mirror of
https://github.com/calibrain/shelfmark.git
synced 2026-11-15 07:28:26 +00:00
9ca1981d00d29fb9bfda971beeff6032f4d1672f
342
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 splitsec2
|
9ca1981d00 |
refactor: move the Socket.IO handlers into core/socket_handlers.py (#1638)
This one is small. The three Socket.IO handlers, connect, disconnect and request_status, move from `main.py` to `core/socket_handlers.py`, registered with `register_socket_handlers(socketio, ...)`. It is the Socket.IO item in the plan from Discussion #1536. They had no tests, so the first commit adds 9 that go through Flask-SocketIO's test client: what a client gets on connect with no auth, as an admin, as a user and as a visitor with no session, a failure reading the status, the manual status request, and the room bookkeeping on disconnect. They pass on current `main`. The second commit is the move. `resolve_status_scope` stays in `main.py` since the HTTP routes use it too, so it is passed in. One test line changed, the patch of `redact_status`, which now points at the new module. Log lines from the handlers now name `shelfmark.core.socket_handlers`. The URL map and the registered socket events are identical before and after, and the full suite, ruff and basedpyright are clean. I also started the app from the branch and the Socket.IO handshake answers. <!-- This is an auto-generated comment: release notes by coderabbit.ai --> <!-- coderabbit-ai-generated:v1 --> ## Summary by CodeRabbit * **Bug Fixes** * Status requests now return a consistent empty update when access is denied and an error message when status retrieval fails. * Status visibility follows the requester’s access level, with non-admin results redacted. * **Refactor** * Connection, disconnection, and status-request handling now follows consistent room and connection-tracking behavior. <!-- end of auto-generated comment: release notes by coderabbit.ai --> |
||
|
splitsec2
|
c6685feba5 |
fix(docker): drop ffmpeg from the default image; install on demand in Debug mode (#1631)
ffmpeg is in the default image for one thing, the Debug mode screen recording (`_start_ffmpeg_recording`). Nothing else calls it, SeleniumBase included, and it brings 93 packages with it. On real builds that's 406 Trivy findings (149 High) and 78 MB of the gzipped amd64 image. This takes it out of the image and installs it when Debug mode needs it. It came out of the base image discussion on #1616 and helps on any base, though on Wolfi the install line would need `apk`. ## What changed - **Dockerfile:** ffmpeg is gone from the full image's install. - **Guard:** `_start_ffmpeg_recording` checks `shutil.which("ffmpeg")` first. If it's missing it logs a warning and returns, and the bypass carries on. Without this, Debug on with no ffmpeg fails every bypass with `FileNotFoundError`. There's a new test for it. - **Entrypoint:** when Debug is on, the internal bypasser is in use and ffmpeg isn't already there, it installs ffmpeg in the background so the app doesn't wait for it. The download and the install each run under a 120 s `timeout` with `DEBIAN_FRONTEND=noninteractive`. Downloading first means a hung mirror can only be killed before dpkg starts, and the next start runs `dpkg --configure -a` in case a restart cut dpkg off. It logs when it starts, succeeds, times out or fails. - **Docs:** a short paragraph in `docs/troubleshooting.md` covering the install and the non-root limit. ## Debug has two sources Debug can come from the `DEBUG` env var or from the Debug Mode setting in the UI, which is saved to `/config/plugins/advanced.json`. The existing block in `entrypoint.sh` only reads the env var, so an install keyed on `$DEBUG` would miss anyone who turned it on in the UI. The new check asks the app instead (`shelfmark.config.env.DEBUG`), which reads both. The UI setting is `requires_restart`, so the restart it already asks for is what brings it through the entrypoint. ## Non-root mode A container started as `1000:1000` can't run `apt-get`, so the entrypoint logs that it can't install ffmpeg, the guard warns on each bypass, and there's no recording. Everything else works. The workaround for a debugging session is to start as root with `PUID`/`PGID`. ## Startup cost None that I could measure. With Debug on, `/api/health` answers in the same time as main (3-4 s with Debug only in config, 9-10 s from env), and ffmpeg was ready 6-8 s later on the runners. The trade-off is that the first bypass after startup may be unrecorded while the install runs. It logs the warning and passes. I saw both outcomes on a bypass started right after health. From reading the code: the install lives in the container, so recreating the container installs it again. It also runs after `tor.sh` and `wireguard.sh`, so apt goes through whichever egress is active. ## Testing Real Docker builds on GitHub Actions on my fork, amd64 and arm64, with the same result on both for every case below. | Case | Result | |---|---| | Debug off | No install, `/api/health` in 3-4 s, same as main | | Debug only in config | Health in 3-4 s, ffmpeg ready 6-8 s later, real mp4 after a bypass | | Debug from env | Health in 9-10 s (same as main), real mp4 after a bypass | | Started as root, app dropped to uid 1000 | Smoke and bypass pass | | Started with `--user 1000`, Debug on | Logs both messages, bypass passes, no mp4 | | apt blocked (closed proxy) | Logs `Could not install ffmpeg`, bypass passes, no mp4 | | apt's proxy on a dropped address, so no TCP reply at all | Health in 9-10 s, a bypass during the install passes with the warning and no mp4. apt gives up by itself in under a minute. Nothing left running, `dpkg --audit` clean | | DNS server on a dropped address | Health in 9-10 s, `Timed out installing ffmpeg after 120s`. Nothing left running, `dpkg --audit` clean | | `--network none`, Debug on | No hang, health 200 from inside the container | | Container restarted while dpkg was running | Next start finishes the install, `dpkg --audit` clean, real mp4 | | Fresh named volume, Debug only in config | Recording made, uid 1000 can write | | Existing uid-1000 volume, Debug only in config | Same, nothing root-owned afterwards | | `find /config /var/log/shelfmark -user root` | 0 files, same as main | | `find /app -user root` | 245, same count as main | | `tests/e2e/platform/run-e2e.sh` with `dns-blocked` and `baseline` | Pass | The bypass checks ran against example.com. A real Cloudflare challenge was served by a demo page in the same runs, and no build solved it, including main, so that part doesn't say anything about this change. Locally: 5494 Python tests pass, and `ruff`, `basedpyright`, `vulture` and `shellcheck` are clean. The full results are in [the base image report](https://github.com/splitsec2/shelfmark/blob/7afcecfb76727724a71561cbc078f091569b58dc/docs/research/2026-11-01-base-image-report.md). |
||
|
splitsec2
|
34c9e6dbd7 |
refactor: move the auth routes into core/auth_routes.py (#1625)
The auth routes are the next group from the plan in Discussion #1536: `/api/auth/login`, `/logout` and `/check`, plus the login lockout helpers, move from `main.py` to `core/auth_routes.py`, registered with `register_auth_routes(app, ...)`. `main.py` is about 370 lines shorter. This is the login path, so I kept it strictly to a move. Two commits. The first adds 59 tests that pin the current behaviour before anything moves: login in each mode (builtin, none, CWA and the OIDC password fallback), wrong passwords, the lockout (the warning when five attempts are left, the 429 and its text on the tenth failure, per-user isolation, expiry), the local-auth switches, logout in each mode and `/check`. They pass on current `main`, and I ran the same file against `main` and against this branch. The second commit is the move, and no existing test changed. The lockout state is one dict, now living in `auth_routes`, and `main.py` imports the same object so its other users (the password-attempt limiter for the self-user routes, and the tests) see it. Everything else that comes from `main` (the user database, the auth mode, the client IP, the CWA database path, the local-auth settings and the password check) is passed in as a callable evaluated per request, so patches on `main` still apply. Two things to know. `main.py` re-exports the lockout names and keeps three lines assigning the view functions (`api_login = app.view_functions["api_login"]` and the other two), because the e2e tests call `main_module.api_login()` directly and I did not change tests in this PR. If you would rather drop both, I can edit those e2e tests instead. The other is that the lockout dict is no longer reset by a reload of `main`, which only matters in tests, and they clear it themselves. The URL map is identical, all 77 rules, and the full suite, ruff and basedpyright are clean. I also ran the app with builtin auth on a scratch instance and went through it over HTTP: a wrong password gets a 401, login works and `/check` shows the session, logout clears it, the tenth failed attempt gets the 429, and the right password is refused while the account is locked. |
||
|
splitsec2
|
d1e959758d |
refactor: move the system routes into core/system_routes.py (#1624)
The system routes are the next group from the plan in Discussion #1536: `/api/config`, `/api/health`, `/api/debug`, `/api/restart` and the routes that serve the frontend (`/`, the SPA catch-all, `/assets/...`, the logo, the favicons and `theme-init.js`) move from `main.py` to `core/system_routes.py`, registered with `register_system_routes(app, login_required, ...)`. `main.py` is about 250 lines shorter. Same two-commit layout as before. The first commit adds 14 tests for the frontend-serving routes and the `/api/config` error paths, which had none, and they pass on current `main`. The second commit is the move, and no existing test changed. The helpers only these routes use moved with them (`admin_required`, built on the `login_required` that is passed in, plus `_base_href` and `_serve_index_html`). What `main.py` still owns, the base path, the frontend folder, the auth mode and the session user id, is passed in as callables evaluated per request, so the tests that patch them on `main` still work. The debug routes are only registered when debug is on, as before, with the stop-GUI and run-script hooks handed in. The thing to look at is the SPA catch-all, which is now registered earlier in `main.py` than it used to be (the old comment said it must be last). Werkzeug picks the most specific rule whatever the registration order, and I checked it: all 76 rule and method pairs in the URL map still match their own endpoint, on `main` and on this branch, and an unknown `/api/...` path still gets the JSON 404. The Flask URL map is identical before and after, all 77 rules, and with debug on (79). The full suite, ruff and basedpyright are clean. I also ran the app from this branch: health, config, the logo, the favicon and the theme script answer as before, a client-side path like `/library/books` gets the index page, an unknown `/api/` path gets the JSON 404, and the page loads in a browser. I did not exercise `/api/restart` or a real debug script run, since nothing tests them. |
||
|
splitsec2
|
d41fa98749 |
refactor: move the metadata routes into core/metadata_routes.py (#1623)
The metadata routes are the next group from the plan in Discussion #1536: the eight `/api/metadata/...` handlers (providers, config, search, field-options, book and the three targets routes) move from `main.py` to `core/metadata_routes.py`, registered with `register_metadata_routes(app, ...)`. `main.py` is about 490 lines shorter. Same two-commit layout as the settings move (#1595). The first commit adds 39 tests that pass on current `main` before anything moves. They cover the 401 on all eight routes, the shape of the providers list and the config, the search errors, the book lookup (200, 404, 400 and 503), and how target errors map to 400, 502 and 500. The second commit is the move, and no test changed in it. Three helpers moved with the routes because nothing else uses them. `_attach_library_ownership` stays in `main.py` since the releases routes need it too, and is passed in along with the login decorator and the auth mode, the way `activity_routes` takes them. The auth mode is a callable evaluated per request, so the one test that patches it on `main` still works. The Flask URL map (every rule, its methods and its endpoint name) is identical before and after, all 77 rules. The full suite, ruff and basedpyright are clean. I also ran the app from this branch and hit the routes over HTTP: providers and config answer 200, a book that doesn't exist gets a 404, and search answers 503 with the "no metadata provider configured" message on an instance with none enabled. I did not look at it in a browser or run the e2e tests. Log lines for these routes now come from `shelfmark.core.metadata_routes`, and the error tuples are local copies so the module doesn't import `main`. One thing to know when reading the diff: the register call sits just after `_attach_library_ownership` and not where the first route used to be, because that function has to be defined before it can be passed in. |
||
|
splitsec2
|
9bd6c51388 |
fix(hardcover): read the featured series from a search result (#1589)
Closes #1585. @joeRob2468 traced this one to the parser, and the answer turned out to be sitting in the search response already. A Hardcover search hit carries the book's featured series as `featured_series`, with the same shape as the `featured_book_series` of a full record: the position and a nested series with its id, name and book count. `_parse_search_result` never read it, so a book requested from a search card was stored with `series_name: null` and the `{Series}` folder and the custom script payload came out empty. It now reads it, through a small helper that `_parse_book` uses too, so there is no extra Hardcover call per request (the issue's other option was fetching the full record, which costs one against the rate limit). I checked it against a live search for "Mistborn The Final Empire". The top hit parses to series "The Mistborn Saga: The Original Trilogy", position 1, id 10001, and its search title is still "Mistborn". That last part is on purpose. I left `_compute_search_title` alone, since passing it the series would change what gets searched for everyone and the series name here doesn't equal the "Mistborn" prefix anyway. Two things that behave differently. A search result now has `series_name` set, so anything in the UI that shows it for a full record (the series link, the bigger cover in the release modal) will show it on search results too. I ran the branch and looked at the release modal for the top Mistborn result: it shows "#1 of 3 in The Mistborn Saga: The Original Trilogy" with the View series link. The modal fetches the full record when it opens, so I can't say how much of that is new on this path. Requests already stored keep the empty series they were saved with. Tests: four in a new file (a hit with a series, hits with none or an incomplete one, a hit nested in a `document`, and the search title staying the same), the first and third fail on current main. The metadata and core suites, ruff and basedpyright are clean. |
||
|
splitsec2
|
cd5a478831 |
refactor: move the onboarding routes into core/onboarding_routes.py (#1594)
First of the route moves from Discussion #1536, following the plan table there: the three onboarding routes (`GET` and `POST /api/onboarding`, `POST /api/onboarding/skip`) leave `main.py` for `core/onboarding_routes.py`, registered with `register_onboarding_routes(app, login_required)` the way `release_inspect_routes` takes it. `main.py` goes down by 72 lines and the handler bodies are the same code (I diffed them after de-indenting: the only difference is the local name of the error tuple). There are two commits so the move can be reviewed on its own. The first adds 15 tests for these routes, which had none, and they pass on current `main` before anything moves. They pin the behaviour I most wanted to keep: a 401 without a session when auth is on, a 403 for a signed-in non-admin (that check lives in `login_required` and goes by path, which is why the same decorator has to be passed in), the 400 on an empty body, the result passed through on success and failure, the 500 with the error text on an operational error, and open access when no auth is configured. The second commit is the move, and no test changed in it. What I checked on the move: - The Flask URL map (every rule, its methods and its endpoint name) is identical before and after: 77 rules, the three onboarding ones included. - Full suite, ruff and basedpyright are clean. - I ran the app from this branch and stepped through the setup wizard in a browser. It loads its steps from the moved route and goes from Search Mode to Metadata Provider, and `POST /api/onboarding/skip` and the empty-body 400 answer as before. Two small things that do change. The log lines for these routes now come from `shelfmark.core.onboarding_routes` instead of `main`, and the error tuple is a local copy (`_ONBOARDING_ERRORS`, the same classes as `_IMPORT_OPERATIONAL_ERRORS`) so the module does not import `main`. Nothing else is different. |
||
|
splitsec2
|
69cbf72698 |
refactor: move the settings routes into core/settings_routes.py (#1595)
Second route move from the plan in Discussion #1536: the four settings routes (`GET /api/settings`, `GET` and `PUT /api/settings/<tab>`, and `POST /api/settings/<tab>/action/<key>`) go from `main.py` to `core/settings_routes.py`, registered with `register_settings_routes(app, login_required)`. `main.py` is 143 lines shorter. These handlers call `settings_registry` directly, so nothing else has to be passed in. The handler bodies are the same code, and the only difference I found after de-indenting them is that `ruff format` wrapped one signature that no longer fits on a line. Same two-commit layout as the onboarding one (#1594). The first commit adds 27 tests for these routes and they pass on current `main` before anything moves. They pin the 404 for an unknown tab, the 400 when a `PUT` body is not a JSON object (a list, a string, a number, or no body at all), the "No changes to save" answer for an empty object, the unsaved form values reaching an action, the result passed through with a 400 on failure, the 500 with the error text on an operational error, a 401 without a session, a 403 for a signed-in non-admin on all four routes, and open access with no auth. The second commit is the move and no test changed in it. The admin check for these paths lives in `login_required`, so the same decorator is passed in. What I checked: - The Flask URL map (every rule, its methods and its endpoint name) is identical before and after, all 77 rules. - Full suite, ruff and basedpyright are clean. - I ran the app from this branch and opened Settings in a browser: the modal lists every tab and the General tab shows its values, which come from the moved `GET`. Over HTTP, a `PUT` of a new search page title saved and read back, a list body got the 400 and an unknown tab the 404. As with #1594, log lines for these routes now come from `shelfmark.core.settings_routes`, and the error tuple is a local copy (`_SETTINGS_ERRORS`) so the module does not import `main`. The two PRs touch neighbouring lines of `main.py` and I checked that they merge together without a conflict, in either order. |
||
|
splitsec2
|
7583f997bb |
fix(oidc): catch joserfc's InvalidClaimError so the issuer guidance shows (#1603)
A wrong OIDC issuer, which is the usual IdP setup mistake, sends the
user back to the login page with a bare "Authentication failed". The
callback has a handler that explains it ("OIDC issuer validation failed.
Verify your discovery URL and IdP issuer/external URL configuration."),
and it never runs.
Authlib validates the ID token with joserfc now, so the error that comes
out of `authorize_access_token()` is `joserfc.errors.InvalidClaimError`.
The handler catches the one in `authlib.jose.errors`, which is a
different class with a different base, so the isinstance check is False.
I confirmed it by validating a real `CodeIDToken` with a wrong `iss` on
Authlib 1.9.1. I haven't run it against a live IdP.
The existing tests passed because they raise the old class from a mock.
The new one raises the error from a real `CodeIDToken`, and it fails on
the old import with `'Authentication failed'` in place of the guidance.
Two small things come with it:
- The new class names its attribute `claim` where the old one used
`claim_name`, so that read changes.
- `authlib.jose` is deprecated in favour of joserfc and Authlib's own
warning says it goes in 2.0. This was the only place Shelfmark used it,
so the import moving also takes us off the module that is going away.
joserfc is already installed as an Authlib requirement, so I added it to
`pyproject.toml` (`>=1.6.8`, Authlib's own floor) and the lock gets the
one entry. No new package.
Ruff, basedpyright and the backend suite pass (5424 passed).
|
||
 CaliBrain
|
947ce7bfe5 |
test(e2e): serve the mock Prowlarr .torrent from Prowlarr's own origin (#1602)
The nightly client-deluge profile has failed since
|
||
|
splitsec2
|
83e61975c4 |
fix(releases): apply the record policy to a provider named alongside a source (#1598)
Two small things I found while going back over what has landed since 1.5.0. `/api/releases` takes a `provider` and a `source`, and the policy check only looked at `source`. A user whose policy blocks one release source but not another could pass the blocked one as `provider` with the allowed one as `source`, and get the blocked source's record back. The record route already refuses that with a 403, so the listing route now applies the same refusal when `provider` names a source that is not being listed. The normal case, where provider and source are the same, is already gated and skips the extra check. The second commit is wording only. The Automatic Downloads setting and the Hardcover sync doc said "synced requests", but the pass picks up every pending Hardcover request, including ones a user made themselves with Request Book. The download is still queued as the user who asked (the admin is only the reviewer, same as a manual approval), and admins already see every user's requests. The text now says that, and a new test pins the behaviour for a user's own request. I opened a scratch instance and looked at the Hardcover Sync tab: the Automatic Downloads heading and the checkbox description read as intended and nothing else on the tab moved. Full suite, ruff and basedpyright pass, and the new policy test fails without the fix. |
||
|
splitsec2
|
cfba44b754 |
refactor: one module for taking credentials and links out of text (#1578)
Draft until #1571 merges: this is stacked on it, so the first two commits here are that PR's and only the last one is new. I will rebase once #1571 is in. Shelfmark strips credentials and links out of text in five places, written in different PRs with different patterns: the log formatter, `_safe_url` and `_redact_urls_in_text` in `torrent_utils.py`, `redact_status_message` in `viewer_redaction.py`, the `sed` rules in `genDebug.sh`, and the password handling in the settings registry. Between them there were three regexes that find a URL in text, two separate userinfo patterns and no shared list of what counts as a secret. #1571 made the gap visible, because every pattern added to the logger had to be remembered somewhere else too. This moves the text and URL part into `shelfmark/core/redaction.py`, which imports nothing from the app so the logger can depend on it: - `redact_secrets` (mask in place, for logs) moved from `core/logger.py`, patterns unchanged. The logger imports it. - `safe_url` and `redact_urls_in_text` (keep scheme, host and path, drop the query) moved from `torrent_utils.py` and are public now, so `alldebrid.py` no longer imports a private name from another module. - `strip_links` (remove links entirely) is what `redact_status_message` now calls. It and `redact_urls_in_text` share one link pattern, the one the viewer redaction already used. Two small behaviour changes come from sharing that pattern. A closing bracket or quote right after a URL is no longer swallowed into it when scrubbing an exception message. And the scheme-less path that requests prints after `url: ` now keeps only what is before its query in `redact_urls_in_text`, where before it passed through as it was (the log formatter masked the key, but the path and query stayed in the message). Also here: the two `sed` rules in `genDebug.sh` that mask userinfo now stop at `?` and `#` like the logger pattern does, with a test that fails on the old script. Not in this PR: the debug bundle zips the live log directory as it is, so log lines written before #1571 are not redacted in it. Fixing that means copying the logs to a staging folder and running them through this module, which I would do as a separate change. Tests: a new file covers `safe_url`, `redact_urls_in_text` and `strip_links` directly, and two checks that the viewer and the logger really use these. Full suite passes. |
||
|
splitsec2
|
7a4f1670dc |
fix(slskd): try a batch enqueue once more when slskd answers 500 (#1590)
slskd answers a bare 500 when the peer doesn't reply to a queue request inside its own 5 second wait (its log says "The wait timed out after 5000 milliseconds"). The peer often answers a moment later, so a Soulseek download that would have worked failed on the first try. The batch enqueue is now tried twice with 5 seconds between. A 4xx or a connection failure is not retried, only a 5xx. While it waits the card reads "The peer did not answer, trying again", and cancelling during the wait stops without the second try. If the second try fails too, the error message carries what slskd said (trimmed to 160 characters). Before, that text was dropped and the card only showed the status line. It's been running on my instance since mid-October, and the timeouts still show up in slskd's log, so I can't say how many it saved. The five new tests cover the retry that works, giving up after the second 500, a 400 and a connection error not being retried, and cancelling mid-wait. Ruff, basedpyright and the slskd tests pass. |
||
|
splitsec2
|
e8b28d05e0 |
fix(sources): take the retry download URL from the release cache only (#1597)
Newznab and Jackett fell back to the download URL sent with the queue request when the release was no longer in the cache. Prowlarr already returns no URL in that case, so this makes the other two match it: the retry URL comes from the cache and nothing else. A request for a release that has expired from the cache now fails with "Release not found in cache" instead of being queued with whatever URL came along. The new tests cover both handlers on a cache miss and with no source_id, and they fail on main. Full suite, ruff and basedpyright pass. |
||
|
splitsec2
|
8b80f32476 |
fix(sources): keep credentials out of a source-unavailable message (#1593)
Code scanning alert 106 (stack trace exposure on the release search 503) turned out to be worth a look. Jackett sends its API key as `?apikey=` in the query string, and when Jackett is unreachable the source builds its "unavailable" message from the requests exception, whose text includes the full request URL. `/api/releases` returns that message to whoever ran the search, so any signed-in user could read the Jackett key while Jackett was down. I reproduced it with Jackett pointed at a closed port: the 503 body had `apikey=<the key>` in it. Jackett was added in #1521 and isn't in a release yet, so nothing shipped with this. `SourceUnavailableError` now masks credentials in its message with the same `redact_secrets` the log formatter uses (query parameters like `apikey=`, `user:pass@` in URLs and so on). I did it in the exception rather than at the two Jackett sites that build the text, since its message is written for the user and any other source that ends up with a key in a URL would hit the same thing. The rest of the message stays readable, for example "Max retries exceeded with url: ...?apikey=(removed)&t=book". Prowlarr and slskd send their keys in headers so they didn't leak a key, but their messages do include the internal host and port. I left that alone here. Tests: two Jackett ones (a connection error on the search and on the indexer lookup, with the key in the exception text) and three on the exception itself. The four that check masking fail on current main. The full suite, ruff and basedpyright are clean. |
||
|
splitsec2
|
72b094db7a |
fix(qbittorrent): map the qBittorrent 5 stopped states and forcedMetaDL (#1570)
On qBittorrent 5 a stopped torrent shows up as an unknown state. 5.0 renamed `pausedDL` and `pausedUP` to `stoppedDL` and `stoppedUP`, and `get_status` only knows the old names, so the torrent comes back as state `unknown` with the raw state name as its message. The paused handling in the download handler never applies to it, for example the grace it gives a torrent that has not started. `forcedMetaDL` is missing from the same table, so a force-started magnet fetching its metadata also shows up as unknown, although it is already in the incomplete-states set and the metadata wait. The three missing states are now mapped like their neighbours: the two stopped states as paused, `forcedMetaDL` as downloading with "Fetching metadata". The state names are the ones `qbittorrent-api` lists in its `TorrentState` enum. Tests: one parametrized test covers the old and new paused names and `forcedMetaDL` (the two old paused names are there as controls). The three new cases fail without the change. Full suite passes. |
||
|
splitsec2
|
f02f0d7365 |
fix(notifications): say what applies on a user's own notification routes page (#1588)
This is the UI side of #1581. A non-admin looking at their own notification routes reads the same help text the admin does, including "or mailtos://...?to={Email} in the admin default". They can't set that default, a personal route can't use mailtos (personal routes are limited to ntfy, Gotify, Discord, Slack, Telegram, Pushover and PushBullet), and nothing on the page says whether an admin default exists. The default's URL is hidden from them on purpose since it can carry credentials, so the page just shows an empty All row, the same as when nothing is set up. What changes, on the user's own page only (the admin's view of a user keeps the full text): - The help text drops the admin default and mailtos part and lists the services a personal route can use. - The edit-context response now carries `adminDefaultRoutesSet`, a boolean that is true when an admin default route with a URL exists. The URL itself is still not sent, and the existing test that checks nothing from the admin's values leaks still passes. - When that is true, the help text adds a line saying a default applies until the user adds their own, and once they have, that theirs replace it. How routes are resolved is unchanged. A user's own routes still replace the default and don't merge with it. I didn't try to add mailtos for personal routes, since keeping SMTP credentials in a user's own URL is what the admin default avoids. Tests: two route tests (flag true with a URL in the admin default, false with an empty row), a vitest file for the description text, and the full frontend suite, oxlint, oxfmt, tsc, ruff and basedpyright are clean. I haven't looked at the page in a browser. |
||
|
splitsec2
|
46b9e7d8a3 |
fix(library): match a shelf title against the full title when the search title is only the series (#1587)
Fixes #1584. Thanks to @joeRob2468 for the write-up, which pinned this down to the line. Hardcover cuts a title like "Mistborn: The Final Empire" down to "Mistborn" for searching when it can't tell the series prefix from a title prefix, and search results don't carry the series. The library check then matched the shelf title "Mistborn: The Final Empire" against "Mistborn", saw "final empire" as extra words and took it for a longer, different work (the "Dune" vs "Dune Messiah" rule), so the book got no In library badge. `match_entries` now tries the search title first as before and, if that finds nothing and the full title is different, tries the full title too. The full title is the more specific of the two so it should rarely add false positives. Another book in the same series still doesn't match, there is a test for that, and the Dune and Dune Messiah test passes unchanged. The matching loop moved into a small helper to run it twice, nothing in it changed. This only touches the library check. The same short search title also feeds release searches and auto-download matching, which I left alone here because loosening or tightening those changes what gets searched and downloaded for everyone. Tests fail on current main without the change. The library tests, ruff and basedpyright are clean. |
||
|
splitsec2
|
ac14e4dd24 |
fix(downloads): a swarm note change is not activity for the stall timer (#1572)
The torrent stall timer can be deferred for as long as the swarm keeps changing. Since #1494 the torrent handler appends a note to its progress message (`0% · 2 seeds`, and `· 0 seeds, giving up in 7m` while the no-seeders clock runs). `update_download_status` treats any changed message as activity, so a seed count that goes 2, 3, 1 between polls, or the give-up countdown ticking down every minute, refreshes `_last_activity` and the torrent never stalls. A torrent whose seed count stays constant still stalls, which is why this is easy to miss. `shelfmark/download/activity.py` already says what counts: a changed status event or a changed progress value, and a message that changes on a timer proves nothing. So `update_download_status` now compares messages through a small `activity_key()` that drops that trailing seed note. The card still gets every message and shows the live seed count, only the stall clock ignores a change in the note. A new progress percentage, a different client message or a status change still counts as activity. The handler builds the note in `_seed_note` and `activity_key` has to match it, so one test builds the note for a range of seed counts and countdowns and checks the key strips all of them. That also covers the gap #1494 left, where no test connected a handler's message to the orchestrator: the other two drive `update_download_status` with real notes and a fake clock. Without the change the note test fails and the key tests cannot even import `activity_key`. The other orchestrator test is a control that a progress or status change still counts. Full suite passes, and each new test also passes when run alone. |
||
  misko903andClaude Opus 5.5
|
d50440f6c2 |
fix(webshare): clear the "Resolving" status once the download starts (#1575)
## What
A Webshare.cz download shows "Resolving Webshare link" for its entire
duration, so a large download (e.g. a 1.3 GB audiobook archive) looks
stuck.
## Why
After the link was resolved, the handler called
`status_callback("downloading", None)`. `update_download_status` treats
a `None` message as "keep the current message", so the "Resolving
Webshare link" detail stayed in the activity panel. The HTTP handler
sends `""` for exactly this reason.
## How
- Report `downloading` with an empty message, which clears the resolving
text.
- When the download response has no `Content-Length`, use the release's
size (`task.size`, parsed with `parse_size_string`) as the total for
progress. Without it, progress never moved, and stall detection would
flag a healthy long download after 5 minutes.
## Testing
- New tests in `tests/webshare/test_handler.py`: the downloading status
clears the message, and progress is computed from the release size when
there is no Content-Length.
- `uv run pytest tests/webshare` is green (34 passed), and `ruff check`
and `ruff format --check` are clean.
Follow-up to #1543.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
|
||
|
splitsec2
|
525b094b57 |
fix(clients): hand a finished wrong-type directory on instead of polling out the timeout (#1573)
A torrent that finishes with the wrong kind of files in it holds the download for the whole completed-path timeout (60 seconds by default, up to an hour) and then fails with "Completed download has no stable supported files yet." An ebook task whose torrent holds only an `.m4b`, an audiobook task that got an `.epub`, or a `.pdf`-only release with the default format list all land here. In v1.4.0 these failed with the specific format error from post-processing, but the wait for the sync to settle (#1464) only counts files of the task's own type, so it keeps polling for files that are never going to appear. If the directory has no supported files, `_wait_for_completed_path` now looks for finished book or audio files that this task does not accept (the audiobook formats, the default ebook formats and pdf). When it finds some and they are the same on two polls in a row, it returns the path so post-processing can say what is wrong with the formats. Cover art and notes alone do not count, because that is what a directory looks like while the real files are still syncing over a remote mount, so those keep waiting as before. A wrong-type file that is still growing keeps waiting too. With no wait configured there is only one sample, and a wrong-type file is handed on straight away. The existing test that pins "no book files, no wait configured" uses a jpg, so it stays as it is. I extended the test helper to take a content type, which the wait tests never passed before. Four new cases fail without the change (a wrong-type directory for ebook, audiobook and pdf tasks), and two more check that sidecar-only and still-growing directories keep waiting. Full suite passes. |
||
|
splitsec2
|
47f0acdd31 |
fix(logging): mask userinfo and path-segment credentials, route stray loggers through the formatter (#1571)
Logs get pasted into issues, and the formatter from #1488 only masks `?key=value` query parameters. I went through what else commonly ends up in a log line and found these unmasked: - credentials in a URL's userinfo, like a proxy's `socks5://user:pass@host:1080` or `http://admin:secret@qbittorrent:8080` - a passkey as a path segment, like `/rss/<key>` or `/passkey/<key>/get.torrent`, which private trackers use - the token in a Telegram bot, Discord webhook or Slack webhook URL, which is what a notification URL looks like when it gets logged All three are masked now, in `redact_secrets`, so they apply to every handler. The path rule only fires after a segment that names a secret (`passkey`, `apikey`, `rsskey`, `authkey`, `token`, `secret`, `rss`, `torznab`) followed by 16 or more token characters. I left out generic words like `download` on purpose, because `/api/download/<md5>/cancel` appears in our own request log and the ids there are useful when debugging. That is the tradeoff: a passkey in a `/download/<passkey>/` path still gets through. Separately, four modules created their logger with plain `logging.getLogger`, so nothing they logged went through the formatter at all: `api/websocket.py`, `download/clients/__init__.py`, and the Jackett, Newznab and Prowlarr package inits. They use `setup_logger` now like everything else. Tests: ten cases, seven for the new shapes and three that must stay readable (an md5 route id, a short `/rss/latest`, an email address), plus two that check the stray modules now carry the redacting formatter. Nine of them fail without the change. Full suite passes. Not covered here: `/api/debug` still zips the existing log files as they are on disk, so anything logged before this change rides along in a bundle. That is a separate change in `genDebug.sh`. |
||
|
splitsec2
|
ef40a38546 |
fix: report a stopped download coordinator in /api/health (#1568)
If the download coordinator thread dies, every task sits at "Queued" and nothing says why. The rest of the app keeps answering, `/api/health` stays green and the container looks fine. #823 and #1166 were both that, and the loop's `except Exception` cannot catch everything that can end a thread. `/api/health` now reports it. The response gets a `coordinator` field, `running` or `stopped`, and when it is stopped the status is `error` and the HTTP code is 503. The Docker healthcheck in the Dockerfile and the one in the readme both use `curl -f`, so the container shows as unhealthy and anything watching health (Uptime Kuma, autoheal) can act on it. It needs no login, same as before, and shows no queue contents. That 503 is the one decision here. Keeping the 200 and only adding a `degraded` entry would not trip a status-code health check, which is the whole point, so I went with 503. Easy to flip if you would rather not. The orchestrator gets a small `coordinator_running()` that asks whether the thread is alive. `start()` already knows how to restart a dead thread, so restarting it automatically would be the true fix, but I would rather the failure be visible first and decide on that separately. Tests: three for the route (ok, 503, no login needed) and three for `coordinator_running` (never started, alive, dead). They fail without the change. One existing test hits `/api/health` through an app that never starts the coordinator, so it now patches that. The readme's health section has a sentence about the 503. Full suite passes, ruff and basedpyright are clean. |
||
|
splitsec2
|
973cf472a2 |
test(image-cache): flush window test no longer depends on float rounding (#1580)
`test_cache_hit_flushes_the_index_once_the_window_has_passed` (from #1544) failed on the first attempt of two of my PRs' Python Tests runs today (#1576, #1579) and passed on rerun, with nothing in either PR touching the image cache. The test sets the clock to exactly `last_saved + ACCESS_FLUSH_SECONDS` and expects a flush. `(now + 30) - now` isn't always 30.0 in floating point. For example `105.65207077218265 + 30 - 105.65207077218265` is `29.999999999999986`, so the `>=` check is false and the flush never happens. That only shows up when the monotonic value is small, which fits a freshly booted runner (on this box the uptime is in the millions of seconds and it never fails). I haven't been able to see the runner's actual uptime in the CI logs, so that part is an inference, but pinning `_last_saved` to that value reproduces the exact failure (`assert [] == [1]`) locally. The test now puts the clock one second past the window. The production check is fine, real elapsed time doesn't have this edge, so only the test changes. |
||
|
splitsec2
|
b248bcd536 |
fix(clients): import a qBittorrent torrent only when every listed file is on disk (#1577)
Fixes #1561, with thanks to @tmussche for the write-up, which had the cause and the fix. When the files reach Shelfmark through a sync (Resilio in the report), the wait for the completed path accepts the folder once two samples a few seconds apart look the same. A sync that copies in batches can pause longer than that, so 10 of 23 files were imported and some of those were still partial copies. For qBittorrent the client can say what the finished download holds, so the wait now checks that. `DownloadClient` gets an optional `get_expected_files()` that returns `ExpectedFile(name, size)` entries, or None when the client cannot say, which is the default and keeps every other client as it was. qBittorrent answers it from `/api/v2/torrents/files`, leaving out files the torrent does not download (priority 0), zero-byte entries and BEP 47 padding files. `_wait_for_completed_path` asks once, and while any of those files is missing or smaller than its final size it keeps waiting, using the same timeout as before. When the wait runs out the error now says how far it got, for example "Completed download is not fully on disk yet (10 of 23 files)." Once everything is there the existing stability check still runs, so nothing is accepted that was accepted before. Names from the client start with the torrent's top-level folder, so each file is looked up with and without it, which also covers a torrent saved without a subfolder. A single-file torrent is checked by size too, since a file that is still syncing exists at a smaller size. Two things I would like a second look at: the comparison is on exact size, so a setup where something rewrites the files after the torrent finishes would never match and would run into the timeout; and the file list is read once per wait, not on every poll. If the client does not answer, nothing changes. I also moved the nested authenticated GET in the path derivation into a `_get_with_auth` method so both calls share it, and added a sentence about this to `docs/configuration.md`. Tests: eight for the wait (partial folder, truncated file, files arriving during the wait, complete folder, no top-level folder, single file both ways, and a client without a list), and six for the qBittorrent file list. The wait tests that depend on the new behaviour fail on current main. Full suite passes. |
||
|
splitsec2
|
9011319ade |
refactor: one lookup for the supported formats, public names for two shared helpers (#1576)
Six places read the Supported Book Formats and Supported Audiobook Formats settings on their own: post-processing, the post-process scan, IRC, slskd, Prowlarr and direct download. For ebooks they already agreed, since they all fall back to `DEFAULT_EBOOK_FORMATS`. For audiobooks they did not. The settings registry and slskd fall back to every audio format plus zip and rar, while post-processing and IRC fell back to m4b and mp3 only. They now share `shelfmark/core/formats.py`, with `get_supported_formats`, `get_supported_audiobook_formats` and a content-type aware `get_supported_formats_for`, and a shared `DEFAULT_AUDIOBOOK_FORMATS` in `core/utils.py` next to the ebook one. `policy.py` re-exports the same functions, so nothing that imports them from there changes. It is not in `policy.py` itself because importing the `postprocess` package pulls in the router, which imports the release sources, so a source could not import from it without a cycle. One behaviour difference, and only in the fallback: with the audiobook setting unset, or emptied, post-processing and IRC used to accept m4b and mp3 and now accept the full default list like slskd and the settings page already did. A configured list is read exactly as before. Direct download's own helper used the raw configured list and now reads the same normalized one, which is identical for any list the settings page can save. Also public now: `parse_size` in the Prowlarr source (slskd and Newznab imported it with the underscore) and `coerce_completed_path_timeout_seconds` in the client base handler (slskd imported that too). Not changed on purpose: slskd's `_coerce_int` looks like a copy of `coerce_int` but is not. It reads `"12.0"` as 12 and `True` as 0, where `coerce_int` gives 0 and 1. And the `_parse_size` in the IRC source is a different function, string to bytes. Tests: a new set checks that post-processing, IRC and slskd all fall back to the shared defaults and that the settings default is the shared one, and the scan-archive and IRC tests now patch the shared config instead of the per-module copies. Full suite passes. |
||
|
splitsec2
|
606e94fde2 |
fix(settings): offer the seed settings for qBittorrent only (#1574)
The two seed settings, Give Up With No Seeders and Minimum Seeders, show up for every torrent client, but only qBittorrent reports a seed count. With Transmission, Deluge or rTorrent you can set both and nothing happens. The first one says "Only qBittorrent reports seed counts" at the end of its description, the second says nothing about it. Both are now shown only when qBittorrent is the selected client, the same `show_when` the other qBittorrent fields use, and Minimum Seeders gets the same sentence. A value someone already saved for another client stays in their config, it just is not shown, and it had no effect before either. `docs/environment-variables.md` is regenerated for the changed description, which is two lines. One test checks that both fields are tied to qBittorrent and say so, and it fails without the change. |
||
|
splitsec2
|
70fe4b8c89 |
fix(hardcover): keep the scheduler alive on an unreadable interval, share one history connection per sync (#1569)
These are the follow-ups from the review notes on #1516 and #1534 that I said I would do once the series had merged. The scheduler waits for the next cycle outside the try that guards the cycle itself, so an interval setting that raised would end the thread and nothing would sync again until a restart. It now logs the error and tries again after six hours. `docs/hardcover-sync.md` says so. The history check in the sync opened a new SQLite connection for every book on the shelf. It opens one per sync now and shares it, which adds up on a slow disk or a small board. The two test nits from #1534 are in as well: a `parametrize` with a single case, and an unused argument on a helper. The thread surviving a failed wait and the shared connection each have a test that fails without the change. Ruff, basedpyright and the backend suite pass (5297 passed). |
||
|
splitsec2
|
a63d9e3e21 |
docs: regenerate the environment variable reference (#1565)
Sorry about this one. The environment variable references had also fallen well behind, and again I was a big part of that. A bunch of the gaps come from my own PRs (Hardcover sync, auto-download, the Audiobookshelf library check) where I charged ahead with the functionality and skipped the boring documentation part (did I mention that documentation is boring?). I again had Claude compare the docs to the code and generate the updates, so it would be good to have another human look this over before it merges. The doc diff is just the generator's output. The part worth a real read is the handful of new entries in `scripts/generate_env_docs.py`, since I wrote those descriptions from `entrypoint.sh` and `tor.sh` and not from a settings definition. What changed: - Regenerated `docs/environment-variables.md`. It had 287 variables and the generator now produces 311. The 21 that were missing are the Hardcover Sync tab (6), auto-download (11) and the Audiobookshelf library check (4). Nothing was removed and no existing default, type or option changed. - `SHELFMARK_API_KEY_READONLY` was only ever hand-edited into the doc, so the generator did not know it and a plain regeneration would have dropped the row. It is in the generator now. - Added `UMASK`, `TOR_CHECK_INTERVAL` and `TOR_MAX_FAILURES`. The code reads all three and no doc mentioned them. - The table of contents built its anchors by lowercasing and swapping spaces, so the Webshare.cz and Moly.hu links were already broken (GitHub drops the dots) and Soulseek (slskd) would have broken with the parentheses. They now follow GitHub's rule. Two tests cover the new entries and the anchors, and both fail without the generator change. To regenerate: `python scripts/generate_env_docs.py` on Python 3.14. A test that fails when the committed file differs from the generator output would have caught this drift, but then every PR that adds a setting has to regenerate the doc. I left that out, your call if you want it. |
||
|
splitsec2
|
d7a4b6b859 |
fix: keep exception text out of the status, config and release search responses (#1562)
Same idea as #1558, for three more places that hand `str(e)` to whoever is logged in. These are code scanning alerts 106, 142 and 143. `/api/status` and `/api/config` returned the raw exception text of an OSError, RuntimeError and so on with the 500. They now return "Could not read the download status" and "Could not load the configuration". The full error is still logged with its trace. For `/api/releases`, a source failing with an unexpected error put `source: <exception text>` into the error list that goes back to the caller. `search_source_releases` takes a new `include_error_detail` flag. The route passes True for admins (and when auth is off, where everyone is treated as an admin) and False for everyone else, who get `source: search failed`. The message of a `SourceUnavailableError` is written for the user, so it is kept either way, and the background auto-download caller keeps the default and sees the full text as before. Left alone on purpose: the `reason` in the release inspect response, since a test covers that it carries the error text and it looks intentional. Tests: two for the routes and two for the helper. All four fail without the change. Full suite passes (5295), ruff and basedpyright are clean. |
||
|
23b3b3a9d8 |
feat: add Jackett release source (#1521)
## Summary Adds Jackett as a separate release source (`jackett`), as requested in #1443. - New `shelfmark/release_sources/jackett/` package: API client, settings tab, source and handler. - Searches Jackett's Torznab endpoints: the `all` aggregate by default, or each selected indexer individually. Failed or timed-out searches raise `SourceUnavailableError` instead of reporting "no releases found". - Reuses Prowlarr code rather than duplicating it: the Torznab parser (now also reads `<jackettindexer>`), `_prowlarr_result_to_release` (new optional `source`/`source_id` args), the release cache, URL/protocol helpers, the timeout clamp and size-limit helper (new optional config-key args), and the handler's `list_files` / retry-restore logic. - Downloads use the existing shared torrent client settings. - `JACKETT_URL` is a trusted origin for `.torrent` fetching (it is usually on the LAN). Prowlarr's `X-Api-Key` is not sent to it. - Added a Jackett section to `docs/environment-variables.md`. ## Screenshots <img width="1117" height="936" alt="Screenshot 2026-10-21 at 13-45-34 Shelfmark" src="https://github.com/user-attachments/assets/00339999-72c7-4226-bd9a-df8b325a7e22" /> <img width="1117" height="938" alt="Screenshot 2026-10-21 at 13-50-38 Shelfmark" src="https://github.com/user-attachments/assets/3c9cd2b4-932a-4b5c-9e10-6ec22b5f987d" /> <img width="1117" height="937" alt="Screenshot 2026-10-21 at 13-50-57 Shelfmark" src="https://github.com/user-attachments/assets/9c4d0ed9-da2f-4323-b188-341b45066efd" /> ## Not included - First-run onboarding does not offer Jackett yet. - No Jackett-specific category settings; it uses the standard 7000 / 3030 categories. ## Testing - New unit tests in `tests/jackett/` (client, source, handler) using mocked responses. - Full suite passes except 11 `tests/e2e` tests that also fail on a clean tree here (cannot write `/var/log/shelfmark`). - Tested against a live Jackett instance. Closes #1443 🤖 Generated with [Claude Code](https://claude.com/claude-code) --------- Co-authored-by: Claude Sonnet 5.5 <noreply@anthropic.com> Co-authored-by: CaliBrain <calibrain@l4n.xyz> |
||
|
splitsec2
|
3633dd3d97 |
fix: keep file links and API keys out of AllDebrid logs, fix Newznab debug log (#1559)
Two log fixes in the same area, both found while going through the open
code scanning alerts.
AllDebrid: the unlock path logged the file link in full. The last-resort
fallback appends `apikey=` to the link and returns it, and the "Failed
to download from ..." error carries that same link into
`logger.exception`, so the key could land in the log and the traceback.
The links are now logged through `_safe_url` (scheme, host and path
only). The GET unlock fallback also logged the exception text at debug
level, and requests puts the request URL in that message, which includes
`apikey=`. That one now logs the exception type only. This covers code
scanning alert 124.
Newznab: `logger.debug("Newznab API: GET %s params=%s", url, *params)`
star-expands the dict keys into the log arguments, so with debug logging
on it raises `TypeError: not all arguments converted during string
formatting` and the line never prints. Logging `params` as is would
print the API key, so it now logs the sorted parameter names.
Tests: the Newznab one formats the logged line and checks the key is
absent. The AllDebrid ones check the fallback and failure paths log no
query string and that the fallback still returns the link with the key.
All three fail without the change. Ruff and basedpyright are clean,
`tests/newznab` and `tests/prowlarr` pass.
|
||
|
splitsec2
|
4f97fc9bec |
fix: keep server paths and URLs out of an error response and a debug log (#1558)
Two small leaks from going through the open CodeQL alerts (code scanning alerts 100 and 123). The local download route returned `str(e)` on an OSError, and for a missing or unreadable file that text is the file's path on the server (`[Errno 2] No such file or directory: '/audiobooks/Author/Title.epub'`). #1422 cuts a non-admin's download path down to the file name everywhere else, so this put it back. The route now answers "Could not read the file" and the detail stays in the log. `should_bypass_proxy` logged the whole request URL at debug when it could not be parsed. URLs can carry credentials, so it logs only the exception now. Each change has a test that fails without it. The full non-e2e suite, ruff and basedpyright pass. |
||
|
splitsec2
|
86fff32297 |
feat(hardcover): owned books, simpler queries, dead sources and failed releases (#1535)
A request whose book turns out to be in the library is closed with the note `[auto] Already in the library.` instead of sitting in Needs Review. A request an admin handled by hand while a pass was running is skipped. A series label in front of a Calibre title no longer hides an owned book, and a Calibre title marked as another edition (full cast, dramatized, GraphicAudio) is not the book. When no source matches, up to three simpler forms of the title are tried on each source under the same strict match. Optionally (off by default), an unmatched audiobook request is looked up on Audible US and UK through Audiobookshelf's metadata search and the request says what the store answered. It never declines anything. A Soulseek peer whose recent transfers all failed, and an AudiobookBay post whose comments say it has no seeders, are tried after the other matches and never excluded. A failed or stalled release is kept out for AUTO_DOWNLOAD_RETRY_FAILED_RELEASE_DAYS and then tried again, and a torrent that died for one request is not offered to another. A request stuck on queued or downloading is settled from the download history. Optionally (off by default) a failed download is retried with another release within minutes. |
||
|
splitsec2
|
9d5b47d17b |
feat(hardcover): retries, error counting and a back-off for requests nothing matches (#1534)
A synced request whose download failed, was interrupted by a restart, or was cancelled by the stall timer goes back to pending after AUTO_DOWNLOAD_RETRY_DAYS (7 by default, 0 turns it off), a few per pass, oldest first. Only what never arrived is retried, the failed release is remembered, and a request is given up on after three failed retries. Downloads cancelled by a person and rejected requests are never retried. Errors count as errors: a search that failed in one source no longer reads as no match, a throttled Hardcover lookup is an error and not a skip, and a request that blows up doesn't abort the pass. Both summaries carry the error count. A request for the same Hardcover book and format is skipped when the history already has a finished or running download for it. A request that finds nothing is searched once a day after three no-match passes, then every two days, up to AUTO_DOWNLOAD_BACKOFF_MAX_DAYS (3 by default, 0 turns it off). |
||
|
splitsec2
|
4389548deb |
feat(hardcover): ebooks, audiobooks or both, and stricter matching for automatic downloads (#1533)
A shelf book can be requested as an ebook, an audiobook or both (one request per format), and each request is checked against its own library, source priority list and format guard. Ranking within a source prefers the most downloaded copy, then size, and one m4b is preferred to a set of chapter files unless it is much smaller than the other copies. Soulseek peers are ranked by free slot, queue length and speed. Automatic downloads refuse multi-book packs, sequels and other works named in the title, study guides, abridged or dramatised versions, two-in-one titles, releases in a language the requester didn't ask for, and an archive for an audiobook request. Matching folds accents on both sides, drops an alternate title for the search and the match, and reads small numbers the same in any form. AudiobookBay releases with no author in the listing title are judged on the authors credited on the detail page. |
||
|
splitsec2
|
ea69db3f8e |
feat(library): Audiobookshelf library check, and the sync fails closed when a library cannot be read (#1519)
Adds an Audiobookshelf provider for the library check, next to the Calibre one. It reads the server over its REST API (title, author, ISBN and ASIN per item), answers for audiobooks, and has its settings in General with a test button that uses the unsaved form values. Read only. The Hardcover sync and automatic downloads now fail closed when an enabled library cannot be read: the sync skips its cycle and an automatic download holds the request. A library that listed items at the last index and now lists none counts as a failed read. Badges still fail open. |
||
 ProcyonX1
|
99a6aa206b |
feat(metadata): add Deutsche Nationalbibliothek (DNB) metadata provider (#1547)
## feat(metadata): add Deutsche Nationalbibliothek (DNB) metadata provider This adds the Deutsche Nationalbibliothek (DNB) as a metadata provider. It covers German editions far better than the existing providers. No API key, disabled by default, and the same pattern as the other providers. It uses DNB's public SRU interface and parses MARC21-XML. Thats one reason why this needs more code than the Google Books or Open Library providers. The others are the search (DNB has no relevance ranking) and the language filter, see below. ### Notes - **Search box.** Search-box text goes from precise to loose until a stage has hits: the complete title, then the title as a phrase, then every word in a title or a person's name, last without German articles. - **"Only show editions in my book languages"** (optional, on by default). DNB also catalogs a lot of translations. DNB filters by the searching user's Default Book Languages, else the system default. - **No covers.** DNB's cover service is behind Anubis bot protection, so the provider returns no cover URLs. - **Proxy support** already included like Open Library since #1526 ### Changes outside the provider - DNB is in the onboarding dropdown. - Calibre's `dnb-idn` identifier is mapped, so DNB results show library ownership - In the env-var docs I added only the DNB section. Regenerating the whole file on main currently also reorders unrelated sections and drops `SHELFMARK_API_KEY_READONLY` (and the Databáze knih section sits inside the Moly.hu block). ### Ideas - **Deduplication.** DNB lists every edition. Kafka's "Die Verwandlung" has around 200 German editions of the same text. Grouping them would touch the API, the frontend and release search. - **Language-aware metadata search** for all providers. The other providers ignore Default Book Languages in metadata search. The `user_id` added here could serve them too. |
||
|
|
2007384058 |
feat(sources): add Webshare.cz as a release source (#1543)
## What Adds [Webshare.cz](https://webshare.cz), the Czech file host, as a release source for ebooks and audiobooks. It follows the Libgen source: disabled by default, with a settings tab for credentials and a **Test Login** button that reports whether the account is VIP. ## Why Webshare is one of the main places Czech and Slovak ebooks and audiobooks are shared, and none of the existing sources reach it. Together with the Databáze knih metadata provider (#1479), this lets Universal mode find Czech and Slovak titles end to end. ## How - **API client** (`webshare/api.py`): Webshare's XML API (`POST /api/<endpoint>/`). Login is the salted exchange the API documents: fetch the salt, send `SHA1(MD5_CRYPT(password, salt))` plus an MD5 digest. Python dropped the `crypt` module in 3.13, so md5-crypt is a small pure-Python port, checked against `openssl passwd -1`. The session token is cached per account and refreshed once when an authenticated call fails. File-level errors (`FILE_LINK_FATAL_*`) are not retried. XML is parsed with `defusedxml`. - **Search** (`webshare/source.py`): Webshare matches file names only, so the file name becomes the release title and its extension the format. Queries are title + author, then title only, or the manual query. Results are filtered to the configured `SUPPORTED_FORMATS` / `SUPPORTED_AUDIOBOOK_FORMATS`, plus `zip`/`rar` for audiobooks, since they are usually uploaded as archives. Password-protected files are skipped. The language is guessed from tags in the name (`CZ`, `SK`, `ENG`, ...), and the vote balance is a sortable column. API failures raise `SourceUnavailableError`. - **Download** (`webshare/handler.py`): resolves a time-limited link per attempt with `file_link`, then streams it into `TMP_DIR` with progress and cancel support. Large audiobook archives are not buffered in memory. - **Settings**: `WEBSHARE_ENABLED`, `WEBSHARE_USERNAME`, `WEBSHARE_PASSWORD` (a `PasswordField`, already covered by the `_PASSWORD` redaction in `genDebug.sh`), and `WEBSHARE_MAX_RESULTS`. The `docs/environment-variables.md` section was added. ## Testing - `tests/webshare/`: 31 offline tests. They cover md5-crypt vectors, password hashing, XML parsing, token reuse and re-login, non-retried file errors, availability gating, format filtering, query fallback, release mapping, language guessing, and the handler's streaming, small-file rejection, link errors and cancel. - `ruff check` and `ruff format --check` are clean. The rest of the unit suite fails the same way with and without this change: Windows-only tests on my machine. - Verified live against webshare.cz: search parsing and the salt endpoint's error path. - Still a draft: I haven't yet verified login and download with a real account. 🤖 Generated with [Claude Code](https://claude.com/claude-code) --------- Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: CaliBrain <calibrain@l4n.xyz> |
||
|
48ae3dbef4 |
feat: optional Hardcover wishlist sync with automated downloading (1 of 5) (#1516)
This replaces #1047, which @InfiniteAvenger opened and calibrain offered to re-review. It is Caleb's feature commit with his authorship kept, rebased onto current main, with small fixes, tests and a docs page on top. Everything is opt-in and off by default, so nothing changes for anyone who doesn't turn it on. What it adds is a **Settings, Hardcover Sync** tab with three parts that can be switched on separately: - **Shelf sync**: a Hardcover shelf (Want to Read by default) is pulled into requests on a schedule, or with a Sync now button. Requests carry the Hardcover metadata and cover. - **Automatic downloads**: synced requests are approved and a release is picked from a source priority list. A strict title, author and format match (plus a seeder minimum for torrents) decides, so anything without a confident match stays pending for manual review. - **Skip books you already own**: sync and automatic downloads ask whichever libraries are enabled under Library Check (Calibre today) and skip a book that is already there. On top of Caleb's commit I made the request owner a lookup instead of assuming user id 1, sent synced cover URLs through the proxy when the request is created, got ruff and basedpyright clean, and added tests and `docs/hardcover-sync.md`. His commit also had its own Audiobookshelf settings and a Test library connection button, written before the library check was split into providers. The button would have raised a TypeError against the current `test_connection`, so those settings are gone from this tab and an Audiobookshelf provider comes as a separate PR, which is the part of #1047 that is missing here (for now anyways...). I've run this daily for about a couple of weeks. Ruff, basedpyright and the backend suite pass (4426 passed). This is 1 of 5. As calibrain suggested, the rest are stacked on this one, each saying which number it is. Closes https://github.com/calibrain/shelfmark/pull/1047#issuecomment-6302071588 --------- Co-authored-by: InfiniteAvenger <calebewest02@gmail.com> Co-authored-by: CaliBrain <calibrain@l4n.xyz> |
||
|
splitsec2andCaliBrain
|
72b5d04939 |
perf: cut repeated work on the status and activity paths (#1545)
Two small cuts to work that repeats on every poll or request. The delivery-state sync behind /api/status and the activity snapshot listed every fulfilled request (SELECT * plus a JSON parse of book_data and release_data per row) before it looked at the queue. That history is never purged, so an idle instance paid for it on every poll. It now returns straight away when no queue bucket holds a task, since nothing could map to a request then. Results and emitted updates are unchanged. The activity snapshot gains less from this, because its status merges download history rows, so it is usually not empty. In proxy auth mode the before-request hook looked the user up by name and by session id on every API call, cover images included, each on a new SQLite connection. UserDB.get_user_cached answers repeat lookups from memory for up to five seconds, bounded at 256 entries. create_user, update_user and delete_user drop the cache as soon as they commit, and a lookup that overlaps a write is not stored, so a role change or deletion made in the app applies on the next request. Changes made outside that UserDB instance (another process, or the separate instances some helpers open) show within the five seconds. Only the middleware uses the cached path; every other get_user caller is unchanged. Each change has a test that failed before it: one counts list_requests calls for an empty queue, the others count get_user calls across repeated lookups and middleware runs, check expiry with a patched clock, and check that writes and caller mutations do not leave a stale row behind. The full suite passes (4432 passed, 48 skipped, e2e excluded), as do ruff and basedpyright on the changed files. I also looked at throttling the full status_update broadcast when only the progress message changes during torrent and usenet polling, and left it out. Those handlers poll every 2 seconds and DOWNLOAD_PROGRESS_UPDATE_INTERVAL defaults to 1 second, so a throttle on that interval would skip nothing at the default setting. A longer window would slow the speed and ETA text in the UI, and without a trailing flush it would hide the last message of a quick burst, such as "Waiting for bot response" arriving just after "Requesting file from bot", for the whole wait that follows. --------- Co-authored-by: CaliBrain <calibrain@l4n.xyz> |
||
|
splitsec2andCaliBrain
|
237ddd53a2 |
fix(redaction): a non-admin's status message loses the links it quotes (#1541)
#1422 hid `source_id` and the download path from non-admins but left the status message alone. A failed add stores the client's exception text there, for example `401 Client Error: Unauthorized for url: <full url>`, and for a requests error that URL carries the path and query: an indexer key, a tracker passkey, a peer name. It reached the owner through `/api/status`, `/api/activity` and the `status_update` socket event. `redact_download_payload` now replaces every link in `status_message` with `(link removed)`. That covers the live queue, the socket push and the history snapshot, since all three go through it. Messages without a link are unchanged and admins still see the full text. Tests cover the payload, a whole status snapshot, and messages that must stay as they are. They fail on main and pass here. Found in a Claude pass over 1.5.0. --------- Co-authored-by: CaliBrain <calibrain@l4n.xyz> |
||
|
splitsec2
|
a430341a93 |
perf: cut repeated work on search and cover paths (#1544)
I went through what Shelfmark does per search and per cover and measured the pieces that showed up. Four of them are cheap to fix, so this is one PR. Shelfmark runs on a single gevent worker, so any of these blocks every other request while it runs, and I'm assuming a lot of people run it on small hardware (a Pi, say). Cover cache. Every cache hit rewrote the whole `cache_index.json` just to bump the entry's last-access time. On my box that is 3.7 ms of CPU per hit at 500 entries and about 100 ms at 15,000 (2.2 MB written each time), and a results page asks for around 40 covers. The access time now stays in memory and the index is written at most every 30 seconds, or whenever something else saves it. A crash inside that window loses some access times, which only changes eviction order. Settings reload. `config.refresh()` rebuilt the cache by reading and parsing a tab's JSON file once per field, so about 250 parses per refresh, 100 ms of CPU, and the metadata search, providers and config routes each trigger it. It now reads each file once per rebuild (several tabs share `settings.json`, so the memo is keyed on the file path). Library match. When the library check is on, each search result is compared against every library entry, and the title was split into words again for every entry. The words are now computed once per result. I measured 33 ms per result at 5,000 library entries and 106 ms at 20,000, which is a second or more per page of 40 results. The matching itself is unchanged. Cover URL probe. `is_covers_cache_enabled()` creates and deletes a file in the config directory to check it is writable, and it runs for every cover, queue task and history row. The answer is remembered for 60 seconds. The probe is 0.045 ms on its own, but it is called a few hundred times per request in the worst case. Each change has a test that fails without it (call counts on the index saves, file opens during a refresh, tokenisation calls, and probe calls). Ruff, basedpyright on the changed files and the non-e2e suite pass. The timings are from a dev machine with an empty config, so the real numbers depend on cache and library size. |
||
|
splitsec2
|
2668705bfa |
perf: reuse connections and cache Anna's Archive download counts (#1546)
Two small changes that stop Shelfmark from opening a new HTTPS connection for requests that go to the same host over and over. Anna's Archive download counts. Every search result gets its count from `/dyn/md5/inline_info/<md5>`, and each of those was a bare `requests.get`, so a search with 50 results opened 50 TCP and TLS connections to the mirror. `search_books` also runs once per title variant, and the variants of one book mostly find the same files, so the same md5 was looked up again for every variant that found it, and again when the release list was reopened. The lookups now share one `requests.Session`, and a successful count is kept for five minutes in a bounded `CacheService` (the same class the metadata cache uses, as its own instance, so it does not depend on the metadata cache toggle). A lookup that fails or comes back without a count is not cached, so it is retried on the next search just like today. The URL, headers and 5 second timeout are unchanged. Debrid clients. Real-Debrid, AllDebrid, Debrid-Link and TorBox also used the bare `requests` helpers. The download handler polls `get_status` every two seconds for each active job, so every poll was a fresh TLS handshake with the debrid API. Each of those modules now keeps one `requests.Session` and sends its API calls through it, the way the qBittorrent, Deluge, Prowlarr and Newznab clients already do. Headers, timeouts, `verify` and error handling are still passed per call exactly as before. The existing tests that patched `<module>.requests.get/post/request` now patch `<module>._session.<verb>` instead; nothing else in them changed. How I checked it: each change has a test that fails on main and passes here. They count the requests that reach `requests.Session.request` and the sessions built while doing it, with no network. On main, three polls of each debrid client build three sessions, and two overlapping batches of AA results look `bbb` up twice. There are also tests for the count cache expiring (with a faked clock) and for a failed lookup not being cached. Full suite: 4432 passed, 48 skipped. ruff and basedpyright are clean on the changed files. I also looked at caching `/api/releases` on the server and left it out. The release modal already caches each book's results per source and content type in the browser for five minutes (longer for IRC and slskd through `cache_ttl_seconds`), and that cache survives closing and reopening the modal, so reopening it does not run a new search today. A server cache would only help after a page reload, and it would quietly undo the modal's "apply filters" and manual search refetches, which drop the browser cache on purpose, unless the frontend also learned to send a bypass. That did not seem worth it for this PR. |
||
|
splitsec2
|
08fdb7ea87 |
fix(startup): the reconcile guard catches the same errors the rest of main.py does (#1540)
The start-up reconcile runs at import, before the coordinator starts, inside a try that named five exception types and left out `TypeError`. `main.py` already defines `_IMPORT_OPERATIONAL_ERRORS` for this shape and the other module-level guards use it, so this switches to it. An unexpected `TypeError` from the reconcile would otherwise fail the import of `shelfmark.main`, and the worker would fail to boot on every restart. I didn't use `except Exception` because BLE001 is on and nothing in main.py does. The test reloads main with the reconcile raising `TypeError`, and with `OSError` as the control (that one already passes on main). Full non-e2e suite passes, ruff and basedpyright are clean. Found in a Claude pass over 1.5.0. |
||
|
splitsec2
|
f560b31ee4 |
fix(db): wait up to 30 s for a busy writer instead of SQLite's 5 s default (#1539)
The users, download history and activity view-state databases open their connections with SQLite's default busy timeout of 5 seconds. A start-up migration, the restart reconcile, or any statement that overlaps a backup or a second process raised "database is locked" after 5 seconds, and the reconcile finished only part of its work. One shared `DB_BUSY_TIMEOUT_SECONDS` (30) in `core/utils.py`, used by all three. The test reads SQLite's `busy_timeout` pragma back from each connection, so it doesn't need a real lock fight. I left `UserDB.initialize()` without a guard on purpose. With 30 seconds the wait covers a backup, and swallowing a schema failure there would hide a worker that can't run. Found in a Claude pass over 1.5.0. |
||
|
splitsec2
|
e27fed1d40 |
fix(queue): a late progress update no longer reopens a finished task (#1538)
#1429 made the first terminal status win, but the guard only refused terminal to terminal changes. A worker's poll that reports DOWNLOADING after a cancel still overwrote CANCELLED (the poll loop calls the status callback before it waits on the cancel flag), and the COMPLETE or ERROR that followed was then accepted as a first terminal status, so the terminal hook ran a second time. This refuses any status change away from a terminal status. `add()` and `enqueue_existing()` reset the status themselves, so retries are unaffected. The new test cancels a downloading task, sends DOWNLOADING and then COMPLETE, and expects CANCELLED and one hook call. It fails on main and passes here. Full non-e2e suite passes, ruff and basedpyright are clean. Found in a Claude pass over 1.5.0, same as my other small fixes from it. |
||
|
splitsec2
|
395c331b4a |
fix(search): give each provider its own slice of the search budget (#1537)
Fixes #1531. All Direct Download providers share the one search budget (`RELEASE_SEARCH_TIMEOUT`, 300s). Anna's Archive runs first, and when its bypass keeps failing the retries use up the whole budget. OceanofPDF then sees an expired deadline, skips its queries, and the search raises Anna's Archive's error even though OceanofPDF never got a turn. I didn't split the budget evenly, since the deadline module itself says a cold-start Anna's Archive solve can legitimately take 60 to 120s and a 150s or 37s slice would break setups that work today. Instead `search_deadline.unit_deadline(waiting=n)` opens a child deadline for one unit of work. It gets the parent's remaining time minus 30s for each unit still waiting, never less than 30s and never more than the parent has left. A single provider still gets everything, a fast provider's unused time rolls on to the next, and a stuck one stops at its own limit with the reserve intact. It's used in the provider loop in `DirectDownloadSource.search`, and in the source loop in `/api/releases` since it's the same shape and four lines. With several sources enabled the first one now gets 30s less per source behind it, which is still well above a cold solve for the usual two or three. Tests: a stuck first provider no longer starves the second (fails on main, passes here), a lone provider gets the full budget, and a child never outlives its parent. Two existing tests in `test_releases_api_budget.py` tripped the deadline in force from inside the source, which is now the child, so they use a short request budget instead. ruff, basedpyright on the changed files and the non-e2e suite are clean. OceanofPDF is inactive on my own instance (no mirror URL set), so I haven't reproduced this live. It's verified by reading the code and by the tests. I had Claude do the pass on this one. |
||
|
splitsec2
|
8e425da527 |
refactor: one default and one reader for the Supported Book Formats list (#1532)
The default for Supported Book Formats (`epub, mobi, azw3, fb2, djvu, cbz, cbr`) was written out in five places: the settings registry, `postprocess/policy.py`, `irc/parser.py`, Prowlarr's `source.py` (added in #1490) and slskd's `source.py` (added in #1428). Both of those two arrived in 1.5.0, and both are mine, so this is cleaning up after myself as much as anything. The setting was also read three different ways. Post-processing split on commas only. slskd split on commas and spaces and stripped a leading dot. IRC split on commas and kept the dot. Prowlarr split on commas and kept `""` as a format when the value was an empty string. So `.epub` or `epub mobi` worked in one source and not another. This puts `DEFAULT_EBOOK_FORMATS` and `normalize_format_list` in `core/utils.py` next to `AUDIOBOOK_FORMATS`, and the registry, post-processing, IRC, Prowlarr and slskd all use them. I took the most lenient reading of the three (comma or space split, leading dot dropped, empties ignored, default when nothing is left), so no existing config gets stricter. I put the reader in `core/utils.py` and not in `policy.py` because the release sources can't import `policy` without pulling in the whole post-processing package. A config with a sensible list reads exactly as before. The only differences are for malformed values: an emptied setting or an empty string now falls back to the default in every place, where IRC used to filter everything out and Prowlarr treated `""` as a format. The audiobook default is a separate story and I left it alone: the registry and slskd default to every processable audiobook format plus the archives, while post-processing and IRC fall back to `m4b, mp3`. They only differ when the setting is unset or emptied, but they do differ, and I'd rather ask which one you want than pick one here. Ruff, basedpyright on the changed files and the backend suite pass (4400 passed, e2e not run). New tests cover the reader and the two fallbacks. |
||
 Amalkrishna A SandCaliBrain
|
a11a63a29e |
feat(notifications): add {User}/{Email} placeholders to per-user routes (#1499)
Closes #1130. Per-user notification route URLs can now use `{Email}` and `{User}`, filled in from each user's account at send time, the same way `{User}` already works in destination paths. One default route like `mailtos://user:pass@smtp.example.com?to={Email}` now reaches every user at their own address, with no per-user setup. - Values are URL-encoded (`alice+books@example.com` keeps its `+`); placeholders are case-insensitive. - A route that needs a value the user doesn't have (e.g. no email on file) is skipped for that user; other routes still send. - The user lookup only runs when a route contains a placeholder. - The personal "Test notification" button expands the placeholders too, so testing a `{Email}` route works. - Added to the routes setting help, `docs/users-and-requests.md`, and tests. @calibrain ready for review. --------- Co-authored-by: CaliBrain <calibrain@l4n.xyz> |