidebar.Menu>
-			</Sidebar.GroupContent>
-			<Sidebar.Group />
-		</Sidebar.Content>
-		<Sidebar.Footer />
-	</Sidebar.Root>
-	<main class="flex w-full overflow-hidden">
-		{#if browser}
-			{#if !isReaderVisible}
-				<div class="absolute inset-0 flex items-center justify-center">
+			{#if !isReady}
+				<div class="flex flex-1 items-center justify-center gap-3">
 					<Spinner />
+					<span class="text-sm text-muted-foreground">Opening…</span>
 				</div>
 			{/if}
 
-			<div class="h-full w-full {isReaderVisible ? '' : 'opacity-0'}">
-				<div class="flex">
-					<Sidebar.Trigger />
-				</div>
+			<div class="min-h-0 flex-1 p-4 {isReady ? '' : 'hidden'}">
+				<!-- The page itself: same colour the book is themed to, so the two
+				     meet seamlessly and the border reads as the edge of the sheet. -->
+				<div class="relative h-full overflow-hidden rounded-lg border bg-card shadow-sm">
+					{#if file}
+						<FoliateView
+							bind:this={viewer}
+							{file}
+							{initialCfi}
+							{initialFraction}
+							settings={settingsState.settings}
+							onrelocate={handleRelocate}
+							onerror={(message) => (loadError = message)}
+							onready={(detail) => {
+								toc = detail.toc;
+								isFixedLayout = detail.isFixedLayout;
+								isReady = true;
+							}}
+						/>
+					{/if}
 
-				<div class="mt-[-30px] flex h-full w-full items-center justify-center">
-					<ChevronLeft
-						onclick={prevPage}
-						class={hasPrevPage
-							? 'text-primary/70 hover:cursor-pointer hover:text-primary'
-							: 'text-muted'}
-					/>
-					<div
-						id="epub-viewer"
-						bind:this={epubViewer}
-						class="h-[{containerHeight}px] w-[{containerWidth}] epub-content mx-8 rounded border-2 p-8 shadow-md"
-					></div>
-					<ChevronRight
-						onclick={nextPage}
-						class={hasNextPage
-							? 'text-primary/70 hover:cursor-pointer hover:text-primary'
-							: 'text-muted'}
-					/>
+					<!-- Overlaid rather than placed beside the page, so they stay within
+					     reach of the text. They sit in foliate's own page margin. -->
+					<Button
+						variant="ghost"
+						size="icon"
+						class="absolute top-1/2 left-1 size-10 -translate-y-1/2 opacity-40 transition-opacity hover:opacity-100"
+						onclick={() => viewer?.goBack()}
+						aria-label="Previous page"
+					>
+						<ChevronLeft class="size-5" />
+					</Button>
+
+					<Button
+						variant="ghost"
+						size="icon"
+						class="absolute top-1/2 right-1 size-10 -translate-y-1/2 opacity-40 transition-opacity hover:opacity-100"
+						onclick={() => viewer?.goForward()}
+						aria-label="Next page"
+					>
+						<ChevronRight class="size-5" />
+					</Button>
 				</div>
 			</div>
 		{/if}
 	</main>
 </Sidebar.Provider>
 
-<style>
-	.epub-content {
-		position: relative;
-		background: white;
-	}
-
-	/* Position separator relative to epub content */
-	@media (min-width: 1209px) {
-		.epub-content:after {
-			/* Calculate position based on content width */
-			--separator-position: calc(var(--content-width, 100%) / 2);
-			position: absolute;
-			width: 1px;
-			border-right: 1px #000 solid;
-			height: 90%;
-			z-index: 1;
-			left: 50%;
-			transform: translateX(-50%);
-			top: 5%;
-			opacity: 0.15;
-			box-shadow: -2px 0 15px rgba(0, 0, 0, 1);
-			content: '';
-		}
-	}
-</style>
+<ReaderSettings bind:open={isSettingsOpen} {isFixedLayout} />

frontend/src/lib/components/reader/foliate-view.svelte

@@ -0,0 +1,305 @@
+<script lang="ts">
+	import { onMount } from 'svelte';
+
+	import { mode } from 'mode-watcher';
+
+	import type {
+		FoliateLoadDetail,
+		FoliateRelocateDetail,
+		FoliateTarget,
+		FoliateTocItem,
+		View
+	} from '$foliate/view.js';
+	import { createFoliateView, loadFoliate } from '$lib/reader/foliate';
+	import type { ReaderSettings } from '$lib/schema/reader';
+	import { toRendererAttributes } from '$lib/reader/settings';
+	import { buildReaderStyles, readReaderPalette } from '$lib/reader/stylesheet';
+
+	interface ReadyDetail {
+		toc: FoliateTocItem[];
+		isFixedLayout: boolean;
+		rtl: boolean;
+	}
+
+	let {
+		file,
+		initialCfi = null,
+		initialFraction = 0,
+		settings,
+		onready,
+		onrelocate,
+		onerror
+	}: {
+		file: File;
+		initialCfi?: string | null;
+		initialFraction?: number;
+		settings: ReaderSettings;
+		onready?: (detail: ReadyDetail) => void;
+		onrelocate?: (detail: FoliateRelocateDetail) => void;
+		onerror?: (message: string) => void;
+	} = $props();
+
+	let container = $state<HTMLDivElement>();
+
+	/** Not $state: nothing in the markup reads it, and it must not be proxied. */
+	let view: View | undefined;
+	let rtl = false;
+
+	// Reactive so the effects below re-run once the book is open — otherwise a
+	// theme flip while it was still loading would never reach the book.
+	let ready = $state(false);
+	let hostWidth = $state(0);
+
+	/**
+	 * How many columns the paginator will actually lay out.
+	 *
+	 * It decides this internally and exposes it only as a custom property inside
+	 * its shadow root, so the formula is mirrored here to place the spread
+	 * divider. Kept in step with paginator.js's `divisor`.
+	 */
+	const columns = $derived(
+		settings.flow === 'scrolled' || hostWidth === 0
+			? 1
+			: Math.min(settings.maxColumnCount, Math.ceil(hostWidth / settings.maxInlineSize))
+	);
+
+	/**
+	 * Where to resume from.
+	 *
+	 * foliate's resolveNavigation logs and swallows its failures, returning
+	 * undefined — and init() then falls through to next(), i.e. page one. So an
+	 * epub.js-authored CFI that does not resolve would silently reset the reader.
+	 * Resolve it up front and fall back to the percentage the backend has stored
+	 * all along, which lands within a page or two.
+	 */
+	function resolveStart(v: View): FoliateTarget | null {
+		if (initialCfi) {
+			const resolved = v.resolveNavigation(initialCfi);
+			if (resolved && Number.isInteger(resolved.index) && resolved.index >= 0) return initialCfi;
+			console.warn('Stored CFI did not resolve; falling back to percentage', initialCfi);
+		}
+		return initialFraction > 0 ? { fraction: initialFraction } : null;
+	}
+
+	function applyAttributes(current: ReaderSettings) {
+		if (!view?.renderer) return;
+		for (const [name, value] of Object.entries(toRendererAttributes(current))) {
+			view.renderer.setAttribute(name, value);
+		}
+	}
+
+	function applyStyles(current: ReaderSettings, dark: boolean) {
+		// Absent on the fixed-layout renderer, which has no reflowable text.
+		view?.renderer?.setStyles?.(buildReaderStyles(current, readReaderPalette(dark)));
+	}
+
+	/**
+	 * Sections render inside iframes, which swallow key events before they reach
+	 * the document — so arrow keys only worked with focus in the app chrome. The
+	 * old reader used epub.js's rendition.on('keydown'); foliate has no equivalent,
+	 * so bind on the section document each time one loads.
+	 */
+	function handleLoad(event: Event) {
+		const { doc } = (event as CustomEvent<FoliateLoadDetail>).detail;
+		doc.addEventListener('keydown', onKeydown);
+	}
+
+	/**
+	 * Replays a shortcut pressed inside the book on the host window.
+	 *
+	 * Key events do not cross document boundaries, so a shortcut pressed while
+	 * the book has focus never reaches the app's handlers — which are bound on
+	 * the window, as the sidebar's ctrl+B is — and the browser's own default runs
+	 * instead. That is why ctrl+B opened bookmarks in the book but toggled the
+	 * chapter sidebar everywhere else.
+	 *
+	 * The original is only cancelled if an app handler actually claimed the
+	 * replay, so combinations the app does not use — ctrl+C above all — keep
+	 * their normal browser behaviour.
+	 */
+	function forwardShortcut(event: KeyboardEvent) {
+		const source = (event.target as Node | null)?.ownerDocument;
+		// isTrusted rules out the replay itself, which would otherwise recurse.
+		if (!event.isTrusted || !source || source === document) return;
+
+		const claimed = !window.dispatchEvent(
+			new KeyboardEvent('keydown', {
+				key: event.key,
+				code: event.code,
+				ctrlKey: event.ctrlKey,
+				metaKey: event.metaKey,
+				shiftKey: event.shiftKey,
+				altKey: event.altKey,
+				bubbles: true,
+				cancelable: true
+			})
+		);
+
+		if (claimed) event.preventDefault();
+	}
+
+	function onKeydown(event: KeyboardEvent) {
+		if (event.ctrlKey || event.metaKey || event.altKey) {
+			forwardShortcut(event);
+			return;
+		}
+
+		switch (event.key) {
+			case 'ArrowLeft':
+			case 'PageUp':
+				event.preventDefault();
+				void goLeft();
+				break;
+			case 'ArrowRight':
+			case 'PageDown':
+			case ' ':
+				event.preventDefault();
+				void goRight();
+				break;
+		}
+	}
+
+	export function goLeft() {
+		return view?.goLeft();
+	}
+
+	export function goRight() {
+		return view?.goRight();
+	}
+
+	/** Direction-aware, so the chevrons stay literal on right-to-left books. */
+	export function goForward() {
+		return rtl ? view?.goLeft() : view?.goRight();
+	}
+
+	export function goBack() {
+		return rtl ? view?.goRight() : view?.goLeft();
+	}
+
+	export function goTo(target: FoliateTarget) {
+		return view?.goTo(target);
+	}
+
+	onMount(() => {
+		let disposed = false;
+
+		const resizeObserver = new ResizeObserver(([entry]) => {
+			hostWidth = entry.contentRect.width;
+		});
+		if (container) resizeObserver.observe(container);
+
+		(async () => {
+			try {
+				await loadFoliate();
+				if (disposed || !container) return;
+
+				view = createFoliateView();
+
+				// Appended before open(): the paginator measures its host, and a
+				// detached element has no size to measure.
+				//
+				// Deliberately outside Svelte's control. <foliate-view> is a custom
+				// element that renders its own iframes; writing it in markup would
+				// have SSR emit an unknown tag and leave Svelte trying to hydrate a
+				// subtree the paginator owns.
+				// eslint-disable-next-line svelte/no-dom-manipulating
+				container.append(view);
+
+				await view.open(file);
+				if (disposed) return;
+
+				rtl = view.book?.dir === 'rtl';
+
+				// Both before init(): the paginator stores the styles and re-applies
+				// them on every section load, so the first page paints already themed
+				// rather than flashing the book's own colours first.
+				applyAttributes(settings);
+				applyStyles(settings, mode.current === 'dark');
+
+				view.addEventListener('load', handleLoad);
+				view.addEventListener('relocate', (event) => {
+					onrelocate?.((event as CustomEvent<FoliateRelocateDetail>).detail);
+				});
+
+				await view.init({ lastLocation: resolveStart(view) });
+				if (disposed) return;
+
+				ready = true;
+				onready?.({
+					toc: view.book?.toc ?? [],
+					isFixedLayout: view.isFixedLayout,
+					rtl
+				});
+			} catch (error) {
+				console.error('Could not open the book', error);
+				onerror?.(error instanceof Error ? error.message : 'The file could not be opened.');
+			}
+		})();
+
+		return () => {
+			disposed = true;
+			resizeObserver.disconnect();
+			for (const { doc } of view?.renderer?.getContents?.() ?? []) {
+				doc.removeEventListener('keydown', onKeydown);
+			}
+			view?.close();
+			view?.remove();
+			view = undefined;
+		};
+	});
+
+	// Re-applies on any settings change and on a light/dark flip.
+	//
+	// setStyles re-paginates internally and waits on document.fonts.ready, so no
+	// manual resize is needed — unlike epub.js, which required a window resize
+	// listener and rendition.resize().
+	//
+	// The styles are deferred by a frame on purpose: mode.current flips before
+	// ModeWatcher writes .dark onto <html>, so reading the tokens in this tick
+	// would style the book from the outgoing palette. That is the bug the old
+	// reader shipped with — a white page against a dark UI.
+	$effect(() => {
+		const current = settings;
+		const dark = mode.current === 'dark';
+		if (!ready) return;
+
+		applyAttributes(current);
+
+		const frame = requestAnimationFrame(() => applyStyles(current, dark));
+		return () => cancelAnimationFrame(frame);
+	});
+</script>
+
+<!--
+	Not {onkeydown}: that shorthand resolves to the global window.onkeydown
+	property, which TypeScript accepts and which is null at runtime, so the
+	handler silently never ran and arrows only worked once the book had focus.
+-->
+<svelte:window onkeydown={onKeydown} />
+
+<div class="relative h-full w-full">
+	<div bind:this={container} class="foliate-host"></div>
+
+	{#if ready && columns === 2}
+		<!-- The gutter between the two pages of a spread. Sits in the column gap,
+		     so it never crosses text. -->
+		<div
+			aria-hidden="true"
+			class="pointer-events-none absolute inset-y-[6%] left-1/2 w-px -translate-x-1/2 bg-border"
+		></div>
+	{/if}
+</div>
+
+<style>
+	/* foliate-view extends bare HTMLElement, so it has no default display. */
+	.foliate-host {
+		height: 100%;
+		width: 100%;
+	}
+
+	.foliate-host :global(foliate-view) {
+		display: block;
+		height: 100%;
+		width: 100%;
+	}
+</style>

frontend/src/lib/components/reader/reader-settings.svelte

@@ -0,0 +1,265 @@
+<script lang="ts">
+	import { RotateCcw } from '@lucide/svelte';
+
+	import type { ReaderFlow, ReaderSettings } from '$lib/schema/reader';
+	import { READER_BOUNDS, READER_FONT_OPTIONS } from '$lib/reader/settings';
+	import { getReaderSettingsState } from '$lib/state/reader-settings.svelte';
+
+	import { Button } from '$lib/components/ui/button/index.js';
+	import { Label } from '$lib/components/ui/label/index.js';
+	import { NativeSelect } from '$lib/components/ui/native-select/index.js';
+	import { Separator } from '$lib/components/ui/separator/index.js';
+	import { Slider } from '$lib/components/ui/slider/index.js';
+	import { Switch } from '$lib/components/ui/switch/index.js';
+	import * as Sheet from '$lib/components/ui/sheet/index.js';
+	import * as ToggleGroup from '$lib/components/ui/toggle-group/index.js';
+
+	let {
+		open = $bindable(false),
+		/**
+		 * Fixed-layout books render through foliate-fxl, which has no setStyles and
+		 * ignores the paginator's attributes — so the controls below would do
+		 * nothing. Hide them rather than leave dead sliders on screen.
+		 */
+		isFixedLayout = false
+	}: {
+		open?: boolean;
+		isFixedLayout?: boolean;
+	} = $props();
+
+	const state = getReaderSettingsState();
+
+	/** Slider binds an array; unwrap to the single value the setting holds. */
+	function slide<K extends keyof ReaderSettings>(key: K) {
+		return (value: number) => state.set(key, value as ReaderSettings[K]);
+	}
+</script>
+
+<Sheet.Root bind:open>
+	<Sheet.Content side="right" class="w-full gap-0 overflow-y-auto sm:max-w-sm">
+		<Sheet.Header>
+			<Sheet.Title class="font-serif">Reading</Sheet.Title>
+			<Sheet.Description>
+				{isFixedLayout
+					? 'This book has a fixed layout, so its typography is set by the publisher.'
+					: 'Applies to every book you read.'}
+			</Sheet.Description>
+		</Sheet.Header>
+
+		<div class="flex flex-col gap-6 px-4 pb-6">
+			{#if !isFixedLayout}
+				<section class="flex flex-col gap-4">
+					<h3 class="font-mono text-[10px] tracking-widest text-muted-foreground uppercase">
+						Typography
+					</h3>
+
+					<div class="flex flex-col gap-2">
+						<Label for="reader-font">Typeface</Label>
+						<NativeSelect
+							id="reader-font"
+							class="w-full"
+							value={state.settings.fontFamily}
+							onchange={(event) => state.set('fontFamily', event.currentTarget.value)}
+						>
+							{#each READER_FONT_OPTIONS as font (font.value)}
+								<option value={font.value}>{font.label}</option>
+							{/each}
+						</NativeSelect>
+					</div>
+
+					<div class="flex flex-col gap-2">
+						<div class="flex items-center justify-between">
+							<Label for="reader-size">Size</Label>
+							<span class="font-mono text-xs text-muted-foreground tabular-nums">
+								{state.settings.fontSize}px
+							</span>
+						</div>
+						<Slider
+							id="reader-size"
+							type="single"
+							value={state.settings.fontSize}
+							min={READER_BOUNDS.fontSize.min}
+							max={READER_BOUNDS.fontSize.max}
+							step={READER_BOUNDS.fontSize.step}
+							onValueChange={slide('fontSize')}
+						/>
+					</div>
+
+					<div class="flex flex-col gap-2">
+						<div class="flex items-center justify-between">
+							<Label for="reader-weight">Weight</Label>
+							<span class="font-mono text-xs text-muted-foreground tabular-nums">
+								{state.settings.fontWeight}
+							</span>
+						</div>
+						<Slider
+							id="reader-weight"
+							type="single"
+							value={state.settings.fontWeight}
+							min={READER_BOUNDS.fontWeight.min}
+							max={READER_BOUNDS.fontWeight.max}
+							step={READER_BOUNDS.fontWeight.step}
+							onValueChange={slide('fontWeight')}
+						/>
+					</div>
+
+					<div class="flex flex-col gap-2">
+						<div class="flex items-center justify-between">
+							<Label for="reader-leading">Line height</Label>
+							<span class="font-mono text-xs text-muted-foreground tabular-nums">
+								{state.settings.lineHeight.toFixed(2)}
+							</span>
+						</div>
+						<Slider
+							id="reader-leading"
+							type="single"
+							value={state.settings.lineHeight}
+							min={READER_BOUNDS.lineHeight.min}
+							max={READER_BOUNDS.lineHeight.max}
+							step={READER_BOUNDS.lineHeight.step}
+							onValueChange={slide('lineHeight')}
+						/>
+					</div>
+
+					<div class="flex flex-col gap-2">
+						<div class="flex items-center justify-between">
+							<Label for="reader-tracking">Letter spacing</Label>
+							<span class="font-mono text-xs text-muted-foreground tabular-nums">
+								{state.settings.letterSpacing.toFixed(2)}em
+							</span>
+						</div>
+						<Slider
+							id="reader-tracking"
+							type="single"
+							value={state.settings.letterSpacing}
+							min={READER_BOUNDS.letterSpacing.min}
+							max={READER_BOUNDS.letterSpacing.max}
+							step={READER_BOUNDS.letterSpacing.step}
+							onValueChange={slide('letterSpacing')}
+						/>
+					</div>
+
+					<div class="flex items-center justify-between gap-4">
+						<Label for="reader-justify" class="font-normal">Justify text</Label>
+						<Switch
+							id="reader-justify"
+							checked={state.settings.justify}
+							onCheckedChange={(checked) => state.set('justify', checked)}
+						/>
+					</div>
+
+					<div class="flex items-center justify-between gap-4">
+						<Label for="reader-hyphenate" class="font-normal">Hyphenate</Label>
+						<Switch
+							id="reader-hyphenate"
+							checked={state.settings.hyphenate}
+							onCheckedChange={(checked) => state.set('hyphenate', checked)}
+						/>
+					</div>
+				</section>
+
+				<Separator />
+			{/if}
+
+			<section class="flex flex-col gap-4">
+				<h3 class="font-mono text-[10px] tracking-widest text-muted-foreground uppercase">
+					Layout
+				</h3>
+
+				<div class="flex flex-col gap-2">
+					<Label>Flow</Label>
+					<ToggleGroup.Root
+						type="single"
+						variant="outline"
+						value={state.settings.flow}
+						onValueChange={(value) => value && state.set('flow', value as ReaderFlow)}
+						class="justify-start"
+					>
+						<ToggleGroup.Item value="paginated">Pages</ToggleGroup.Item>
+						<ToggleGroup.Item value="scrolled">Scroll</ToggleGroup.Item>
+					</ToggleGroup.Root>
+				</div>
+
+				{#if !isFixedLayout}
+					<div class="flex flex-col gap-2">
+						<Label>Columns</Label>
+						<ToggleGroup.Root
+							type="single"
+							variant="outline"
+							value={String(state.settings.maxColumnCount)}
+							onValueChange={(value) => value && state.set('maxColumnCount', Number(value))}
+							class="justify-start"
+						>
+							<ToggleGroup.Item value="1">Single</ToggleGroup.Item>
+							<ToggleGroup.Item value="2">Spread</ToggleGroup.Item>
+						</ToggleGroup.Root>
+						<p class="text-xs text-muted-foreground">
+							A spread needs the window to be wide enough for two columns.
+						</p>
+					</div>
+
+					<div class="flex flex-col gap-2">
+						<div class="flex items-center justify-between">
+							<Label for="reader-width">Line width</Label>
+							<span class="font-mono text-xs text-muted-foreground tabular-nums">
+								{state.settings.maxInlineSize}px
+							</span>
+						</div>
+						<Slider
+							id="reader-width"
+							type="single"
+							value={state.settings.maxInlineSize}
+							min={READER_BOUNDS.maxInlineSize.min}
+							max={READER_BOUNDS.maxInlineSize.max}
+							step={READER_BOUNDS.maxInlineSize.step}
+							onValueChange={slide('maxInlineSize')}
+						/>
+					</div>
+
+					<div class="flex flex-col gap-2">
+						<div class="flex items-center justify-between">
+							<Label for="reader-gap">Column gap</Label>
+							<span class="font-mono text-xs text-muted-foreground tabular-nums">
+								{state.settings.gap}%
+							</span>
+						</div>
+						<Slider
+							id="reader-gap"
+							type="single"
+							value={state.settings.gap}
+							min={READER_BOUNDS.gap.min}
+							max={READER_BOUNDS.gap.max}
+							step={READER_BOUNDS.gap.step}
+							onValueChange={slide('gap')}
+						/>
+					</div>
+				{/if}
+
+				<div class="flex flex-col gap-2">
+					<div class="flex items-center justify-between">
+						<Label for="reader-margin">Margins</Label>
+						<span class="font-mono text-xs text-muted-foreground tabular-nums">
+							{state.settings.margin}px
+						</span>
+					</div>
+					<Slider
+						id="reader-margin"
+						type="single"
+						value={state.settings.margin}
+						min={READER_BOUNDS.margin.min}
+						max={READER_BOUNDS.margin.max}
+						step={READER_BOUNDS.margin.step}
+						onValueChange={slide('margin')}
+					/>
+				</div>
+			</section>
+
+			<Separator />
+
+			<Button variant="outline" onclick={() => state.reset()} class="self-start">
+				<RotateCcw class="size-4" />
+				Reset to defaults
+			</Button>
+		</div>
+	</Sheet.Content>
+</Sheet.Root>

frontend/src/lib/components/reader/reader-toc.svelte

@@ -0,0 +1,86 @@
+<script lang="ts">
+	import { ChevronDown } from '@lucide/svelte';
+
+	import type { FoliateTocItem } from '$foliate/view.js';
+	import * as Collapsible from '$lib/components/ui/collapsible/index';
+	import * as Sidebar from '$lib/components/ui/sidebar/index';
+
+	let {
+		toc,
+		activeId = null,
+		onnavigate
+	}: {
+		toc: FoliateTocItem[];
+		/** From relocate.tocItem.id — foliate assigns these, books do not carry them. */
+		activeId?: number | null;
+		onnavigate: (href: string) => void;
+	} = $props();
+
+	/** True when this item or any descendant is the current chapter. */
+	function contains(item: FoliateTocItem, id: number | null): boolean {
+		if (id === null) return false;
+		if (item.id === id) return true;
+		return item.subitems?.some((sub) => contains(sub, id)) ?? false;
+	}
+</script>
+
+<Sidebar.Root>
+	<Sidebar.Header class="px-4 py-3">
+		<p class="font-mono text-[10px] tracking-widest text-muted-foreground uppercase">Chapters</p>
+	</Sidebar.Header>
+	<Sidebar.Content>
+		<Sidebar.GroupContent>
+			<Sidebar.Menu>
+				{#each toc as chapter (chapter.id)}
+					{#if chapter.subitems?.length}
+						<Collapsible.Root class="group/collapsible" open={contains(chapter, activeId)}>
+							<div class="flex w-full items-center gap-1">
+								<Sidebar.MenuItem class="min-w-0 flex-1">
+									<Sidebar.MenuButton
+										class="w-full"
+										isActive={chapter.id === activeId}
+										onclick={() => onnavigate(chapter.href)}
+									>
+										<span class="block truncate" title={chapter.label}>{chapter.label}</span>
+									</Sidebar.MenuButton>
+								</Sidebar.MenuItem>
+								<Collapsible.Trigger class="flex-shrink-0 p-2">
+									<ChevronDown
+										class="size-4 transition-transform group-data-[state=open]/collapsible:rotate-180"
+									/>
+								</Collapsible.Trigger>
+							</div>
+							<Collapsible.Content>
+								<Sidebar.MenuSub>
+									{#each chapter.subitems ?? [] as subchapter (subchapter.id)}
+										<Sidebar.MenuSubItem class="min-w-0">
+											<Sidebar.MenuButton
+												class="w-full"
+												isActive={subchapter.id === activeId}
+												onclick={() => onnavigate(subchapter.href)}
+											>
+												<span class="block truncate" title={subchapter.label}>
+													{subchapter.label}
+												</span>
+											</Sidebar.MenuButton>
+										</Sidebar.MenuSubItem>
+									{/each}
+								</Sidebar.MenuSub>
+							</Collapsible.Content>
+						</Collapsible.Root>
+					{:else}
+						<Sidebar.MenuItem class="min-w-0">
+							<Sidebar.MenuButton
+								class="w-full"
+								isActive={chapter.id === activeId}
+								onclick={() => onnavigate(chapter.href)}
+							>
+								<span class="block truncate" title={chapter.label}>{chapter.label}</span>
+							</Sidebar.MenuButton>
+						</Sidebar.MenuItem>
+					{/if}
+				{/each}
+			</Sidebar.Menu>
+		</Sidebar.GroupContent>
+	</Sidebar.Content>
+</Sidebar.Root>

frontend/src/lib/components/ui/avatar/avatar-badge.svelte

@@ -0,0 +1,26 @@
+<script lang="ts">
+	import { cn, type WithElementRef } from "$lib/utils.js";
+	import type { HTMLAttributes } from "svelte/elements";
+
+	let {
+		ref = $bindable(null),
+		class: className,
+		children,
+		...restProps
+	}: WithElementRef<HTMLAttributes<HTMLSpanElement>> = $props();
+</script>
+
+<span
+	bind:this={ref}
+	data-slot="avatar-badge"
+	class={cn(
+		"bg-primary text-primary-foreground ring-background absolute right-0 bottom-0 z-10 inline-flex items-center justify-center rounded-full bg-blend-color ring-2 select-none",
+		"group-data-[size=sm]/avatar:size-2 group-data-[size=sm]/avatar:[&>svg]:hidden",
+		"group-data-[size=default]/avatar:size-2.5 group-data-[size=default]/avatar:[&>svg]:size-2",
+		"group-data-[size=lg]/avatar:size-3 group-data-[size=lg]/avatar:[&>svg]:size-2",
+		className
+	)}
+	{...restProps}
+>
+	{@render children?.()}
+</span>

frontend/src/lib/components/ui/avatar/avatar-fallback.svelte

@@ -0,0 +1,20 @@
+<script lang="ts">
+	import { Avatar as AvatarPrimitive } from "bits-ui";
+	import { cn } from "$lib/utils.js";
+
+	let {
+		ref = $bindable(null),
+		class: className,
+		...restProps
+	}: AvatarPrimitive.FallbackProps = $props();
+</script>
+
+<AvatarPrimitive.Fallback
+	bind:ref
+	data-slot="avatar-fallback"
+	class={cn(
+		"rounded-full bg-muted text-muted-foreground flex size-full items-center justify-center text-sm group-data-[size=sm]/avatar:text-xs",
+		className
+	)}
+	{...restProps}
+/>

frontend/src/lib/components/ui/avatar/avatar-group-count.svelte

@@ -0,0 +1,23 @@
+<script lang="ts">
+	import { cn, type WithElementRef } from "$lib/utils.js";
+	import type { HTMLAttributes } from "svelte/elements";
+
+	let {
+		ref = $bindable(null),
+		class: className,
+		children,
+		...restProps
+	}: WithElementRef<HTMLAttributes<HTMLDivElement>> = $props();
+</script>
+
+<div
+	bind:this={ref}
+	data-slot="avatar-group-count"
+	class={cn(
+		"size-8 rounded-full bg-muted text-sm text-muted-foreground group-has-data-[size=lg]/avatar-group:size-10 group-has-data-[size=sm]/avatar-group:size-6 [&>svg]:size-4 group-has-data-[size=lg]/avatar-group:[&>svg]:size-5 group-has-data-[size=sm]/avatar-group:[&>svg]:size-3 relative flex shrink-0 items-center justify-center ring-2 ring-background",
+		className
+	)}
+	{...restProps}
+>
+	{@render children?.()}
+</div>

frontend/src/lib/components/ui/avatar/avatar-group.svelte

@@ -0,0 +1,23 @@
+<script lang="ts">
+	import { cn, type WithElementRef } from "$lib/utils.js";
+	import type { HTMLAttributes } from "svelte/elements";
+
+	let {
+		ref = $bindable(null),
+		class: className,
+		children,
+		...restProps
+	}: WithElementRef<HTMLAttributes<HTMLDivElement>> = $props();
+</script>
+
+<div
+	bind:this={ref}
+	data-slot="avatar-group"
+	class={cn(
+		"cn-avatar-group group/avatar-group flex -space-x-2 *:data-[slot=avatar]:ring-2 *:data-[slot=avatar]:ring-background",
+		className
+	)}
+	{...restProps}
+>
+	{@render children?.()}
+</div>

frontend/src/lib/components/ui/avatar/avatar-image.svelte

@@ -0,0 +1,17 @@
+<script lang="ts">
+	import { Avatar as AvatarPrimitive } from "bits-ui";
+	import { cn } from "$lib/utils.js";
+
+	let {
+		ref = $bindable(null),
+		class: className,
+		...restProps
+	}: AvatarPrimitive.ImageProps = $props();
+</script>
+
+<AvatarPrimitive.Image
+	bind:ref
+	data-slot="avatar-image"
+	class={cn("rounded-full aspect-square size-full object-cover", className)}
+	{...restProps}
+/>

frontend/src/lib/components/ui/avatar/avatar.svelte

@@ -0,0 +1,26 @@
+<script lang="ts">
+	import { Avatar as AvatarPrimitive } from "bits-ui";
+	import { cn } from "$lib/utils.js";
+
+	let {
+		ref = $bindable(null),
+		loadingStatus = $bindable("loading"),
+		size = "default",
+		class: className,
+		...restProps
+	}: AvatarPrimitive.RootProps & {
+		size?: "default" | "sm" | "lg";
+	} = $props();
+</script>
+
+<AvatarPrimitive.Root
+	bind:ref
+	bind:loadingStatus
+	data-slot="avatar"
+	data-size={size}
+	class={cn(
+		"size-8 rounded-full after:rounded-full data-[size=lg]:size-10 data-[size=sm]:size-6 group/avatar relative flex shrink-0 select-none after:absolute after:inset-0 after:border after:border-border after:mix-blend-darken dark:after:mix-blend-lighten",
+		className
+	)}
+	{...restProps}
+/>

frontend/src/lib/components/ui/avatar/index.ts

@@ -0,0 +1,22 @@
+import Badge from "./avatar-badge.svelte";
+import Fallback from "./avatar-fallback.svelte";
+import GroupCount from "./avatar-group-count.svelte";
+import Group from "./avatar-group.svelte";
+import Image from "./avatar-image.svelte";
+import Root from "./avatar.svelte";
+
+export {
+	Root,
+	Image,
+	Fallback,
+	Badge,
+	Group,
+	GroupCount,
+	//
+	Root as Avatar,
+	Image as AvatarImage,
+	Fallback as AvatarFallback,
+	Badge as AvatarBadge,
+	Group as AvatarGroup,
+	GroupCount as AvatarGroupCount,
+};

frontend/src/lib/components/ui/button/button.svelte

@@ -15,7 +15,8 @@
 				secondary: 'bg-secondary text-secondary-foreground shadow-xs hover:bg-secondary/80',
 				ghost: 'hover:bg-accent hover:text-accent-foreground dark:hover:bg-accent/50',
 				link: 'text-primary underline-offset-4 hover:underline',
-				accent: 'bg-green-500 hover:bg-green-500/90 dark:text-primary dark:bg-green-500/90'
+				accent:
+					'bg-primary text-primary-foreground shadow-xs hover:bg-primary/90'
 			},
 			size: {
 				default: 'h-9 px-4 py-2 has-[>svg]:px-3',

frontend/src/lib/components/ui/file-drop-zone/file-drop-zone.svelte

@@ -3,6 +3,7 @@
 -->
 
 <script lang="ts">
+	import { untrack } from 'svelte';
 	import { cn } from '$lib/utils';
 	import UploadIcon from '@lucide/svelte/icons/upload';
 	import { displaySize } from '.';
@@ -26,11 +27,14 @@
 		...rest
 	}: FileDropZoneProps = $props();
 
+	// A one-off sanity check at init, not a reactive concern.
+	untrack(() => {
 		if (maxFiles !== undefined && fileCount === undefined) {
 			console.warn(
 				'Make sure to provide FileDropZone with `fileCount` when using the `maxFiles` prompt'
 			);
 		}
+	});
 
 	let uploading = $state(false);
 

frontend/src/lib/components/ui/sidebar/sidebar-provider.svelte

@@ -9,6 +9,7 @@
 		SIDEBAR_WIDTH_ICON
 	} from './constants.js';
 	import { setSidebar } from './context.svelte.js';
+	import { untrack } from 'svelte';
 
 	let {
 		ref = $bindable(null),
@@ -36,7 +37,7 @@
 			// This sets the cookie to keep the sidebar state.
 			document.cookie = `${SIDEBAR_COOKIE_NAME}_${contextKey}=${open}; path=/; max-age=${SIDEBAR_COOKIE_MAX_AGE}`;
 		},
-		contextKey: contextKey
+		contextKey: untrack(() => contextKey)
 	});
 </script>
 

frontend/src/lib/components/ui/sidebar/sidebar-rail.svelte

@@ -2,6 +2,7 @@
 	import { cn, type WithElementRef } from '$lib/utils.js';
 	import type { HTMLAttributes } from 'svelte/elements';
 	import { useSidebar } from './context.svelte.js';
+	import { untrack } from 'svelte';
 
 	let {
 		ref = $bindable(null),
@@ -13,7 +14,8 @@
 		contextKey?: string;
 	} = $props();
 
-	const sidebar = useSidebar(contextKey);
+	// contextKey selects which sidebar this rail drives; fixed per instance.
+	const sidebar = useSidebar(untrack(() => contextKey));
 </script>
 
 <button

frontend/src/lib/components/ui/slider/index.ts

@@ -0,0 +1,7 @@
+import Root from './slider.svelte';
+
+export {
+	Root,
+	//
+	Root as Slider
+};

frontend/src/lib/components/ui/slider/slider.svelte

@@ -0,0 +1,63 @@
+<script lang="ts">
+	import { Slider as SliderPrimitive } from 'bits-ui';
+	import { cn, type WithoutChildrenOrChild } from '$lib/utils.js';
+
+	let {
+		ref = $bindable(null),
+		value = $bindable(),
+		orientation = 'horizontal',
+		class: className,
+		...restProps
+	}: WithoutChildrenOrChild<SliderPrimitive.RootProps> = $props();
+</script>
+
+<!--
+Discriminated Unions + Destructing (required for bindable) do not
+get along, so we shut typescript up by casting `value` to `never`.
+
+The generated component styled the track with `data-horizontal:` /
+`data-vertical:` variants, which Tailwind compiles to `[data-horizontal]` —
+an attribute nothing sets, since the orientation is carried as
+`data-orientation="horizontal"`. The track therefore had no height and only
+the thumb was visible. These use the same `data-[orientation=…]` form as
+separator.svelte, which is the convention everywhere else in ui/.
+-->
+<SliderPrimitive.Root
+	bind:ref
+	bind:value={value as never}
+	data-slot="slider"
+	{orientation}
+	class={cn(
+		'relative flex w-full touch-none items-center select-none data-disabled:opacity-50 data-[orientation=vertical]:h-full data-[orientation=vertical]:min-h-40 data-[orientation=vertical]:w-auto data-[orientation=vertical]:flex-col',
+		className
+	)}
+	{...restProps}
+>
+	{#snippet children({ thumbItems })}
+		<span
+			data-slot="slider-track"
+			data-orientation={orientation}
+			class={cn(
+				'relative grow overflow-hidden rounded-full bg-muted',
+				'data-[orientation=horizontal]:h-1.5 data-[orientation=horizontal]:w-full',
+				'data-[orientation=vertical]:h-full data-[orientation=vertical]:w-1.5'
+			)}
+		>
+			<SliderPrimitive.Range
+				data-slot="slider-range"
+				data-orientation={orientation}
+				class={cn(
+					'absolute bg-primary select-none',
+					'data-[orientation=horizontal]:h-full data-[orientation=vertical]:w-full'
+				)}
+			/>
+		</span>
+		{#each thumbItems as thumb (thumb.index)}
+			<SliderPrimitive.Thumb
+				data-slot="slider-thumb"
+				index={thumb.index}
+				class="relative block size-4 shrink-0 rounded-full border border-primary bg-background ring-ring/50 transition-[color,box-shadow] select-none after:absolute after:-inset-2 hover:ring-3 focus-visible:ring-3 focus-visible:outline-hidden active:ring-3 disabled:pointer-events-none disabled:opacity-50"
+			/>
+		{/each}
+	{/snippet}
+</SliderPrimitive.Root>

frontend/src/lib/components/ui/tags-input/tags-input.svelte

@@ -206,6 +206,6 @@
 		{placeholder}
 		data-invalid={invalid}
 		onkeydown={keydown}
-		class="min-w-16 shrink grow basis-0 border-none bg-transparent px-2 outline-hidden placeholder:text-muted-foreground focus:outline-hidden disabled:cursor-not-allowed data-[invalid=true]:text-red-500 md:text-sm"
+		class="min-w-16 shrink grow basis-0 border-none bg-transparent px-2 outline-hidden placeholder:text-muted-foreground focus:outline-hidden disabled:cursor-not-allowed data-[invalid=true]:text-destructive md:text-sm"
 	/>
 </div>

frontend/src/lib/components/ui/toggle-group/toggle-group.svelte

@@ -23,9 +23,16 @@
 		...restProps
 	}: ToggleGroupPrimitive.RootProps & ToggleVariants = $props();
 
+	// Getters rather than values: consumers read `ctx.variant` / `ctx.size` as
+	// plain properties, so this stays source-compatible while making the props
+	// actually reactive — passing them by value captured the initial ones.
 	setToggleGroupCtx({
-		variant,
-		size
+		get variant() {
+			return variant;
+		},
+		get size() {
+			return size;
+		}
 	});
 </script>
 

frontend/src/lib/components/view/batch-operations-toolbar.svelte

@@ -1,15 +1,27 @@
 <script lang="ts">
 	import { buttonVariants } from '$lib/components/ui/button/button.svelte';
 	import * as Tooltip from '$lib/components/ui/tooltip/index';
-	import { BookOpenCheck, Download, Trash2, SquareCheckBig, X, Album, PlusIcon } from '@lucide/svelte';
+	import {
+		BookOpenCheck,
+		Download,
+		GitMerge,
+		Trash2,
+		SquareCheckBig,
+		SlidersHorizontal,
+		X,
+		Album,
+		PlusIcon
+	} from '@lucide/svelte';
 	import * as DropdownMenu from '$lib/components/ui/dropdown-menu/index';
 	import { getBookOperationsState } from '$lib/state/bookOperations.svelte';
 	import { getBookSelectionState } from '$lib/state/bookSelection.svelte';
 	import { getBookCollectionState } from '$lib/state/bookCollection.svelte';
 	import { getLibraryState } from '$lib/state/library.svelte';
 	import { getBookshelfState } from '$lib/state/bookshelf.svelte';
-	import { invalidate } from '$app/navigation';
 	import ShelfCreateDialog from '../forms/shelf-create-dialog.svelte';
+	import MergeBooks from '../forms/merge-books/merge-books.svelte';
+	import { mergeInto } from '../forms/merge-books/merge';
+	import type { Book } from '$lib/schema';
 
 	const libraryState = getLibraryState();
 	const bookshelfState = getBookshelfState();
@@ -17,10 +29,50 @@
 	const bookOps = getBookOperationsState();
 	const collectionState = getBookCollectionState();
 
-	let selectedBooks = $derived(selectionState.getSelectedBooks())
+	let selectedBooks = $derived(selectionState.getSelectedBooks());
 
-	let createShelfDialogOpen = $state(false)
+	let createShelfDialogOpen = $state(false);
 
+	// The books the merge dialog opened on. Snapshotted rather than read live from
+	// the selection, so clearing the selection on success cannot empty the dialog
+	// underneath itself.
+	let mergingBooks = $state<Book[] | null>(null);
+
+	// A quick merge has no dialog to disable, so the flag is what stops a slow
+	// round trip being started twice.
+	let quickMerging = $state(false);
+
+	/** Put the library back in step with a merge that went through. */
+	function afterMerge(folded: Book[]) {
+		// Only the folded records are gone; the survivor is still in the library.
+		libraryState.activeLibrary!.total! -= folded.length;
+		bookshelfState.deletedBooks(folded);
+		selectionState.deselectAll();
+	}
+
+	/**
+	 * Merge without opening the workbench, keeping the book selected first.
+	 *
+	 * Selection is held in insertion order, so the first record is the one the
+	 * reader started from — the one they were looking at when they decided the
+	 * rest were copies of it.
+	 */
+	async function quickMerge() {
+		if (quickMerging) return;
+
+		// Read once: the merge clears the selection, and the bookkeeping afterwards
+		// still needs the records that went away.
+		const [survivor, ...folded] = selectedBooks;
+		if (!survivor || folded.length === 0) return;
+
+		quickMerging = true;
+
+		const merged = await mergeInto(libraryState.activeLibrary!.id, survivor, folded);
+
+		quickMerging = false;
+
+		if (merged) afterMerge(folded);
+	}
 </script>
 
 <!-- Mark selected as finished button -->
@@ -59,8 +111,7 @@
 					<DropdownMenu.Content>
 						<DropdownMenu.Group>
 							<DropdownMenu.GroupHeading>Shelves</DropdownMenu.GroupHeading>
-							<DropdownMenu.Separator />
-							{#each bookshelfState.getBookshelves(libraryState.activeLibrary!.id) as shelf}
+							{#each bookshelfState.getBookshelves(libraryState.activeLibrary!.id) ?? [] as shelf (shelf.id)}
 								<DropdownMenu.CheckboxItem
 									checked={selectedBooks.every(
 										(book) => book.lists.findIndex((sh) => sh.id === shelf.id) !== -1
@@ -97,8 +148,9 @@
 						</DropdownMenu.Group>
 						<DropdownMenu.Separator />
 						<DropdownMenu.Item
-							onclick={() => createShelfDialogOpen = true}
-							class="text-muted-foreground ">
+							onclick={() => (createShelfDialogOpen = true)}
+							class="text-muted-foreground "
+						>
 							<PlusIcon class="size-4" />
 							New Shelf
 						</DropdownMenu.Item>
@@ -129,6 +181,44 @@
 	</Tooltip.Root>
 </Tooltip.Provider>
 
+<!-- Merge selected button. Two books is the smallest thing a merge can mean, so
+     it appears only once there are two. -->
+{#if selectedBooks.length > 1}
+	<Tooltip.Provider>
+		<Tooltip.Root ignoreNonKeyboardFocus>
+			<Tooltip.Trigger>
+				{#snippet child({ props })}
+					<!-- A choice rather than a jump straight into the workbench: two
+					     copies of one book have nothing to resolve, and the dialog is a
+					     step in the way of saying so. -->
+					<DropdownMenu.Root>
+						<DropdownMenu.Trigger
+							{...props}
+							disabled={quickMerging}
+							class={buttonVariants({ variant: 'ghost', size: 'icon' })}
+						>
+							<GitMerge />
+						</DropdownMenu.Trigger>
+						<DropdownMenu.Content>
+							<DropdownMenu.Item onclick={quickMerge}>
+								<GitMerge class="size-4 shrink-0" />
+								Merge into first selected
+							</DropdownMenu.Item>
+							<DropdownMenu.Item onclick={() => (mergingBooks = selectedBooks)}>
+								<SlidersHorizontal class="size-4 shrink-0" />
+								Manual merge…
+							</DropdownMenu.Item>
+						</DropdownMenu.Content>
+					</DropdownMenu.Root>
+				{/snippet}
+			</Tooltip.Trigger>
+			<Tooltip.Content>
+				<p>Merge {selectedBooks.length} books</p>
+			</Tooltip.Content>
+		</Tooltip.Root>
+	</Tooltip.Provider>
+{/if}
+
 <!-- Delete selected button -->
 <Tooltip.Provider>
 	<Tooltip.Root ignoreNonKeyboardFocus>
@@ -137,8 +227,8 @@
 				bookOps.deleteDialogTitle = `Delete ${selectionState.getSelectedIds().length} books?`;
 				bookOps.deleteFn = async (deleteFiles: boolean) => {
 					await bookOps.deleteBooks(selectionState.getSelectedIds(), deleteFiles);
-					libraryState.activeLibrary!.total! -= selectedBooks.length
-					bookshelfState.deletedBooks(selectedBooks)
+					libraryState.activeLibrary!.total! -= selectedBooks.length;
+					bookshelfState.deletedBooks(selectedBooks);
 					selectionState.deselectAll();
 				};
 				bookOps.deleteDialogOpen = true;
@@ -187,14 +277,43 @@
 	</Tooltip.Root>
 </Tooltip.Provider>
 
-
 <ShelfCreateDialog
 	bind:open={createShelfDialogOpen}
 	onSubmit={async (name: string) => {
-
-		const bookshelf = await bookshelfState.addBookshelf(name, libraryState.activeLibrary!.id, selectionState.getSelectedIds())
-		selectedBooks.forEach(book => book.lists.push(bookshelf))
-		selectionState.deselectAll()
-		createShelfDialogOpen = false
+		const bookshelf = await bookshelfState.addBookshelf(
+			name,
+			libraryState.activeLibrary!.id,
+			selectionState.getSelectedIds()
+		);
+		selectedBooks.forEach((book) => book.lists.push(bookshelf));
+		selectionState.deselectAll();
+		createShelfDialogOpen = false;
 	}}
 />
+
+<!--
+	Keyed on the selection so opening merge on a different pair starts from those
+	records rather than the previous dialog's draft.
+-->
+{#if mergingBooks}
+	{#key mergingBooks.map((book) => book.id).join()}
+		<MergeBooks
+			books={mergingBooks}
+			libraryId={libraryState.activeLibrary!.id}
+			bind:open={
+				() => mergingBooks !== null,
+				(value) => {
+					// Bound, not passed as a bare `true`: the dialog closes itself on an
+					// outside click or Escape, and if that never reaches `mergingBooks`
+					// the snapshot stays set — the menu item then re-assigns the same
+					// selection, nothing changes, and the dialog never reopens.
+					if (!value) mergingBooks = null;
+				}
+			}
+			onmerged={(_survivor, folded) => {
+				afterMerge(folded);
+				mergingBooks = null;
+			}}
+		/>
+	{/key}
+{/if}