BoulderBadgeDad
3490092e3d
video search: each source returns its own results (was identical)
...
mock_search now takes the source and gives Soulseek / Torrent / Usenet distinct
hits — different release-group pools, seeder magnitudes, and which qualities show
up (Soulseek drops the top remux, Usenet drops the cam, etc.), the way real
indexers differ. Endpoint forwards body.source. Fixes 'click Torrent shows the
same results as Soulseek'. +1 test.
2026-06-19 13:56:04 -07:00
BoulderBadgeDad
44b074772b
video search: evaluate/rank pipeline + mock indexer + /downloads/search
...
Phase 2 of live search (indexer mocked, pipeline real):
- quality_eval.evaluate_release(parsed, profile, scope, want_season/episode, size_gb)
→ {accepted, score, rejected, tier, quality_label}. Sonarr-style: rejects junk
sources/codecs/3D, requires an ENABLED ladder tier, honours HDR-require + size caps,
and VALIDATES scope (episode wants SxxExx, season wants the whole-season PACK, show
wants a complete-series pack, with season/episode-number checks). Scores keepers by
resolution/source/codec-pref/HDR/audio/repack for ranking.
- mock_search.py: deterministic stand-in indexer returning scope-shaped raw hits
(the single swap-point for real slskd/Prowlarr).
- POST /downloads/search: mock → parse → evaluate → rank (accepted, score, seeders).
16 tests, isolation guard + ruff clean. UI wiring next.
2026-06-19 13:39:09 -07:00
