A search entry often carries no `format`, so `format_of(None)` produced
`Other("")`, that was stashed at search time and handed back as the
download's format, and conversion had no parser to pick:
4073669 could not be converted from to SRT: no parser for ""
`POST /download` takes `sub_format`, and the API converts on its side, so
asking for srt makes the answer srt whatever the uploader posted. The
format guessed from the search entry is dead weight and goes with it.
The download mock matched only `file_id`, so it never exercised what the
real API returns.
Closes#272
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
`login` posted the credentials as HTTP basic auth with no request body.
The API takes them as JSON, answers basic auth with a 401, and the lane
reported that back as "opensubtitles rejected the configured
credentials" — for correct credentials.
Search was unaffected and hid this: it authenticates with the Api-Key
header alone and never logs in, so candidates were found and only the
download failed.
`mount_login` matched on method and path, so the mock answered a token
to any request shape. It now matches the body.
Closes#271
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The id lane sent `tmdb_movie_id` and `tmdb_series_id`. Neither is a
parameter of this API, and unknown ones are ignored rather than refused,
so every search silently degraded to a `moviehash` lookup: one result for
a release someone had already hashed, none at all for anything else.
Two further requirements the API documents and answers a 301 to when
missed: parameters sorted by name, and language codes lowercase and
sorted.
The integration tests asserted the old names against a mock, which
answers whatever it is asked for. They pinned the bug rather than
catching it.
Closes#270
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A release whose only subtitle is a PGS or VobSub track was stuck both
ways: the track satisfied English so no English SRT was ever fetched,
and bitmaps can never feed a translator. §15 is amended to separate
satisfying viewing from providing a translation source, and to carve a
fetch made to obtain a source out of the no-upgrade rule.
Timings come from the disc rather than from alass guessing at the audio.
At import, each non-forced image track's packet timestamps are paired
show-to-clear into a cue skeleton and stored; the fetched source is then
aligned against that skeleton, and the translation made from it skips
the post-translation pass, which could only move disc-exact timings off.
Pairing is validated before it is trusted — even packet count, plausible
durations, sane density for the runtime — because PGS allows several
composition segments per subtitle and a slipped pairing is quietly half
a second out. A track that fails validation gets no skeleton and falls
back to aligning against the video, as does any file imported before
this: there is no backfill.
Closes#268
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The base URL and model move into a cell the backend re-reads per request,
so an operator can repoint it without a restart. Migration 0028 adds the
two columns; DESIGN.md §15 calls both database rows. Construction never
depends on the API key — llama.cpp serves without one.
Syncer::run passed [subtitle, video] where alass expects
<reference-file> <incorrect-sub-file>; every real invocation failed
before comparing timings, silently degrading to SyncState::NotRun.
#193 moved them out of openai.rs so the command backend could share them,
but left them ungated: with neither LLM backend compiled in they are dead
code. just ci missed it because it lints the workspace with --all-features;
building arr-api alone does not enable arr-subs' features.
Backend::characters_billed() polled a cumulative counter that races
under concurrent closes and can't attribute cost to one call. Drop it
in favour of the caller counting source characters before it sends
anything, which #197's budget needs anyway.
Folds an implausible result and an unusable alass binary into one
SyncState both the grab handler (#199) and reconcile loop (#196)
can record without re-deriving the same match arms.
Every translation backend sits behind a default-off cargo feature, and the
gate ran with the default set, so clippy and the test run never saw a line
of arr-subs' openai, deepl or google modules. #191 and #192 each reported
it after verifying their own work by hand. lint and test now pass
--all-features.
Also drops a duplicated wiremock suppression the #188 merge left behind and
puts the module list back in order.
Behind the existing translate-openai feature. One JSON-in/JSON-out
chat/completions request per batch; refusal, content-filter and
truncated replies fail loudly instead of validating as a short batch.
The Podnapisi zip fixtures never reached the branch: a global gitignore
excludes *.zip, so git add skipped them silently and the worker's CI passed
against untracked files. Recovered, with a fixtures .gitignore that keeps
the next binary fixture from vanishing the same way.
Two providers landing in parallel also disagreed on names and on which
crate dependencies each integration-test target uses. PodnapisiProvider is
now Podnapisi, matching OpenSubtitles, and both test targets declare the
dependencies they do not use so the per-target lint stays quiet.
Searches by moviehash computed from the media file and by TMDB id
with season/episode for TV, ranks candidates through
arr_core::subs::rank, and downloads under a lazily-fetched user
token. The daily download cap (429/406) surfaces as Error::RateLimited
so the loop can show a queue state. Credentials come from config or
environment only; tests run against wiremock fixtures.
The Backend trait plus everything the backends (#191-#193) share, so no
backend can skip it: chunking into character-budgeted batches, rejection
of replies whose cue count or numbering drifted, and reassembly onto the
original timings. Timing data never leaves arr; pt-PT and pt-BR are
distinct targets a backend must refuse rather than conflate.
Translation reassembles translated text onto original timings (§15), so
cues need a structured form. Parsing is tolerant of real files (CRLF,
BOM, missing indices, dot milliseconds); rendering is strict and
renumbers from 1.
#184 and #185 were built in parallel and their candidate types did not
meet. Ranking returned no identity for a candidate, so the winner of a
rank() could not be handed back to Provider::download -- rank() reorders,
so the caller could not recover it by position either.
RankedSubtitle now carries the index of the candidate in the slice it was
given, and Candidate::to_core is the one place the two shapes are mapped:
hash_match against a compared hash, sdh against hearing_impaired, group
against release_group, and the two optional tiebreakers defaulted to sort
last rather than block a candidate.
The skeleton the subtitles milestone hangs off: provider domain types, an
object-safe Provider trait, the crate's own error type, and the cargo
features the translation backends will sit behind. No provider, no
translation, no ranking — DESIGN.md §15 keeps ranking pure in arr-core.
