diff options
| author | Christophe Besson <cbesson@gmail.com> | 2026-08-26 01:08:58 +0200 |
|---|---|---|
| committer | Christophe Besson <cbesson@gmail.com> | 2026-08-26 01:08:58 +0200 |
| commit | 2d144d76cee55cf8faaacf196e716a0930dfd7e9 (patch) | |
| tree | 1220d39cb30ccc8a58a80f85ba020b5b30bea9b7 /packages/meshbay-node | |
| parent | 704cfe37506fc5316c997b025004fbc9c4d2a47b (diff) | |
| download | meshbay-2d144d76cee55cf8faaacf196e716a0930dfd7e9.tar.gz | |
fix(node,hub): key music/media metadata lookups by file_id, not path
IndexEntry.path is the *folder* a file is in (indexer.py's
_virtual_dir docstring: "the directory a file appears in"), not the
file itself. GroupIndex.get_entry_by_path() treated it as if it named
one file, and every one of its four callers did too:
_do_music_meta_request, _do_media_meta_request, _do_tmdb_override, and
_admin_exec_tmdb_override. Any two files sharing a folder — an album is
one folder with many tracks, a season is one folder with many episodes
— collided: a lookup by path silently returned whichever entry the
index happened to iterate to first, regardless of which file the
client actually asked about.
Found live (2026-08-25): three unrelated albums ("High Tone - Various",
two "Le Peuple de l'Herbe" albums) all showed the same MusicBrainz
cover, because all their representative tracks happened to sit in one
"high_tone" folder alongside a track that legitimately matched that
cover. A force-reload didn't help — the bug is server-side, not a
stale client state.
Fixed by keying these four request/response pairs by `file_id` (the
entry's own content hash — already unique, already how every other
lookup in the system identifies a file) instead of `path`, both in the
wire messages (music_meta_req/resp, media_meta_req/resp, tmdb_override)
and in music-app.js/video-app.js's own hooks. GroupIndex.get_entry_by_path
is now unused and removed — GroupIndex.get_entry(file_id) already did
the right thing.
No test previously exercised either handler with two entries sharing a
folder — the only existing coverage (test_tmdb_override_policy.py) gave
each entry its own folder, so the bug never had a chance to show up.
Added that scenario there and in two new test files, all confirmed
failing against the pre-fix code before being confirmed green against
the fix.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013XSohfUQQiaE77qyFLgSv3
Diffstat (limited to 'packages/meshbay-node')
5 files changed, 362 insertions, 62 deletions
diff --git a/packages/meshbay-node/src/meshbay_node/indexer/group_index.py b/packages/meshbay-node/src/meshbay_node/indexer/group_index.py index 1ce4e0a..25081ef 100644 --- a/packages/meshbay-node/src/meshbay_node/indexer/group_index.py +++ b/packages/meshbay-node/src/meshbay_node/indexer/group_index.py @@ -78,19 +78,6 @@ class GroupIndex: def get_entry(self, file_id: str) -> IndexEntry | None: return self._entries.get(file_id) - def get_entry_by_path(self, path: str) -> IndexEntry | None: - """ - Linear scan — entries are keyed by content id, not path, and nothing - before the Videos app needed to go the other way (a client always - already has the id from index_sync/index_delta). Fine for an - on-demand, per-tile lookup against a few thousand entries; revisit - if a future caller makes this hot. - """ - for entry in self._entries.values(): - if entry.path == path: - return entry - return None - @property def entries(self) -> list[IndexEntry]: return list(self._entries.values()) diff --git a/packages/meshbay-node/src/meshbay_node/transport/webrtc_server.py b/packages/meshbay-node/src/meshbay_node/transport/webrtc_server.py index b4db051..6709fbc 100644 --- a/packages/meshbay-node/src/meshbay_node/transport/webrtc_server.py +++ b/packages/meshbay-node/src/meshbay_node/transport/webrtc_server.py @@ -2759,21 +2759,30 @@ class WebRTCPeerSession: async def _do_music_meta_request(self, msg: dict) -> None: """ - docs/musicbay.md §4.3: MusicBrainz metadata for one path, resolved - from the group's index. Album-level (release), the direct analogue - of Videos' show-level TMDB caching: one search per (artist, album) - pair serves cover art and canonical naming to every track of the - same release, keyed off the `artist`/`album` fields enrich_audio.py - already populated at index time (from tags, or the filename-parse - fallback) — never re-parsed here. + docs/musicbay.md §4.3: MusicBrainz metadata for one track, resolved + from the group's index by its content id. Album-level (release), the + direct analogue of Videos' show-level TMDB caching: one search per + (artist, album) pair serves cover art and canonical naming to every + track of the same release, keyed off the `artist`/`album` fields + enrich_audio.py already populated at index time (from tags, or the + filename-parse fallback) — never re-parsed here. + + Keyed by `file_id` (the entry's own content hash), not `path`: found + live (2026-08-25) — `IndexEntry.path` is the *folder* a file is in + (indexer.py's `_virtual_dir`), so any two tracks in the same folder + (routinely true — an album is one folder, many tracks) shared the + same `.path`, and looking a track up by it silently resolved to + whichever entry happened to be first in the index. Three unrelated + albums showed the same wrong cover before this fix, all sharing one + folder with the track that legitimately matched it. """ - path = msg.get("path") - log.debug("music_meta_req path=%r", path) - if not isinstance(path, str) or not path: - self._send({"type": "error", "detail": "Missing path"}) + file_id = msg.get("file_id") + log.debug("music_meta_req file_id=%r", file_id) + if not isinstance(file_id, str) or not file_id: + self._send({"type": "error", "detail": "Missing file_id"}) return ctx = self._group_ctx() - entry = ctx["index"].get_entry_by_path(path) + entry = ctx["index"].get_entry(file_id) if not entry: self._send({"type": "error", "detail": "File not found"}) return @@ -2789,7 +2798,7 @@ class WebRTCPeerSession: or not ctx.get("musicbrainz_enabled", True) or not entry.artist or not entry.album): self._send({"type": MNP.MUSIC_META_RESP, "v": MNP_VERSION, - "path": path, "confidence": 0}) + "file_id": file_id, "confidence": 0}) return mbid = await media_cache.get_file_mbid(entry.id) @@ -2799,7 +2808,7 @@ class WebRTCPeerSession: result, ratio = await musicbrainz_client.search_release(entry.artist, entry.album) if result is None or ratio < 0.6: self._send({"type": MNP.MUSIC_META_RESP, "v": MNP_VERSION, - "path": path, "confidence": 0}) + "file_id": file_id, "confidence": 0}) return mbid = result.get("id") artist_credit = result.get("artist-credit") or [] @@ -2814,11 +2823,11 @@ class WebRTCPeerSession: cover_thumb_hash = await self._fetch_and_cache_cover( media_cache, musicbrainz_client, mbid) - log.debug("music_meta_req path=%r: replying mbid=%s cover=%s", - path, mbid, cover_thumb_hash) + log.debug("music_meta_req file_id=%r: replying mbid=%s cover=%s", + file_id, mbid, cover_thumb_hash) self._send({ - "type": MNP.MUSIC_META_RESP, "v": MNP_VERSION, "path": path, + "type": MNP.MUSIC_META_RESP, "v": MNP_VERSION, "file_id": file_id, "mbid": mbid, "artist": meta.get("artist"), "album": meta.get("album"), @@ -2830,17 +2839,25 @@ class WebRTCPeerSession: async def _do_media_meta_request(self, msg: dict) -> None: """ - docs/mediacenter.md §5.4: TMDB metadata for one path, resolved from - the group's index (root+relpath the client already knows from - index_sync/index_delta — never a raw filesystem path off the wire). + docs/mediacenter.md §5.4: TMDB metadata for one file, resolved from + the group's index by its content id (root+relpath the client already + knows from index_sync/index_delta identify the entry; its own `id` + is what actually names one file — never a raw filesystem path off + the wire). + + Keyed by `file_id`, not `path`: `IndexEntry.path` is the *folder* a + file is in (indexer.py's `_virtual_dir`), so two files in the same + folder — any multi-episode season, routinely — shared the same + `.path`, and a lookup by it could silently resolve to the wrong + entry (found live via the Music app's identical bug, 2026-08-25). """ - path = msg.get("path") - log.debug("media_meta_req path=%r", path) - if not isinstance(path, str) or not path: - self._send({"type": "error", "detail": "Missing path"}) + file_id = msg.get("file_id") + log.debug("media_meta_req file_id=%r", file_id) + if not isinstance(file_id, str) or not file_id: + self._send({"type": "error", "detail": "Missing file_id"}) return ctx = self._group_ctx() - entry = ctx["index"].get_entry_by_path(path) + entry = ctx["index"].get_entry(file_id) if not entry: self._send({"type": "error", "detail": "File not found"}) return @@ -2853,7 +2870,7 @@ class WebRTCPeerSession: # TMDB match" as the ordinary case. if media_cache is None or tmdb_client is None or not ctx.get("tmdb_enabled", True): self._send({"type": MNP.MEDIA_META_RESP, "v": MNP_VERSION, - "path": path, "confidence": 0}) + "file_id": file_id, "confidence": 0}) return is_show = entry.season is not None and entry.episode is not None @@ -2870,7 +2887,7 @@ class WebRTCPeerSession: result, ratio = await self._tmdb_search(tmdb_client, entry, is_show) if result is None or ratio < 0.6: self._send({"type": MNP.MEDIA_META_RESP, "v": MNP_VERSION, - "path": path, "confidence": 0}) + "file_id": file_id, "confidence": 0}) return tmdb_id = str(result["id"]) meta = await self._tmdb_build_meta(tmdb_client, tmdb_id, media_type, result) @@ -2881,11 +2898,11 @@ class WebRTCPeerSession: media_cache, tmdb_client, meta.get("poster_path")) backdrop_thumb_hash = await self._fetch_and_cache_poster( media_cache, tmdb_client, meta.get("backdrop_path")) - log.debug("media_meta_req path=%r: replying tmdb_id=%s poster=%s backdrop=%s", - path, tmdb_id, poster_thumb_hash, backdrop_thumb_hash) + log.debug("media_meta_req file_id=%r: replying tmdb_id=%s poster=%s backdrop=%s", + file_id, tmdb_id, poster_thumb_hash, backdrop_thumb_hash) resp = { - "type": MNP.MEDIA_META_RESP, "v": MNP_VERSION, "path": path, + "type": MNP.MEDIA_META_RESP, "v": MNP_VERSION, "file_id": file_id, "tmdb_id": tmdb_id, "title": meta.get("title"), "original_title": meta.get("original_title"), "overview": meta.get("overview"), @@ -3013,25 +3030,28 @@ class WebRTCPeerSession: (§3.4/§V6) — not just the one file the operator happened to be looking at, so the correction actually sticks regardless of which episode a future render picks as representative. + + Keyed by `file_id`, not `path` — see `_do_media_meta_request`'s + docstring for why a folder-level path cannot name one file. """ - path = msg.get("path") + file_id = msg.get("file_id") tmdb_id = msg.get("tmdb_id") media_type = msg.get("media_type") - if not isinstance(path, str) or not path: - self._send({"type": "error", "detail": "Missing path"}) + if not isinstance(file_id, str) or not file_id: + self._send({"type": "error", "detail": "Missing file_id"}) return if not isinstance(tmdb_id, str) or not tmdb_id or media_type not in ("movie", "tv"): self._send({"type": "error", "detail": "Missing tmdb_id or media_type"}) return ctx = self._group_ctx() - entry = ctx["index"].get_entry_by_path(path) + entry = ctx["index"].get_entry(file_id) if not entry: self._send({"type": "error", "detail": "File not found"}) return if not self._has_admin_authority(): self._send({"type": "error", "detail": "No authorized key for this"}) return - subject = f"path={path},tmdb_id={tmdb_id},media_type={media_type}" + subject = f"file_id={file_id},tmdb_id={tmdb_id},media_type={media_type}" self._issue_admin_challenge(OP_TMDB_OVERRIDE, subject) async def _admin_exec_tmdb_override( @@ -3043,10 +3063,10 @@ class WebRTCPeerSession: self._audit("admin_auth_failed", f"tmdb_override:{subject}") return fields = dict(part.split("=", 1) for part in subject.split(",")) - path, tmdb_id, media_type = fields["path"], fields["tmdb_id"], fields["media_type"] + file_id, tmdb_id, media_type = fields["file_id"], fields["tmdb_id"], fields["media_type"] ctx = self._group_ctx() - entry = ctx["index"].get_entry_by_path(path) + entry = ctx["index"].get_entry(file_id) media_cache = self._ctx.get("media_cache") if entry is None or media_cache is None: self._send({"type": "error", "detail": "File or media cache not available"}) @@ -3058,7 +3078,7 @@ class WebRTCPeerSession: await media_cache.set_file_tmdb(e.id, tmdb_id, media_type) self._audit("tmdb_override", subject) - notice = {"type": MNP.TMDB_OVERRIDE_ACK, "v": MNP_VERSION, "path": path, + notice = {"type": MNP.TMDB_OVERRIDE_ACK, "v": MNP_VERSION, "file_id": file_id, "tmdb_id": tmdb_id, "media_type": media_type} for uid, session in list(self._peer_registry().items()): try: diff --git a/packages/meshbay-node/tests/test_media_meta_request.py b/packages/meshbay-node/tests/test_media_meta_request.py new file mode 100644 index 0000000..a7355e8 --- /dev/null +++ b/packages/meshbay-node/tests/test_media_meta_request.py @@ -0,0 +1,130 @@ +""" +`_do_media_meta_request`, keyed by `file_id` (2026-08-25 fix) — same +regression as test_music_meta_request.py, one app over: `IndexEntry.path` +is the *folder* a file is in, not the file itself, so a lookup by path +alone (the pre-fix `GroupIndex.get_entry_by_path`) silently resolved to +whichever entry the index happened to return first for that folder — a +real risk here too, since a season folder routinely holds many episodes. +""" + +import pytest +from cryptography.hazmat.primitives.asymmetric.ed25519 import Ed25519PrivateKey +from meshbay_common.protocol import IndexEntry +from meshbay_node.indexer.group_index import GroupIndex +from meshbay_node.media_cache import MediaCache +from meshbay_node.transport.webrtc_server import WebRTCPeerSession + +pytestmark = pytest.mark.asyncio + + +class FakeTmdbClient: + """Returns a distinct, deterministic match per entry — real enough to + prove the server searched using the *right* entry's own fields.""" + + def __init__(self): + self.searched = [] + + async def search_movie(self, title): + self.searched.append(("movie", title)) + return {"id": 1000 + len(self.searched), "title": title, + "release_date": "2001-01-01"}, 1.0 + + async def search_tv(self, title): + self.searched.append(("tv", title)) + return {"id": 2000 + len(self.searched), "name": title, + "first_air_date": "2001-01-01"}, 1.0 + + async def fetch_image(self, url): + return f"image-bytes-for-{url}".encode() + + @staticmethod + def poster_url(path): + return f"https://image.tmdb.org/t/p/w500{path}" + + +def _entry(path: str, name: str, file_id: str, display_title: str) -> IndexEntry: + return IndexEntry( + id=file_id, name=name, path=path, size=1, type="video", added_at=0, + display_title=display_title, + ) + + +@pytest.fixture +async def media_cache(tmp_path): + c = MediaCache(db_path=tmp_path / "media_cache.db") + await c.open() + yield c + await c.close() + + +def _session(index, media_cache, tmdb_client): + session = WebRTCPeerSession.__new__(WebRTCPeerSession) + session._ctx = { + "index": index, + "media_cache": media_cache, + "tmdb_client": tmdb_client, + "tmdb_enabled": True, + } + session._group_id = None + session.sent = [] + session._send = session.sent.append + # _tmdb_search/_tmdb_build_meta are the real methods (not part of this + # regression) — stub the search ladder to a single direct call by + # display_title so this test is about routing, not TMDB matching. + async def _search(tmdb_client_, entry, is_show): + return await (tmdb_client_.search_tv(entry.display_title) if is_show + else tmdb_client_.search_movie(entry.display_title)) + session._tmdb_search = lambda *a: _search(*a) + + async def _build_meta(tmdb_client_, tmdb_id, media_type, result): + return { + "title": result.get("title") or result.get("name"), + "original_title": result.get("title") or result.get("name"), + "release_date": result.get("release_date"), + "first_air_date": result.get("first_air_date"), + "confidence": 1.0, + } + session._tmdb_build_meta = lambda *a: _build_meta(*a) + return session + + +async def test_two_episodes_in_the_same_season_folder_each_get_their_own_metadata(media_cache): + """The Videos-side analogue of the Music bug: two episodes share a + season folder, and each must resolve against its own entry — not + whichever one the index happens to return first for that folder.""" + index = GroupIndex(group_id="g" * 32, sk_node=Ed25519PrivateKey.generate()) + ep1 = _entry("shows/Show/Season 1", "s01e01.mkv", "id-1", "War of the Worlds") + ep2 = _entry("shows/Show/Season 1", "s01e02.mkv", "id-2", "A Different Show") + index.add_entry(ep1) + index.add_entry(ep2) + client = FakeTmdbClient() + session = _session(index, media_cache, client) + + await session._do_media_meta_request({"file_id": "id-1"}) + await session._do_media_meta_request({"file_id": "id-2"}) + + resp_1, resp_2 = session.sent + assert resp_1["file_id"] == "id-1" + assert resp_1["title"] == "War of the Worlds" + assert resp_2["file_id"] == "id-2" + assert resp_2["title"] == "A Different Show" + assert resp_1["tmdb_id"] != resp_2["tmdb_id"], ( + "two different shows sharing a season folder must not resolve to the same match") + + +async def test_missing_file_id_is_refused(media_cache): + index = GroupIndex(group_id="g" * 32, sk_node=Ed25519PrivateKey.generate()) + session = _session(index, media_cache, FakeTmdbClient()) + + await session._do_media_meta_request({}) + + assert session.sent == [{"type": "error", "detail": "Missing file_id"}] + + +async def test_unknown_file_id_is_refused(media_cache): + index = GroupIndex(group_id="g" * 32, sk_node=Ed25519PrivateKey.generate()) + session = _session(index, media_cache, FakeTmdbClient()) + + await session._do_media_meta_request({"file_id": "nope"}) + + assert session.sent == [{"type": "error", "detail": "File not found"}] diff --git a/packages/meshbay-node/tests/test_music_meta_request.py b/packages/meshbay-node/tests/test_music_meta_request.py new file mode 100644 index 0000000..96307f2 --- /dev/null +++ b/packages/meshbay-node/tests/test_music_meta_request.py @@ -0,0 +1,136 @@ +""" +`_do_music_meta_request`, keyed by `file_id` (2026-08-25 fix). + +Regression found live: `IndexEntry.path` is the *folder* a track is in +(indexer.py's `_virtual_dir`), not the track itself — an album is one +folder with many tracks in it, so looking a track up by `.path` alone +(the pre-fix behaviour, `GroupIndex.get_entry_by_path`) silently resolved +every track in that folder to whichever entry the index happened to +return first. Three unrelated albums showed the same wrong MusicBrainz +cover in production before this was found and fixed. +""" + +import pytest +from cryptography.hazmat.primitives.asymmetric.ed25519 import Ed25519PrivateKey +from meshbay_common.protocol import MNP, IndexEntry +from meshbay_node.indexer.group_index import GroupIndex +from meshbay_node.media_cache import MediaCache +from meshbay_node.transport.webrtc_server import WebRTCPeerSession + +pytestmark = pytest.mark.asyncio + + +class FakeMusicBrainzClient: + """Returns a distinct, deterministic match per (artist, album) pair — + real enough to prove the server routed to the *right* track's own tags, + not a stand-in for musicbrainz.py's own search-quality tests.""" + + def __init__(self): + self.calls = [] + + async def search_release(self, artist, album): + self.calls.append((artist, album)) + return ( + {"id": f"mbid-for-{artist}-{album}", "title": album, + "artist-credit": [{"name": artist}]}, + 1.0, + ) + + async def fetch_cover_art(self, mbid): + return f"cover-bytes-for-{mbid}".encode() + + +def _entry(path: str, name: str, file_id: str, artist: str, album: str) -> IndexEntry: + return IndexEntry( + id=file_id, name=name, path=path, size=1, type="audio", added_at=0, + artist=artist, album=album, display_title=name, + ) + + +@pytest.fixture +async def media_cache(tmp_path): + c = MediaCache(db_path=tmp_path / "media_cache.db") + await c.open() + yield c + await c.close() + + +def _session(index, media_cache, musicbrainz_client): + session = WebRTCPeerSession.__new__(WebRTCPeerSession) + session._ctx = { + "index": index, + "media_cache": media_cache, + "musicbrainz_client": musicbrainz_client, + "musicbrainz_enabled": True, + } + session._group_id = None + session.sent = [] + session._send = session.sent.append + return session + + +async def test_two_tracks_in_the_same_folder_each_get_their_own_metadata(media_cache): + """The exact production scenario: two tracks share a folder (an album), + with different artist/album tags of their own (one mistagged, sitting + in the wrong physical folder — a real, if messy, real-world case). Each + must resolve against its *own* tags, not whichever track the index + happens to return first for that shared folder path.""" + index = GroupIndex(group_id="g" * 32, sk_node=Ed25519PrivateKey.generate()) + track_a = _entry("music/high_tone", "a.mp3", "id-a", "High Tone", "Future Dub (1)") + track_b = _entry("music/high_tone", "b.mp3", "id-b", "Le Peuple de l'Herbe", "Triple Zero") + index.add_entry(track_a) + index.add_entry(track_b) + client = FakeMusicBrainzClient() + session = _session(index, media_cache, client) + + await session._do_music_meta_request({"file_id": "id-a"}) + await session._do_music_meta_request({"file_id": "id-b"}) + + resp_a, resp_b = session.sent + assert resp_a["file_id"] == "id-a" + assert resp_a["artist"] == "High Tone" + assert resp_a["album"] == "Future Dub (1)" + assert resp_b["file_id"] == "id-b" + assert resp_b["artist"] == "Le Peuple de l'Herbe" + assert resp_b["album"] == "Triple Zero" + assert resp_a["cover_thumb_hash"] != resp_b["cover_thumb_hash"], ( + "two different tracks sharing a folder must not end up with the same cover") + assert set(client.calls) == {("High Tone", "Future Dub (1)"), + ("Le Peuple de l'Herbe", "Triple Zero")} + + +async def test_missing_file_id_is_refused(media_cache): + index = GroupIndex(group_id="g" * 32, sk_node=Ed25519PrivateKey.generate()) + session = _session(index, media_cache, FakeMusicBrainzClient()) + + await session._do_music_meta_request({}) + + assert session.sent == [{"type": "error", "detail": "Missing file_id"}] + + +async def test_unknown_file_id_is_refused(media_cache): + index = GroupIndex(group_id="g" * 32, sk_node=Ed25519PrivateKey.generate()) + session = _session(index, media_cache, FakeMusicBrainzClient()) + + await session._do_music_meta_request({"file_id": "nope"}) + + assert session.sent == [{"type": "error", "detail": "File not found"}] + + +async def test_no_confidence_below_threshold_still_answers_by_file_id(media_cache): + index = GroupIndex(group_id="g" * 32, sk_node=Ed25519PrivateKey.generate()) + index.add_entry(_entry("music/x", "a.mp3", "id-a", "Some Artist", "Some Album")) + + class LowConfidenceClient(FakeMusicBrainzClient): + async def search_release(self, artist, album): + return {"id": "mbid", "title": album, + "artist-credit": [{"name": artist}]}, 0.1 + + session = _session(index, media_cache, LowConfidenceClient()) + + await session._do_music_meta_request({"file_id": "id-a"}) + + assert session.sent == [{ + "type": MNP.MUSIC_META_RESP, "v": session.sent[0]["v"], + "file_id": "id-a", "confidence": 0, + }] diff --git a/packages/meshbay-node/tests/test_tmdb_override_policy.py b/packages/meshbay-node/tests/test_tmdb_override_policy.py index b8fa6f9..cd1cee1 100644 --- a/packages/meshbay-node/tests/test_tmdb_override_policy.py +++ b/packages/meshbay-node/tests/test_tmdb_override_policy.py @@ -65,7 +65,7 @@ def _entry(path: str, name: str, display_title: str) -> IndexEntry: # ── Refused before a challenge is even issued ─────────────────────────────── -async def test_missing_path_is_refused(tmp_path): +async def test_missing_file_id_is_refused(tmp_path): session = _session(tmp_path, "op", operator="op") session._has_admin_authority = lambda: True issued = [] @@ -79,24 +79,25 @@ async def test_missing_path_is_refused(tmp_path): async def test_missing_tmdb_id_is_refused(tmp_path): session = _session(tmp_path, "op", operator="op") - session._ctx["index"].add_entry(_entry("shared", "ep.mkv", "Show")) + entry = _entry("shared", "ep.mkv", "Show") + session._ctx["index"].add_entry(entry) session._has_admin_authority = lambda: True issued = [] session._issue_admin_challenge = lambda op, subject: issued.append((op, subject)) - session._do_tmdb_override({"path": "shared", "media_type": "tv"}) + session._do_tmdb_override({"file_id": entry.id, "media_type": "tv"}) assert not issued assert [m for m in session.sent if m.get("type") == "error"] -async def test_unknown_path_is_refused(tmp_path): +async def test_unknown_file_id_is_refused(tmp_path): session = _session(tmp_path, "op", operator="op") session._has_admin_authority = lambda: True issued = [] session._issue_admin_challenge = lambda op, subject: issued.append((op, subject)) - session._do_tmdb_override({"path": "nope", "tmdb_id": "123", "media_type": "tv"}) + session._do_tmdb_override({"file_id": "nope", "tmdb_id": "123", "media_type": "tv"}) assert not issued assert [m for m in session.sent if m.get("type") == "error"] @@ -104,25 +105,51 @@ async def test_unknown_path_is_refused(tmp_path): async def test_a_request_with_nobody_to_authorize_it_is_refused(tmp_path): session = _session(tmp_path, "member-1", operator="the-operator") - session._ctx["index"].add_entry(_entry("shared", "ep.mkv", "Show")) + entry = _entry("shared", "ep.mkv", "Show") + session._ctx["index"].add_entry(entry) session._has_admin_authority = lambda: False - session._do_tmdb_override({"path": "shared", "tmdb_id": "123", "media_type": "tv"}) + session._do_tmdb_override({"file_id": entry.id, "tmdb_id": "123", "media_type": "tv"}) assert [m for m in session.sent if m.get("type") == "error"] async def test_a_valid_request_is_signed(tmp_path): session = _session(tmp_path, "op", operator="op") - session._ctx["index"].add_entry(_entry("shared", "ep.mkv", "War of the Worlds")) + entry = _entry("shared", "ep.mkv", "War of the Worlds") + session._ctx["index"].add_entry(entry) session._has_admin_authority = lambda: True issued = [] session._issue_admin_challenge = lambda op, subject: issued.append((op, subject)) - session._do_tmdb_override({"path": "shared", "tmdb_id": "2255", "media_type": "tv"}) + session._do_tmdb_override({"file_id": entry.id, "tmdb_id": "2255", "media_type": "tv"}) assert issued == [(OP_TMDB_OVERRIDE, - "path=shared,tmdb_id=2255,media_type=tv")] + f"file_id={entry.id},tmdb_id=2255,media_type=tv")] + + +async def test_two_files_in_the_same_folder_are_told_apart(tmp_path): + """ + Regression (found live, 2026-08-25): `IndexEntry.path` is the *folder* a + file is in, not the file itself — two files in the same folder (any + multi-episode season) used to collide when looked up by path, silently + resolving to whichever entry the index happened to return first. Keyed + by `file_id` now, so two entries sharing a folder must resolve to their + own, distinct entries. + """ + session = _session(tmp_path, "op", operator="op") + e1 = _entry("shared/Season 1", "s01e01.mkv", "War of the Worlds") + e2 = _entry("shared/Season 1", "s01e02.mkv", "War of the Worlds") + session._ctx["index"].add_entry(e1) + session._ctx["index"].add_entry(e2) + session._has_admin_authority = lambda: True + issued = [] + session._issue_admin_challenge = lambda op, subject: issued.append((op, subject)) + + session._do_tmdb_override({"file_id": e2.id, "tmdb_id": "2255", "media_type": "tv"}) + + assert issued == [(OP_TMDB_OVERRIDE, + f"file_id={e2.id},tmdb_id=2255,media_type=tv")] # ── Applying the override ─────────────────────────────────────────────────── @@ -151,7 +178,7 @@ async def test_override_updates_every_entry_sharing_the_display_title(tmp_path): session._peer_registry = lambda: {"peer-1": peer} await session._admin_exec_tmdb_override( - {"subject": "path=shared/S1,tmdb_id=999,media_type=tv"}, + {"subject": f"file_id={s1.id},tmdb_id=999,media_type=tv"}, b"transcript", b"sig") for e in (s1, s2, s3): |