[pull] main from mhdzumair:main#269
Merged
Merged
Conversation
- store_external_id: add created_at = NOW() to INSERT so the NOT NULL column doesn't silently fail the insert via let _ = ... - link_to_catalogs: upsert the catalog row before linking so fresh installs with an empty catalog table don't produce 0-row results - discover_prewarm link_to_catalog: same upsert-then-link pattern Fixes #681 — native scrapers populate media/stream_media_link but media_external_id, catalog, and media_catalog_link stay empty.
- meta.rs get_media_meta: convert query_as! to non-macro query_as so LEFT JOIN LATERAL columns (poster_url, background_url, imdb_rating) are decoded as Option<T> at runtime instead of non-nullable as inferred by the offline sqlx cache - meta.rs get_episodes: same fix for thumbnail_url and fml.media_id - streams.rs fetch_stream_playback_info: add ts.torrent_file to the GROUP BY clause in the non-series path Fixes runtime errors: "meta query [tt...]: error decoding column 13: unexpected null" "episodes for media ...: error decoding column 5: unexpected null" "fetch_stream_playback_info: column 'ts.torrent_file' must appear in GROUP BY"
Previously fingerprint() used only file:line, so all warnings from the same source location were bucketed together regardless of message — the traceback stored the first message while the displayed message showed the latest, causing the mismatch visible in the admin exception panel. Changes: - fingerprint() now hashes file:line + normalized message prefix (digit runs collapsed to 'N', first 80 chars) so errors with genuinely different text get separate entries; numeric variations like page numbers still cluster together - store_event update path now also refreshes traceback to stay consistent with the updated message field
- response_json: when JSON parse fails and HTTP status is 401/403, return invalid_token.mp4 instead of api_error.mp4 — covers the case where a proxy/CDN returns an HTML 401 page instead of the API's JSON error body - realdebrid check_cached: replace hand-rolled HTTP loop with get_torrent_list() so the request goes through rd_get → response_json → check_rd_error, giving consistent auth-error handling - premiumize check_pm_error: detect "Not logged in" / "Invalid API key" / "unauthorized" messages and map them to invalid_token.mp4 instead of transfer_error.mp4
- contributions.rs + import_helpers.rs: bind initial_status / status as ContributionStatus enum instead of &str — Postgres rejects the text → contributionstatus cast at runtime - playback_dedup.rs: downgrade stale-lock reclaim log from warn to debug; this is expected crash-recovery behavior, not an error condition - encrypt.rs: downgrade AES padding error from warn to debug; malformed tokens are expected user-input noise, fully handled with a 400 response
Four tests that protect against the runtime bugs fixed in this batch:
- get_media_meta_null_poster_and_rating_external_id: movie with no
poster/background/rating via tt_ lookup must not panic on null columns
- get_media_meta_null_poster_and_rating_internal_id: same via mf{id} path
- get_episodes_null_thumbnail_and_null_file_link: series with no
episode_image/file_media_link must return episodes with None fields
- fetch_stream_playback_info_movie_group_by_is_complete: torrent stream
with non-NULL torrent_file exercises the GROUP BY path that was missing
ts.torrent_file
… token auth - Add ENABLE_PROMETHEUS_METRICS (default false) so the /api/v1/metrics endpoint is opt-in; disabled instances return 404 (route not registered). - Add PROMETHEUS_METRICS_TOKEN: when set, every scrape request must carry "Authorization: Bearer <token>" — enforced even on public instances. - Update prometheus.yml, prometheus-dev.yml, and docker-compose-perf.yml to use the new metrics path and Authorization bearer credential; replace the defunct PROMETHEUS_DISABLE_AUTH flag with ENABLE_PROMETHEUS_METRICS. fix(sports): store sports categories as catalogs only, not genres - Remove genre_name param from find_or_create_sports_stub and drop all link_genre() calls inside it — sports sub-categories (Football, Formula Racing, etc.) should live in media_catalog_link, not the genre table. - Update all four callers (dmm_hashlist, sport_video, ext_to, user_library) to drop the now-removed argument. - Remove the genre field from CatalogSpec in ext_to.rs (was only used to feed the deleted parameter). - Remove sports_category_to_genre from the parser re-export (function still exists in sports_parser.rs but has no external callers).
Serve debrid watchlist catalogs natively (hash lookup, delete-all meta/stream), fix Seedr listing via root folder names, align Prowlarr credential resolution with Python, and downgrade expected auth/rate-limit logs to debug.
Rebuild genre lists with parallel table scans instead of a heavy SQL join, filter blocked genres via the admin keyword cache at manifest time, and use shared provider short names (including Seedr as SDR).
…board
The admin request-metrics dashboard was always empty because
metrics_middleware only updated in-memory counters and never wrote to
Redis, while the admin readers expected Redis-backed data.
- metrics_middleware: after each non-health request, spawn a background
task that writes aggregate stats to req_metrics:agg:{METHOD}:{route}
hashes (HINCRBY/HINCRBYFLOAT for atomic increments), tracks endpoints
in a req_metrics:endpoints sorted set, and pushes individual request
JSON to a req_metrics:recent list (capped at 10k, TTL 1h)
- admin_extended: rewrite all five request-metrics handlers to read from
the correct Redis schema (ZRANGE sorted set + HGETALL hashes for
endpoint stats; LRANGE list for recent requests) instead of the old
dead LIST+string schema that was never populated
Meta requests now create missing media on miss; stream requests enqueue background re-scrapes that run the full torrent+usenet fan-out with per-indexer parallel Prowlarr/Jackett tasks and dual Zilean endpoints.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
See Commits and Changes for more details.
Created by
pull[bot] (v2.0.0-alpha.4)
Can you help keep this open source service alive? 💖 Please sponsor : )