From c66849155519c45cd065325299c95242b1c96863 Mon Sep 17 00:00:00 2001 From: chodeus <190988615+chodeus@users.noreply.github.com> Date: Thu, 19 Nov 2026 05:56:40 +0800 Subject: [PATCH 1/6] fix(media): read-only collections list; collection API stores titles as a sync does - Media Manage lists Plex collections read-only: its Create, Rename and Delete never worked (wrong payload, and the next Plex sync undoes any change). - POST/PUT /api/media/collections check title and alternate_titles and normalize them with normalize_titles; PUT updates the row in place and answers 409 on a taken title. - One parse_alternate_titles helper for the manual picker and the poster renamerr, so a non-string alternate title no longer crashes the picker. --- backend/api/media_api.py | 68 ++++---- backend/modules/poster_renamerr.py | 11 +- backend/util/asset_candidates.py | 9 +- backend/util/database/collection_cache.py | 40 ++++- backend/util/helper.py | 9 + frontend/src/pages/media/MediaManagePage.jsx | 164 +----------------- .../src/pages/media/MediaManagePage.test.jsx | 22 ++- frontend/src/utils/api/media.js | 15 -- tests/test_helper.py | 12 ++ tests/test_media_api_queries.py | 60 +++++++ tests/test_posters_queries.py | 13 ++ 11 files changed, 200 insertions(+), 223 deletions(-) diff --git a/backend/api/media_api.py b/backend/api/media_api.py index e534ca3fa..ded300a2c 100644 --- a/backend/api/media_api.py +++ b/backend/api/media_api.py @@ -33,6 +33,7 @@ from backend.util.arr_delete import ArrDeleter, ArrOutcome from backend.util.arr_instances import instance_detail from backend.util.config import ConfigError, load_config +from backend.util.normalization import normalize_titles from backend.util.public_error import failure_text from backend.util.database import ( INCOMPLETE_METADATA_FIELDS, @@ -746,6 +747,22 @@ def delete_collection( ) +def _title_fields(payload: dict) -> Tuple[Dict[str, Any], Optional[str]]: + """The title columns ``payload`` sets, normalized as a Plex sync stores them, or why they are invalid.""" + fields: Dict[str, Any] = {} + if "title" in payload: + title = payload["title"] + if not isinstance(title, str) or not title.strip(): + return {}, "Field 'title' must be a non-empty string" + fields.update(title=title, normalized_title=normalize_titles(title)) + if "alternate_titles" in payload: + alts = payload["alternate_titles"] + if not isinstance(alts, list) or not all(isinstance(t, str) and t.strip() for t in alts): + return {}, "Field 'alternate_titles' must be a list of non-empty strings" + fields.update(alternate_titles=alts, normalized_alternate_titles=[normalize_titles(t) for t in alts]) + return fields, None + + @router.post( "/collections", summary="Create collection", @@ -763,7 +780,7 @@ def delete_collection( } }, }, - 400: {"description": "Missing 'title' or 'instance_name'"}, + 400: {"description": "Missing 'title' or 'instance_name', or a title field of the wrong type"}, 413: {"description": "Request body too large"}, }, ) @@ -801,19 +818,19 @@ async def create_collection( status_code=400, ) + titles, why = _title_fields(payload) + if why: + return error(why, code="INVALID_FIELDS", status_code=400) record = { - "title": title, - "normalized_title": title.lower().strip() if title else None, "year": payload.get("year"), "tmdb_id": payload.get("tmdb_id"), "tvdb_id": payload.get("tvdb_id"), "imdb_id": payload.get("imdb_id"), "folder": payload.get("folder"), "library_name": payload.get("library_name"), - "alternate_titles": payload.get("alternate_titles", []), - "normalized_alternate_titles": payload.get( - "normalized_alternate_titles", [] - ), + "alternate_titles": [], + "normalized_alternate_titles": [], + **titles, } db.collection.upsert(record, instance_name) @@ -852,7 +869,10 @@ async def create_collection( } } }, - } + }, + 400: {"description": "A title field of the wrong type"}, + 404: {"description": "No collection with that id"}, + 409: {"description": "Another collection in this library already has that title"}, }, ) async def update_collection( @@ -897,31 +917,21 @@ async def update_collection( status_code=404, ) - # Merge updates into existing record - updatable_fields = [ - "title", - "year", - "tmdb_id", - "tvdb_id", - "imdb_id", - "folder", - "library_name", - "alternate_titles", - "normalized_alternate_titles", - ] - for field in updatable_fields: + titles, why = _title_fields(payload) + if why: + return error(why, code="INVALID_FIELDS", status_code=400) + for field in ("year", "tmdb_id", "tvdb_id", "imdb_id", "folder", "library_name"): if field in payload: existing[field] = payload[field] - - # Update normalized_title if title changed - if "title" in payload: - existing["normalized_title"] = ( - payload["title"].lower().strip() if payload["title"] else None + existing.update(titles) + # In place by id: an upsert keyed on the new title would add a second row. + if not db.collection.update_row(collection_id, existing): + return error( + "Another collection in this library already has that title", + code="COLLECTION_EXISTS", + status_code=409, ) - instance_name = existing.get("instance_name") - db.collection.upsert(existing, instance_name) - # Fetch the updated record updated = db.collection.get_by_id(collection_id) diff --git a/backend/modules/poster_renamerr.py b/backend/modules/poster_renamerr.py index 9e7d22c76..5fa64c7ff 100755 --- a/backend/modules/poster_renamerr.py +++ b/backend/modules/poster_renamerr.py @@ -27,6 +27,7 @@ ) from backend.util.helper import ( as_list, + parse_alternate_titles, classify_match, is_id_match_reason, create_table, @@ -529,15 +530,7 @@ def find_asset_candidate( season_number = media.get("season_number") expected_asset_type = "collection" if is_collection else asset_type - alt_titles = [] - try: - alt_titles = json.loads(media.get("alternate_titles") or "[]") - except Exception: - pass - # The collections API stores these unchecked: keep only the strings. - if not isinstance(alt_titles, list): - alt_titles = [] - alt_titles = [t for t in alt_titles if isinstance(t, str) and t] + alt_titles = parse_alternate_titles(media.get("alternate_titles")) reasons = [] candidates = [] diff --git a/backend/util/asset_candidates.py b/backend/util/asset_candidates.py index 283e237a2..92358e41a 100644 --- a/backend/util/asset_candidates.py +++ b/backend/util/asset_candidates.py @@ -6,10 +6,9 @@ """ import difflib -import json from typing import Any, Dict, List, Optional, Tuple -from backend.util.helper import is_match +from backend.util.helper import is_match, parse_alternate_titles from backend.util.normalization import normalize_titles @@ -22,11 +21,7 @@ def rank_candidates( ) -> List[Tuple[Dict[str, Any], bool, float, str]]: """Best-first (row, would_match, similarity, reason) tuples for a media row.""" season_number = row.get("season_number") - try: - alts = json.loads(row.get("alternate_titles") or "[]") - except (ValueError, TypeError): - alts = [] - search_titles = [row.get("title")] + [a for a in alts if a] + search_titles = [row.get("title")] + parse_alternate_titles(row.get("alternate_titles")) row_norm = row.get("normalized_title") or normalize_titles(row.get("title") or "") seen = set() diff --git a/backend/util/database/collection_cache.py b/backend/util/database/collection_cache.py index 1a869097e..e18f41a7a 100755 --- a/backend/util/database/collection_cache.py +++ b/backend/util/database/collection_cache.py @@ -1,4 +1,5 @@ import json +import sqlite3 from typing import Any, List, Optional from .db_base import DatabaseBase @@ -10,18 +11,45 @@ class CollectionCache(DatabaseBase): Provides CRUD and sync operations. """ - def upsert(self, record: dict, instance_name: str) -> None: - """Insert or update a collection record for a given instance.""" - record["instance_name"] = instance_name - record["asset_type"] = "collection" - - # Serialize list/dict fields to JSON + @staticmethod + def _store_json(record: dict) -> None: + """Serialize ``record``'s title lists to JSON in place; a missing list is stored empty.""" for key in ("alternate_titles", "normalized_alternate_titles"): if isinstance(record.get(key), (list, dict)): record[key] = json.dumps(record[key]) elif record.get(key) is None: record[key] = json.dumps([]) + def update_row(self, id: int, record: dict) -> bool: + """Write ``record``'s fields onto row ``id`` in place; False when its title is taken in that library and instance.""" + self._store_json(record) + columns = ( + "title", + "normalized_title", + "alternate_titles", + "normalized_alternate_titles", + "year", + "tmdb_id", + "tvdb_id", + "imdb_id", + "folder", + "library_name", + ) + try: + self.execute_query( + f"UPDATE collections_cache SET {', '.join(c + '=?' for c in columns)} WHERE id=?", + (*(record.get(c) for c in columns), id), + ) + except sqlite3.IntegrityError: + return False + return True + + def upsert(self, record: dict, instance_name: str) -> None: + """Insert or update a collection record for a given instance.""" + record["instance_name"] = instance_name + record["asset_type"] = "collection" + self._store_json(record) + self.execute_query( """ INSERT INTO collections_cache diff --git a/backend/util/helper.py b/backend/util/helper.py index a2596e600..5a5a0ca78 100755 --- a/backend/util/helper.py +++ b/backend/util/helper.py @@ -684,6 +684,15 @@ def classify_match(matched: bool, reason: str) -> Tuple[str, float]: return "matched", _MATCH_CONF_EXACT +def parse_alternate_titles(raw: Any) -> List[str]: + """The non-empty string titles in a row's ``alternate_titles`` JSON; [] when it is unreadable or not a list.""" + try: + titles = json.loads(raw or "[]") + except (TypeError, ValueError): + return [] + return [t for t in titles if isinstance(t, str) and t] if isinstance(titles, list) else [] + + def generate_title_variants(title: str) -> Dict[str, List[str]]: """Produce alternate title candidates plus their normalized forms. diff --git a/frontend/src/pages/media/MediaManagePage.jsx b/frontend/src/pages/media/MediaManagePage.jsx index aca0c99d9..fa1eeeb94 100644 --- a/frontend/src/pages/media/MediaManagePage.jsx +++ b/frontend/src/pages/media/MediaManagePage.jsx @@ -6,7 +6,7 @@ import { mediaAPI } from '../../utils/api/media.js'; import { nestarrAPI } from '../../utils/api/nestarr.js'; import { apiCore } from '../../utils/api/core.js'; import { Modal } from '../../components/modals/Modal'; -import { Button, LoadingButton, IconButton } from '../../components/ui/index.js'; +import { Button, LoadingButton } from '../../components/ui/index.js'; import { PageHeader } from '../../components/ui/PageHeader'; import Spinner from '../../components/ui/Spinner.jsx'; import { LibraryMaintenance } from '../../components/maintenance/LibraryMaintenance.jsx'; @@ -426,10 +426,6 @@ const MediaManagePage = () => { const [resolveDeleteFiles, setResolveDeleteFiles] = useState(false); const [resolveMembers, setResolveMembers] = useState(null); const [resolveMembersLoading, setResolveMembersLoading] = useState(false); - const [showCreateCollection, setShowCreateCollection] = useState(false); - const [newCollectionName, setNewCollectionName] = useState(''); - const [editCollection, setEditCollection] = useState(null); - const [editCollectionName, setEditCollectionName] = useState(''); const [nestedIssues, setNestedIssues] = useState([]); const [lastScanTime, setLastScanTime] = useState(null); const [unmatchedEnabled, setUnmatchedEnabled] = useState(true); @@ -447,40 +443,11 @@ const MediaManagePage = () => { mediaAPI.exportMedia({ format: 'json' }) ); - const { execute: createCollection, isLoading: isCreatingCollection } = useApiMutation( - name => mediaAPI.createCollection({ name }), - { - successMessage: 'Collection created', - onSuccess: () => { - setShowCreateCollection(false); - setNewCollectionName(''); - refreshCollections(); - }, - } - ); - - const { execute: updateCollectionMutation, isLoading: isUpdatingCollection } = useApiMutation( - // execute() forwards a single arg — take one object, not (id, data). - ({ id, data }) => mediaAPI.updateCollection(id, data), - { - successMessage: 'Collection updated', - onSuccess: () => { - setEditCollection(null); - setEditCollectionName(''); - refreshCollections(); - }, - } - ); - - const { data: collectionsData, refresh: refreshCollections } = useApiData({ + const { data: collectionsData } = useApiData({ apiFunction: mediaAPI.fetchCollections, options: { showErrorToast: false }, }); - const { execute: deleteCollection } = useApiMutation(id => mediaAPI.deleteCollection(id), { - successMessage: 'Collection deleted', - }); - const { execute: deleteItem, isLoading: isDeleting } = useApiMutation( () => mediaAPI.deleteMediaItem(deleteTarget?.id, { deleteFiles }), { successMessage: 'Media item deleted' } @@ -675,24 +642,6 @@ const MediaManagePage = () => { } }; - const handleCreateCollection = async () => { - if (!newCollectionName.trim()) return; - try { - await createCollection(newCollectionName.trim()); - } catch { - toast.error('Failed to create collection'); - } - }; - - const handleDeleteCollection = async id => { - try { - await deleteCollection(id); - refreshCollections(); - } catch { - toast.error('Failed to delete collection'); - } - }; - const handleDelete = async () => { if (!deleteTarget) return; try { @@ -1108,20 +1057,16 @@ const MediaManagePage = () => { title="Collections" count={`${collections.length}`} countClass="bg-primary/15 text-primary-hover" - > - - + /> +

+ Your Plex collections as of the last sync. Create, rename or delete them in + Plex. +

{collections.map((col, i) => (
@@ -1133,27 +1078,6 @@ const MediaManagePage = () => { )}
-
- { - setEditCollection(col); - setEditCollectionName(col.title || col.name || ''); - }} - /> - handleDeleteCollection(col.id)} - /> -
))}
@@ -1363,78 +1287,6 @@ const MediaManagePage = () => { - {/* Create Collection Modal */} - { - setShowCreateCollection(false); - setNewCollectionName(''); - }} - size="small" - > - Create Collection - - - setNewCollectionName(e.target.value)} - placeholder="My Collection" - className="w-full p-2 bg-input border border-border rounded-md text-fg text-sm focus:border-primary focus:outline-none" - /> - - - - - - - {/* Edit Collection Modal */} - { - setEditCollection(null); - setEditCollectionName(''); - }} - size="small" - > - Edit Collection - - - setEditCollectionName(e.target.value)} - className="w-full p-2 bg-input border border-border rounded-md text-fg text-sm focus:border-primary focus:outline-none" - /> - - - - - - {/* Fix Nested Media Modal — shows preview before executing */} { expect(media.fetchDuplicates).toHaveBeenCalledTimes(1); }); }); + +describe('MediaManagePage Collections', () => { + beforeEach(() => { + media.fetchDuplicates.mockResolvedValue({ data: { duplicates: [] } }); + media.fetchCollections.mockResolvedValue({ + data: { collections: [{ id: 1, title: 'Marvel Cinematic Universe', year: 2008 }] }, + }); + }); + + it('lists Plex collections read-only and says where to change them', async () => { + render(); + const section = (await screen.findByRole('heading', { name: 'Collections' })).closest( + 'section' + ); + + expect(within(section).getByText('Marvel Cinematic Universe')).toBeTruthy(); + expect(within(section).getByText(/Create, rename or delete them in Plex/)).toBeTruthy(); + expect(within(section).queryAllByRole('button')).toEqual([]); + }); +}); diff --git a/frontend/src/utils/api/media.js b/frontend/src/utils/api/media.js index 9bcab96db..da3df216c 100755 --- a/frontend/src/utils/api/media.js +++ b/frontend/src/utils/api/media.js @@ -117,21 +117,6 @@ export const mediaAPI = { }); }, - /** Create a collection from { name, description, mediaIds }. */ - createCollection: collectionData => { - return apiCore.post('/media/collections', collectionData); - }, - - /** Update a collection. */ - updateCollection: (collectionId, collectionData) => { - return apiCore.put(`/media/collections/${collectionId}`, collectionData); - }, - - /** Delete a collection. */ - deleteCollection: collectionId => { - return apiCore.delete(`/media/collections/${collectionId}`); - }, - /** Duplicate groups and folder collisions. */ fetchDuplicates: (options = {}) => { const params = new URLSearchParams(options); diff --git a/tests/test_helper.py b/tests/test_helper.py index 30df14174..60edc8a99 100644 --- a/tests/test_helper.py +++ b/tests/test_helper.py @@ -2,8 +2,11 @@ import os +import pytest + from backend.util.helper import ( + parse_alternate_titles, classify_match, compare_strings, create_bar, @@ -624,3 +627,12 @@ def test_is_match_both_years_present_and_differ_still_blocks(): media = {"title": "The Lovers", "year": 2023} matched, _ = is_match(asset, media) assert matched is False + + +@pytest.mark.parametrize( + "raw, titles", + [(None, []), ("", []), ("not json", []), ('{"Duna": 1}', []), ('"Duna"', []), ('[1, "", "Duna"]', ["Duna"])], +) +def test_parse_alternate_titles_keeps_only_string_titles(raw, titles): + parsed = parse_alternate_titles(raw) + assert parsed == titles diff --git a/tests/test_media_api_queries.py b/tests/test_media_api_queries.py index 3a1de7da3..e003387fe 100644 --- a/tests/test_media_api_queries.py +++ b/tests/test_media_api_queries.py @@ -447,3 +447,63 @@ def test_create_collection_route_reads_back_the_created_row(db): collection = resp.json()["data"]["collection"] assert collection["title"] == "Marvel" assert collection["instance_name"] == "plex1" + + +def _collection_rows(db): + rows = db.collection.execute_query( + "SELECT title, normalized_title, normalized_alternate_titles FROM collections_cache ORDER BY id", + fetch_all=True, + ) + return [(r["title"], r["normalized_title"], r["normalized_alternate_titles"]) for r in rows] + + +def test_create_collection_stores_titles_normalized_as_a_sync_does(db): + resp = _client(db).post( + "/api/media/collections", + json={"title": "The Matrix Collection", "instance_name": "plex1", "alternate_titles": ["Matrix"]}, + ) + rows = _collection_rows(db) + assert (resp.status_code, rows) == (200, [("The Matrix Collection", "thematrixcollection", '["matrix"]')]) + + +@pytest.mark.parametrize( + "body", + [{"alternate_titles": "Matrix"}, {"alternate_titles": ["Matrix", 7]}, {"alternate_titles": [" "]}, {"title": 7}], + ids=["string", "number-in-list", "blank", "numeric-title"], +) +def test_create_collection_refuses_title_fields_of_the_wrong_shape(db, body): + resp = _client(db).post("/api/media/collections", json={"title": "Marvel", "instance_name": "plex1", **body}) + assert (resp.status_code, resp.json()["error_code"], _collection_rows(db)) == (400, "INVALID_FIELDS", []) + + +def test_update_collection_renames_in_place_and_renormalizes(db): + client = _client(db) + created = client.post("/api/media/collections", json={"title": "Matrix", "instance_name": "plex1"}) + cid = created.json()["data"]["collection"]["id"] + resp = client.put( + f"/api/media/collections/{cid}", + json={"title": "The Matrix Collection", "alternate_titles": ["Matrix"]}, + ) + rows = _collection_rows(db) + assert (resp.status_code, resp.json()["data"]["collection"]["title"], rows) == ( + 200, + "The Matrix Collection", + [("The Matrix Collection", "thematrixcollection", '["matrix"]')], + ) + + +@pytest.mark.parametrize( + "body, status, code", + [({"title": ""}, 400, "INVALID_FIELDS"), ({"title": "Marvel"}, 409, "COLLECTION_EXISTS")], + ids=["blank-title", "taken-title"], +) +def test_update_collection_refuses_a_title_it_cannot_store(db, body, status, code): + client = _client(db) + for title in ("Marvel", "Matrix"): + client.post( + "/api/media/collections", json={"title": title, "instance_name": "plex1", "library_name": "Movies"} + ) + cid = db.collection.get_by_title_and_instance("Matrix", "plex1", "Movies")["id"] + resp = client.put(f"/api/media/collections/{cid}", json=body) + titles = [title for title, *_ in _collection_rows(db)] + assert (resp.status_code, resp.json()["error_code"], titles) == (status, code, ["Marvel", "Matrix"]) diff --git a/tests/test_posters_queries.py b/tests/test_posters_queries.py index 48eec20f0..17511cf37 100644 --- a/tests/test_posters_queries.py +++ b/tests/test_posters_queries.py @@ -934,3 +934,16 @@ def _boom(*_a, **_kw): "SELECT file FROM poster_cache", fetch_one=True )["file"] assert stored == str(poster) # row still points at the surviving file + + +@pytest.mark.parametrize("stored", ['{"Duna": 1}', '[1, "Duna"]', "not json"]) +def test_rank_candidates_ranks_a_row_with_odd_alternate_titles(db, stored): + """A row stored with non-string alternate titles still ranks: only its string titles are looked up.""" + from backend.util.asset_candidates import rank_candidates + + _seed_poster(db, "Dune", file="/src/Dune (2021).jpg") + row = {**db.media.get_by_id(_seed_media(db, "Dune")), "alternate_titles": stored} + + ranked = rank_candidates(db, row, "movie") + + assert [c["title"] for c, *_ in ranked] == ["Dune"] From 6d4bc4feb2450cc83e53910db8da439086a358f2 Mon Sep 17 00:00:00 2001 From: chodeus <190988615+chodeus@users.noreply.github.com> Date: Thu, 19 Nov 2026 05:59:29 +0800 Subject: [PATCH 2/6] refactor(helper): one json_list for every alternate_titles read is_match's nested copy and tmdb's _json_list become helper.json_list; parse_alternate_titles builds on it. --- backend/util/helper.py | 32 +++++++++++++++----------------- backend/util/tmdb.py | 19 ++++--------------- tests/test_helper.py | 10 +++++++++- 3 files changed, 28 insertions(+), 33 deletions(-) diff --git a/backend/util/helper.py b/backend/util/helper.py index 5a5a0ca78..5ab80d01b 100755 --- a/backend/util/helper.py +++ b/backend/util/helper.py @@ -453,17 +453,6 @@ def is_match( ) media["normalized_folder_title"] = normalize_titles(media["folder_title"]) - def json_list(value: Any) -> List[Any]: - if isinstance(value, list): - return value - if isinstance(value, str): - try: - parsed = json.loads(value) - return parsed if isinstance(parsed, list) else [] - except Exception: - return [] - return [] - asset_alt_titles = json_list(asset.get("alternate_titles")) asset_norm_alt_titles = json_list(asset.get("normalized_alternate_titles")) media_alt_titles = json_list(media.get("alternate_titles")) @@ -684,13 +673,22 @@ def classify_match(matched: bool, reason: str) -> Tuple[str, float]: return "matched", _MATCH_CONF_EXACT +def json_list(value: Any) -> List[Any]: + """``value`` as a list: itself, or the list its JSON holds; [] for anything else.""" + if isinstance(value, list): + return value + if isinstance(value, str): + try: + parsed = json.loads(value) + except ValueError: + return [] + return parsed if isinstance(parsed, list) else [] + return [] + + def parse_alternate_titles(raw: Any) -> List[str]: - """The non-empty string titles in a row's ``alternate_titles`` JSON; [] when it is unreadable or not a list.""" - try: - titles = json.loads(raw or "[]") - except (TypeError, ValueError): - return [] - return [t for t in titles if isinstance(t, str) and t] if isinstance(titles, list) else [] + """The non-empty string titles in a row's ``alternate_titles`` (a list or its JSON).""" + return [t for t in json_list(raw) if isinstance(t, str) and t] def generate_title_variants(title: str) -> Dict[str, List[str]]: diff --git a/backend/util/tmdb.py b/backend/util/tmdb.py index ce8abc951..55918b14f 100644 --- a/backend/util/tmdb.py +++ b/backend/util/tmdb.py @@ -23,6 +23,7 @@ from backend.util.config import TMDBConfig from backend.util.database import ChubDB +from backend.util.helper import json_list from backend.util.normalization import normalize_titles @@ -638,18 +639,6 @@ def _norm_set(values: Iterable[Optional[str]]) -> Set[str]: return out -def _json_list(value) -> List: - if isinstance(value, list): - return value - if isinstance(value, str): - try: - parsed = json.loads(value) - return parsed if isinstance(parsed, list) else [] - except (ValueError, TypeError): - return [] - return [] - - def run_match_quality( db: ChubDB, cfg: TMDBConfig, @@ -734,7 +723,7 @@ def _verify_and_hydrate( ) row_titles = _norm_set( [row.get("title"), row.get("normalized_title")] - + _json_list(row.get("alternate_titles")) + + json_list(row.get("alternate_titles")) ) # row's normalized_title is already normalized; include raw too row_titles |= {n for n in [row.get("normalized_title")] if n} @@ -778,8 +767,8 @@ def _flag_review(db: ChubDB, media_id: int, reason: str) -> None: def _merge_akas(db: ChubDB, row: dict, tmdb_akas: List[str]) -> bool: """Union TMDB AKAs into a row's alternate_titles / normalized_alternate_titles. Returns True if anything new was added.""" - existing = _json_list(row.get("alternate_titles")) - existing_norm = set(_json_list(row.get("normalized_alternate_titles"))) + existing = json_list(row.get("alternate_titles")) + existing_norm = set(json_list(row.get("normalized_alternate_titles"))) added = False for aka in tmdb_akas: n = normalize_titles(aka) diff --git a/tests/test_helper.py b/tests/test_helper.py index 60edc8a99..49cf68217 100644 --- a/tests/test_helper.py +++ b/tests/test_helper.py @@ -631,7 +631,15 @@ def test_is_match_both_years_present_and_differ_still_blocks(): @pytest.mark.parametrize( "raw, titles", - [(None, []), ("", []), ("not json", []), ('{"Duna": 1}', []), ('"Duna"', []), ('[1, "", "Duna"]', ["Duna"])], + [ + (None, []), + ("", []), + ("not json", []), + ('{"Duna": 1}', []), + ('"Duna"', []), + ('[1, "", "Duna"]', ["Duna"]), + (["Duna", 1], ["Duna"]), + ], ) def test_parse_alternate_titles_keeps_only_string_titles(raw, titles): parsed = parse_alternate_titles(raw) From ffbc9d2786471090516330905314904ac77dad01 Mon Sep 17 00:00:00 2001 From: chodeus <190988615+chodeus@users.noreply.github.com> Date: Thu, 19 Nov 2026 06:10:31 +0800 Subject: [PATCH 3/6] fix(media): collection API derives titles with the sync's helper and checks every field - collection_title_fields owns the sync's title derivation (unescape, ASCII, normalize, variants); the Plex sync and the API both call it, so a rename regenerates variants. - One _collection_fields check for POST and PUT: ids and year are positive whole numbers, text fields are strings, library_name is required on create; a PUT with nothing to set answers NO_FIELDS. - update_row refuses a clash in one statement with IS, so a NULL library no longer slips past the UNIQUE key and a 409 logs no database error; a row removed mid-request answers 404. - json_list also absorbs RecursionError; the poster renamerr's last raw json.loads reads through parse_alternate_titles. --- backend/api/media_api.py | 89 +++++++----- backend/modules/poster_renamerr.py | 4 +- backend/util/database/collection_cache.py | 25 ++-- backend/util/helper.py | 12 +- backend/util/plex.py | 16 +-- tests/test_helper.py | 1 + tests/test_media_api_queries.py | 161 +++++++++++++++++----- tests/test_poster_renamerr.py | 5 + tests/test_posters_queries.py | 2 +- 9 files changed, 216 insertions(+), 99 deletions(-) diff --git a/backend/api/media_api.py b/backend/api/media_api.py index ded300a2c..a73654fc1 100644 --- a/backend/api/media_api.py +++ b/backend/api/media_api.py @@ -33,6 +33,7 @@ from backend.util.arr_delete import ArrDeleter, ArrOutcome from backend.util.arr_instances import instance_detail from backend.util.config import ConfigError, load_config +from backend.util.helper import collection_title_fields from backend.util.normalization import normalize_titles from backend.util.public_error import failure_text from backend.util.database import ( @@ -747,18 +748,38 @@ def delete_collection( ) -def _title_fields(payload: dict) -> Tuple[Dict[str, Any], Optional[str]]: - """The title columns ``payload`` sets, normalized as a Plex sync stores them, or why they are invalid.""" +def _is_text(value: Any) -> bool: + """A string with more than whitespace in it.""" + return isinstance(value, str) and bool(value.strip()) + + +def _collection_fields(payload: dict) -> Tuple[Dict[str, Any], Optional[str]]: + """The collection columns ``payload`` sets, titles derived as a Plex sync derives them, or why one is invalid.""" fields: Dict[str, Any] = {} + for key in ("year", "tmdb_id", "tvdb_id"): + if key in payload: + number = None if payload[key] is None else _media_id(payload[key]) + if payload[key] is not None and number is None: + return {}, f"Field '{key}' must be a positive whole number or null" + fields[key] = number + for key in ("imdb_id", "folder"): + if key in payload: + if payload[key] is not None and not isinstance(payload[key], str): + return {}, f"Field '{key}' must be a string or null" + fields[key] = payload[key] + if "library_name" in payload: + if not _is_text(payload["library_name"]): + return {}, "Field 'library_name' must be a non-empty string" + fields["library_name"] = payload["library_name"] if "title" in payload: - title = payload["title"] - if not isinstance(title, str) or not title.strip(): + if not _is_text(payload["title"]): return {}, "Field 'title' must be a non-empty string" - fields.update(title=title, normalized_title=normalize_titles(title)) + fields.update(collection_title_fields(payload["title"].strip())) if "alternate_titles" in payload: alts = payload["alternate_titles"] - if not isinstance(alts, list) or not all(isinstance(t, str) and t.strip() for t in alts): + if not isinstance(alts, list) or not all(_is_text(t) for t in alts): return {}, "Field 'alternate_titles' must be a list of non-empty strings" + alts = [t.strip() for t in alts] fields.update(alternate_titles=alts, normalized_alternate_titles=[normalize_titles(t) for t in alts]) return fields, None @@ -780,7 +801,7 @@ def _title_fields(payload: dict) -> Tuple[Dict[str, Any], Optional[str]]: } }, }, - 400: {"description": "Missing 'title' or 'instance_name', or a title field of the wrong type"}, + 400: {"description": "Missing 'title', 'instance_name' or 'library_name', or a field of the wrong type"}, 413: {"description": "Request body too large"}, }, ) @@ -804,8 +825,7 @@ async def create_collection( return body_too_large_error() logger.debug(f"Serving POST /api/media/collections with payload: {payload}") - title = payload.get("title") - if not title: + if "title" not in payload: return error( "Field 'title' is required", code="MISSING_TITLE", status_code=400 ) @@ -818,26 +838,21 @@ async def create_collection( status_code=400, ) - titles, why = _title_fields(payload) + if "library_name" not in payload: + return error( + "Field 'library_name' is required", + code="MISSING_LIBRARY", + status_code=400, + ) + record, why = _collection_fields(payload) if why: return error(why, code="INVALID_FIELDS", status_code=400) - record = { - "year": payload.get("year"), - "tmdb_id": payload.get("tmdb_id"), - "tvdb_id": payload.get("tvdb_id"), - "imdb_id": payload.get("imdb_id"), - "folder": payload.get("folder"), - "library_name": payload.get("library_name"), - "alternate_titles": [], - "normalized_alternate_titles": [], - **titles, - } db.collection.upsert(record, instance_name) - # Fetch the created/updated record + # Read back by the stored title: the sync's derivation may have changed it. created = db.collection.get_by_title_and_instance( - title, instance_name, record["library_name"] + record["title"], instance_name, record["library_name"] ) return ok("Collection created successfully", {"collection": created}) @@ -870,7 +885,7 @@ async def create_collection( } }, }, - 400: {"description": "A title field of the wrong type"}, + 400: {"description": "No updatable field, or a field of the wrong type"}, 404: {"description": "No collection with that id"}, 409: {"description": "Another collection in this library already has that title"}, }, @@ -885,7 +900,7 @@ async def update_collection( Update a media collection by ID. Fetches the existing collection, merges in the provided updates, - and upserts the result back to the database. + and writes the result back onto that row. Args: collection_id: The unique identifier of the collection to update @@ -898,7 +913,7 @@ async def update_collection( if payload is BODY_TOO_LARGE: return body_too_large_error() # An unusable body used to raise into the 500 handler; keep that - # response — falling through would upsert the record unchanged. + # response — falling through would write the record unchanged. if not isinstance(payload, dict) or not payload: return error( "Error updating collection", @@ -917,24 +932,28 @@ async def update_collection( status_code=404, ) - titles, why = _title_fields(payload) + fields, why = _collection_fields(payload) if why: return error(why, code="INVALID_FIELDS", status_code=400) - for field in ("year", "tmdb_id", "tvdb_id", "imdb_id", "folder", "library_name"): - if field in payload: - existing[field] = payload[field] - existing.update(titles) + if not fields: + return error("No valid fields to update", code="NO_FIELDS", status_code=400) + existing.update(fields) # In place by id: an upsert keyed on the new title would add a second row. - if not db.collection.update_row(collection_id, existing): + written = db.collection.update_row(collection_id, existing) + updated = db.collection.get_by_id(collection_id) + if updated is None: + return error( + f"Collection {collection_id} not found", + code="COLLECTION_NOT_FOUND", + status_code=404, + ) + if not written: return error( "Another collection in this library already has that title", code="COLLECTION_EXISTS", status_code=409, ) - # Fetch the updated record - updated = db.collection.get_by_id(collection_id) - return ok("Collection updated successfully", {"collection": updated}) except Exception as e: diff --git a/backend/modules/poster_renamerr.py b/backend/modules/poster_renamerr.py index 5fa64c7ff..4d560f63e 100755 --- a/backend/modules/poster_renamerr.py +++ b/backend/modules/poster_renamerr.py @@ -637,9 +637,7 @@ def _title_match_reason( """is_match()'s reason when a title-lookup poster belongs to ``media``, else "".""" if cand.get("season_number") != media.get("season_number"): return "" - cand_alt_titles = set( - json.loads(cand.get("normalized_alternate_titles", "[]") or "[]") - ) + cand_alt_titles = set(parse_alternate_titles(cand.get("normalized_alternate_titles"))) if ( cand.get("normalized_title", "") not in all_titles and not all_titles & cand_alt_titles diff --git a/backend/util/database/collection_cache.py b/backend/util/database/collection_cache.py index e18f41a7a..9db4173d1 100755 --- a/backend/util/database/collection_cache.py +++ b/backend/util/database/collection_cache.py @@ -1,5 +1,4 @@ import json -import sqlite3 from typing import Any, List, Optional from .db_base import DatabaseBase @@ -21,7 +20,7 @@ def _store_json(record: dict) -> None: record[key] = json.dumps([]) def update_row(self, id: int, record: dict) -> bool: - """Write ``record``'s fields onto row ``id`` in place; False when its title is taken in that library and instance.""" + """Write ``record``'s fields onto row ``id`` in place; False when that row is gone or another holds its key.""" self._store_json(record) columns = ( "title", @@ -35,14 +34,20 @@ def update_row(self, id: int, record: dict) -> bool: "folder", "library_name", ) - try: - self.execute_query( - f"UPDATE collections_cache SET {', '.join(c + '=?' for c in columns)} WHERE id=?", - (*(record.get(c) for c in columns), id), - ) - except sqlite3.IntegrityError: - return False - return True + # `IS`: the UNIQUE key treats NULL libraries as distinct, so it never refuses a clash there. + written = self.execute_query( + f"UPDATE collections_cache SET {', '.join(c + '=?' for c in columns)} WHERE id=? AND NOT EXISTS " + "(SELECT 1 FROM collections_cache WHERE title IS ? AND library_name IS ? AND instance_name IS ? AND id<>?)", + ( + *(record.get(c) for c in columns), + id, + record.get("title"), + record.get("library_name"), + record.get("instance_name"), + id, + ), + ) + return bool(written) def upsert(self, record: dict, instance_name: str) -> None: """Insert or update a collection record for a given instance.""" diff --git a/backend/util/helper.py b/backend/util/helper.py index 5ab80d01b..8a83f9090 100755 --- a/backend/util/helper.py +++ b/backend/util/helper.py @@ -1,6 +1,7 @@ # util/helper.py import copy +import html import json import os import re @@ -12,6 +13,7 @@ import yaml from tqdm import tqdm +from unidecode import unidecode from backend.util.constants import ( folder_year_regex, @@ -680,14 +682,14 @@ def json_list(value: Any) -> List[Any]: if isinstance(value, str): try: parsed = json.loads(value) - except ValueError: + except (ValueError, RecursionError): return [] return parsed if isinstance(parsed, list) else [] return [] def parse_alternate_titles(raw: Any) -> List[str]: - """The non-empty string titles in a row's ``alternate_titles`` (a list or its JSON).""" + """The non-empty strings in an alternate-titles column, raw or normalized (a list or its JSON).""" return [t for t in json_list(raw) if isinstance(t, str) and t] @@ -763,3 +765,9 @@ def generate_title_variants(title: str) -> Dict[str, List[str]]: "alternate_titles": alternate_titles, "normalized_alternate_titles": normalized_alternate_titles, } + + +def collection_title_fields(title: str) -> Dict[str, Any]: + """The title columns collections_cache stores for a Plex collection named ``title``.""" + plain = unidecode(html.unescape(title)) + return {"title": plain, "normalized_title": normalize_titles(plain), **generate_title_variants(plain)} diff --git a/backend/util/plex.py b/backend/util/plex.py index 69d7f1217..6ec17770a 100755 --- a/backend/util/plex.py +++ b/backend/util/plex.py @@ -1,4 +1,3 @@ -import html import itertools import os import re @@ -12,12 +11,11 @@ from plexapi.exceptions import NotFound from pathvalidate import sanitize_filename from plexapi.server import PlexServer -from unidecode import unidecode from backend.util.config import get_config_path from backend.util.helper import ( YEAR_MATCH_TOLERANCE, - generate_title_variants, + collection_title_fields, progress, ) from backend.util.normalization import normalize_titles @@ -250,9 +248,8 @@ def get_collections( leave=False, ) as inner: for collection in inner: - title_unescaped = unidecode(html.unescape(collection.title)) - normalized_title = normalize_titles(title_unescaped) - alternate_titles = generate_title_variants(title_unescaped) + titles = collection_title_fields(collection.title) + title_unescaped = titles["title"] # Use pathvalidate rather than the bare regex so the folder name # is safe across Linux + Windows + Samba shares (handles # reserved names like CON/AUX, max-length truncation, and @@ -271,15 +268,10 @@ def get_collections( imdb_id = None tvdb_id = None media_item = { - "title": title_unescaped, - "normalized_title": normalized_title, + **titles, "location": library_name, "year": year, "folder": folder, - "alternate_titles": alternate_titles["alternate_titles"], - "normalized_alternate_titles": alternate_titles[ - "normalized_alternate_titles" - ], "library_name": library_name, "asset_type": "collection", "tmdb_id": tmdb_id, diff --git a/tests/test_helper.py b/tests/test_helper.py index 49cf68217..0e823760d 100644 --- a/tests/test_helper.py +++ b/tests/test_helper.py @@ -639,6 +639,7 @@ def test_is_match_both_years_present_and_differ_still_blocks(): ('"Duna"', []), ('[1, "", "Duna"]', ["Duna"]), (["Duna", 1], ["Duna"]), + pytest.param("[" * 100_000, [], id="deep-nesting"), ], ) def test_parse_alternate_titles_keeps_only_string_titles(raw, titles): diff --git a/tests/test_media_api_queries.py b/tests/test_media_api_queries.py index e003387fe..48bfbee87 100644 --- a/tests/test_media_api_queries.py +++ b/tests/test_media_api_queries.py @@ -441,7 +441,7 @@ def test_create_collection_route_reads_back_the_created_row(db): """POST /collections returns the row it just upserted.""" resp = _client(db).post( "/api/media/collections", - json={"title": "Marvel", "instance_name": "plex1", "year": 2012}, + json={"title": "Marvel", "instance_name": "plex1", "library_name": "Movies", "year": 2012}, ) assert resp.status_code == 200 collection = resp.json()["data"]["collection"] @@ -457,53 +457,142 @@ def _collection_rows(db): return [(r["title"], r["normalized_title"], r["normalized_alternate_titles"]) for r in rows] -def test_create_collection_stores_titles_normalized_as_a_sync_does(db): - resp = _client(db).post( - "/api/media/collections", - json={"title": "The Matrix Collection", "instance_name": "plex1", "alternate_titles": ["Matrix"]}, - ) - rows = _collection_rows(db) - assert (resp.status_code, rows) == (200, [("The Matrix Collection", "thematrixcollection", '["matrix"]')]) +def _plex_sync(db, titles, library="Movies", instance="plex1"): + """The real PlexClient.get_collections + sync_collections_cache over fake plexapi collections.""" + from types import SimpleNamespace + + from backend.util.plex import PlexClient + + found = [SimpleNamespace(title=t, smart=False, minYear=None) for t in titles] + section = SimpleNamespace(search=lambda libtype: found, type="movie") + client = PlexClient.__new__(PlexClient) + client.logger = _StubLog() + client.plex = SimpleNamespace(library=SimpleNamespace(section=lambda name: section)) + db.collection.sync_collections_cache(instance, library, client.get_collections(library, include_smart=True)) + + +def _post_collection(client, **fields): + body = {"title": "Marvel", "instance_name": "plex1", "library_name": "Movies", **fields} + return client.post("/api/media/collections", json=body) + + +@pytest.mark.parametrize("title", ["The Matrix Collection", "Amélie Collection", "Fast & Furious Collection"]) +def test_create_collection_stores_the_title_columns_a_sync_stores(tmp_path, title): + with ChubDB(_StubLog(), db_path=str(tmp_path / "api.db")) as api_db, ChubDB( + _StubLog(), db_path=str(tmp_path / "sync.db") + ) as sync_db: + resp = _post_collection(_client(api_db), title=f" {title} ") + _plex_sync(sync_db, [title]) + assert (resp.status_code, _collection_rows(api_db)) == (200, _collection_rows(sync_db)) + + +@pytest.mark.parametrize( + "body, code", + [ + ({"instance_name": "plex1", "library_name": "Movies"}, "MISSING_TITLE"), + ({"title": "Marvel", "instance_name": "plex1"}, "MISSING_LIBRARY"), + ], + ids=["no-title", "no-library"], +) +def test_create_collection_refuses_a_body_missing_a_key_field(db, body, code): + resp = _client(db).post("/api/media/collections", json=body) + assert (resp.status_code, resp.json()["error_code"], _collection_rows(db)) == (400, code, []) @pytest.mark.parametrize( "body", - [{"alternate_titles": "Matrix"}, {"alternate_titles": ["Matrix", 7]}, {"alternate_titles": [" "]}, {"title": 7}], - ids=["string", "number-in-list", "blank", "numeric-title"], + [ + {"alternate_titles": "Matrix"}, + {"alternate_titles": ["Matrix", 7]}, + {"alternate_titles": [" "]}, + {"title": 7}, + {"title": " "}, + {"library_name": ""}, + {"year": [2008]}, + {"year": True}, + {"tmdb_id": 2**63}, + {"imdb_id": 7}, + ], + ids=["string", "number-in-list", "blank", "numeric-title", "blank-title", "blank-library", "list-year", + "bool-year", "huge-id", "numeric-imdb"], ) -def test_create_collection_refuses_title_fields_of_the_wrong_shape(db, body): - resp = _client(db).post("/api/media/collections", json={"title": "Marvel", "instance_name": "plex1", **body}) +def test_create_collection_refuses_fields_of_the_wrong_shape(db, body): + resp = _post_collection(_client(db), **body) assert (resp.status_code, resp.json()["error_code"], _collection_rows(db)) == (400, "INVALID_FIELDS", []) -def test_update_collection_renames_in_place_and_renormalizes(db): - client = _client(db) - created = client.post("/api/media/collections", json={"title": "Matrix", "instance_name": "plex1"}) - cid = created.json()["data"]["collection"]["id"] - resp = client.put( - f"/api/media/collections/{cid}", - json={"title": "The Matrix Collection", "alternate_titles": ["Matrix"]}, - ) - rows = _collection_rows(db) - assert (resp.status_code, resp.json()["data"]["collection"]["title"], rows) == ( - 200, - "The Matrix Collection", - [("The Matrix Collection", "thematrixcollection", '["matrix"]')], - ) +@pytest.mark.parametrize( + "body, row", + [ + ({"title": "Star Wars Collection"}, ("Star Wars Collection", "starwarscollection", '["starwarscollection", "starwars"]')), + ( + {"title": "Star Wars Collection", "alternate_titles": ["Star Wars"]}, + ("Star Wars Collection", "starwarscollection", '["starwars"]'), + ), + ], + ids=["variants-regenerated", "variants-given"], +) +def test_update_collection_renames_in_place_and_rederives_its_titles(db, body, row): + _plex_sync(db, ["The Matrix Collection"]) + cid = db.collection.get_by_title_and_instance("The Matrix Collection", "plex1", "Movies")["id"] + resp = _client(db).put(f"/api/media/collections/{cid}", json=body) + assert (resp.status_code, resp.json()["data"]["collection"]["id"], _collection_rows(db)) == (200, cid, [row]) + + +class _ErrorLog(_StubLog): + """Keeps error() lines; every other log call is a no-op.""" + + def __init__(self): + self.errors = [] + + def error(self, msg, *_a, **_kw): + self.errors.append(str(msg)) @pytest.mark.parametrize( "body, status, code", - [({"title": ""}, 400, "INVALID_FIELDS"), ({"title": "Marvel"}, 409, "COLLECTION_EXISTS")], - ids=["blank-title", "taken-title"], + [ + ({"title": ""}, 400, "INVALID_FIELDS"), + ({"library_name": None}, 400, "INVALID_FIELDS"), + ({"name": "Avengers"}, 400, "NO_FIELDS"), + ({"title": "Marvel"}, 409, "COLLECTION_EXISTS"), + ], + ids=["blank-title", "null-library", "no-updatable-field", "taken-title"], ) -def test_update_collection_refuses_a_title_it_cannot_store(db, body, status, code): - client = _client(db) - for title in ("Marvel", "Matrix"): - client.post( - "/api/media/collections", json={"title": title, "instance_name": "plex1", "library_name": "Movies"} +def test_update_collection_refuses_a_change_it_cannot_store(tmp_path, body, status, code): + log = _ErrorLog() + with ChubDB(log, db_path=str(tmp_path / "chub.db")) as db: + _plex_sync(db, ["Marvel", "Matrix"]) + cid = db.collection.get_by_title_and_instance("Matrix", "plex1", "Movies")["id"] + resp = _client(db).put(f"/api/media/collections/{cid}", json=body) + titles = [title for title, *_ in _collection_rows(db)] + assert (resp.status_code, resp.json()["error_code"], titles, log.errors) == ( + status, + code, + ["Marvel", "Matrix"], + [], ) - cid = db.collection.get_by_title_and_instance("Matrix", "plex1", "Movies")["id"] - resp = client.put(f"/api/media/collections/{cid}", json=body) + + +def test_update_collection_refuses_a_taken_title_in_no_library(db): + for title in ("Marvel", "Matrix"): + db.collection.upsert({"title": title, "normalized_title": title.lower(), "library_name": None}, "plex1") + cid = db.collection.get_by_title_and_instance("Matrix", "plex1", None)["id"] + resp = _client(db).put(f"/api/media/collections/{cid}", json={"title": "Marvel"}) + titles = [title for title, *_ in _collection_rows(db)] + assert (resp.status_code, resp.json()["error_code"], titles) == (409, "COLLECTION_EXISTS", ["Marvel", "Matrix"]) + + +def test_update_collection_answers_404_when_a_sync_removed_the_row_first(db, monkeypatch): + _plex_sync(db, ["Marvel"]) + cid = db.collection.get_by_title_and_instance("Marvel", "plex1", "Movies")["id"] + write = db.collection.update_row + + def sync_lands_first(id, record): + _plex_sync(db, ["Avengers"]) + return write(id, record) + + monkeypatch.setattr(db.collection, "update_row", sync_lands_first) + resp = _client(db).put(f"/api/media/collections/{cid}", json={"year": 2012}) titles = [title for title, *_ in _collection_rows(db)] - assert (resp.status_code, resp.json()["error_code"], titles) == (status, code, ["Marvel", "Matrix"]) + assert (resp.status_code, resp.json()["error_code"], titles) == (404, "COLLECTION_NOT_FOUND", ["Avengers"]) diff --git a/tests/test_poster_renamerr.py b/tests/test_poster_renamerr.py index 9584a72f3..67aae84b3 100644 --- a/tests/test_poster_renamerr.py +++ b/tests/test_poster_renamerr.py @@ -2987,3 +2987,8 @@ def _broken(*_a): m.run() assert m.logger.messages["warning"][-1] == "Could not log the run summary: summary bug" + +def test_title_match_reason_skips_a_poster_with_unreadable_alternate_titles(): + poster = {"season_number": None, "normalized_title": "other", "normalized_alternate_titles": "matrix"} + reason = PosterRenamerr._title_match_reason(poster, {"season_number": None}, {"matrix"}, []) + assert reason == "" diff --git a/tests/test_posters_queries.py b/tests/test_posters_queries.py index 17511cf37..b3f0defba 100644 --- a/tests/test_posters_queries.py +++ b/tests/test_posters_queries.py @@ -938,7 +938,7 @@ def _boom(*_a, **_kw): @pytest.mark.parametrize("stored", ['{"Duna": 1}', '[1, "Duna"]', "not json"]) def test_rank_candidates_ranks_a_row_with_odd_alternate_titles(db, stored): - """A row stored with non-string alternate titles still ranks: only its string titles are looked up.""" + """A row whose stored alternate titles are not a list of strings still ranks instead of raising.""" from backend.util.asset_candidates import rank_candidates _seed_poster(db, "Dune", file="/src/Dune (2021).jpg") From ae64a5faf66a9589683209f06f7959c2c9d878b1 Mon Sep 17 00:00:00 2001 From: chodeus <190988615+chodeus@users.noreply.github.com> Date: Thu, 19 Nov 2026 06:14:21 +0800 Subject: [PATCH 4/6] refactor(media): split the collection field check below the complexity line; widen its tests _scalar_fields takes the year, ids, folder and library; the title test compares the raw alternate titles too, and a null now has a test. --- backend/api/media_api.py | 12 ++++++++++-- tests/test_media_api_queries.py | 26 +++++++++++++++++++++----- 2 files changed, 31 insertions(+), 7 deletions(-) diff --git a/backend/api/media_api.py b/backend/api/media_api.py index a73654fc1..827918b81 100644 --- a/backend/api/media_api.py +++ b/backend/api/media_api.py @@ -753,8 +753,8 @@ def _is_text(value: Any) -> bool: return isinstance(value, str) and bool(value.strip()) -def _collection_fields(payload: dict) -> Tuple[Dict[str, Any], Optional[str]]: - """The collection columns ``payload`` sets, titles derived as a Plex sync derives them, or why one is invalid.""" +def _scalar_fields(payload: dict) -> Tuple[Dict[str, Any], Optional[str]]: + """The year, ids, folder and library ``payload`` sets, or why one is invalid.""" fields: Dict[str, Any] = {} for key in ("year", "tmdb_id", "tvdb_id"): if key in payload: @@ -771,6 +771,14 @@ def _collection_fields(payload: dict) -> Tuple[Dict[str, Any], Optional[str]]: if not _is_text(payload["library_name"]): return {}, "Field 'library_name' must be a non-empty string" fields["library_name"] = payload["library_name"] + return fields, None + + +def _collection_fields(payload: dict) -> Tuple[Dict[str, Any], Optional[str]]: + """The collection columns ``payload`` sets (a title derived as a Plex sync derives it), or why one is invalid.""" + fields, why = _scalar_fields(payload) + if why: + return {}, why if "title" in payload: if not _is_text(payload["title"]): return {}, "Field 'title' must be a non-empty string" diff --git a/tests/test_media_api_queries.py b/tests/test_media_api_queries.py index 48bfbee87..89193f601 100644 --- a/tests/test_media_api_queries.py +++ b/tests/test_media_api_queries.py @@ -451,10 +451,10 @@ def test_create_collection_route_reads_back_the_created_row(db): def _collection_rows(db): rows = db.collection.execute_query( - "SELECT title, normalized_title, normalized_alternate_titles FROM collections_cache ORDER BY id", + "SELECT title, normalized_title, alternate_titles, normalized_alternate_titles FROM collections_cache ORDER BY id", fetch_all=True, ) - return [(r["title"], r["normalized_title"], r["normalized_alternate_titles"]) for r in rows] + return [(r["title"], r["normalized_title"], r["alternate_titles"], r["normalized_alternate_titles"]) for r in rows] def _plex_sync(db, titles, library="Movies", instance="plex1"): @@ -524,10 +524,18 @@ def test_create_collection_refuses_fields_of_the_wrong_shape(db, body): @pytest.mark.parametrize( "body, row", [ - ({"title": "Star Wars Collection"}, ("Star Wars Collection", "starwarscollection", '["starwarscollection", "starwars"]')), ( - {"title": "Star Wars Collection", "alternate_titles": ["Star Wars"]}, - ("Star Wars Collection", "starwarscollection", '["starwars"]'), + {"title": "Star Wars Collection"}, + ( + "Star Wars Collection", + "starwarscollection", + '["Star Wars Collection", "Star Wars"]', + '["starwarscollection", "starwars"]', + ), + ), + ( + {"title": "Star Wars Collection", "alternate_titles": [" Star Wars "]}, + ("Star Wars Collection", "starwarscollection", '["Star Wars"]', '["starwars"]'), ), ], ids=["variants-regenerated", "variants-given"], @@ -539,6 +547,14 @@ def test_update_collection_renames_in_place_and_rederives_its_titles(db, body, r assert (resp.status_code, resp.json()["data"]["collection"]["id"], _collection_rows(db)) == (200, cid, [row]) +def test_update_collection_clears_the_fields_it_is_sent_as_null(db): + created = _post_collection(_client(db), year=2012, tmdb_id=86311, folder="Marvel") + cid = created.json()["data"]["collection"]["id"] + resp = _client(db).put(f"/api/media/collections/{cid}", json={"year": None, "tmdb_id": None, "folder": None}) + row = db.collection.get_by_id(cid) + assert (resp.status_code, row["year"], row["tmdb_id"], row["folder"]) == (200, None, None, None) + + class _ErrorLog(_StubLog): """Keeps error() lines; every other log call is a no-op.""" From cda5d8f00a94343c11dccddacf3756a5e2355fc1 Mon Sep 17 00:00:00 2001 From: chodeus <190988615+chodeus@users.noreply.github.com> Date: Thu, 19 Nov 2026 06:25:41 +0800 Subject: [PATCH 5/6] fix(media): refuse a collection title that decodes to blank; an empty PUT answers NO_FIELDS - A title like ' ' or an emoji-only one passed the blank check but stored a blank title once decoded to ASCII; the derived title is checked too. - A supplied alternate title that normalizes to nothing is refused, so it cannot match every poster whose normalized title is empty. - PUT {} reaches the NO_FIELDS answer; only a non-object body keeps the old 500. --- backend/api/media_api.py | 15 ++++++++++----- tests/test_media_api_queries.py | 23 ++++++++++++++++++++--- 2 files changed, 30 insertions(+), 8 deletions(-) diff --git a/backend/api/media_api.py b/backend/api/media_api.py index 827918b81..6bd5a4084 100644 --- a/backend/api/media_api.py +++ b/backend/api/media_api.py @@ -782,13 +782,19 @@ def _collection_fields(payload: dict) -> Tuple[Dict[str, Any], Optional[str]]: if "title" in payload: if not _is_text(payload["title"]): return {}, "Field 'title' must be a non-empty string" - fields.update(collection_title_fields(payload["title"].strip())) + titles = collection_title_fields(payload["title"].strip()) + if not _is_text(titles["title"]): + return {}, "Field 'title' is blank once decoded to ASCII" + fields.update(titles) if "alternate_titles" in payload: alts = payload["alternate_titles"] if not isinstance(alts, list) or not all(_is_text(t) for t in alts): return {}, "Field 'alternate_titles' must be a list of non-empty strings" alts = [t.strip() for t in alts] - fields.update(alternate_titles=alts, normalized_alternate_titles=[normalize_titles(t) for t in alts]) + normalized = [normalize_titles(t) for t in alts] + if not all(normalized): + return {}, "Field 'alternate_titles' has a title that is blank once normalized" + fields.update(alternate_titles=alts, normalized_alternate_titles=normalized) return fields, None @@ -920,9 +926,8 @@ async def update_collection( payload = await read_request_json(request) if payload is BODY_TOO_LARGE: return body_too_large_error() - # An unusable body used to raise into the 500 handler; keep that - # response — falling through would write the record unchanged. - if not isinstance(payload, dict) or not payload: + # A non-object body used to raise into the 500 handler; keep that response. + if not isinstance(payload, dict): return error( "Error updating collection", code="COLLECTION_UPDATE_ERROR", diff --git a/tests/test_media_api_queries.py b/tests/test_media_api_queries.py index 89193f601..1dcf641be 100644 --- a/tests/test_media_api_queries.py +++ b/tests/test_media_api_queries.py @@ -505,16 +505,32 @@ def test_create_collection_refuses_a_body_missing_a_key_field(db, body, code): {"alternate_titles": "Matrix"}, {"alternate_titles": ["Matrix", 7]}, {"alternate_titles": [" "]}, + {"alternate_titles": ["Marvel", " "]}, {"title": 7}, {"title": " "}, + {"title": " "}, + {"title": "\U0001F600"}, {"library_name": ""}, {"year": [2008]}, {"year": True}, {"tmdb_id": 2**63}, {"imdb_id": 7}, ], - ids=["string", "number-in-list", "blank", "numeric-title", "blank-title", "blank-library", "list-year", - "bool-year", "huge-id", "numeric-imdb"], + ids=[ + "string", + "number-in-list", + "blank", + "entity-blank-alternate", + "numeric-title", + "blank-title", + "entity-blank", + "emoji-only", + "blank-library", + "list-year", + "bool-year", + "huge-id", + "numeric-imdb", + ], ) def test_create_collection_refuses_fields_of_the_wrong_shape(db, body): resp = _post_collection(_client(db), **body) @@ -571,9 +587,10 @@ def error(self, msg, *_a, **_kw): ({"title": ""}, 400, "INVALID_FIELDS"), ({"library_name": None}, 400, "INVALID_FIELDS"), ({"name": "Avengers"}, 400, "NO_FIELDS"), + ({}, 400, "NO_FIELDS"), ({"title": "Marvel"}, 409, "COLLECTION_EXISTS"), ], - ids=["blank-title", "null-library", "no-updatable-field", "taken-title"], + ids=["blank-title", "null-library", "no-updatable-field", "empty-object", "taken-title"], ) def test_update_collection_refuses_a_change_it_cannot_store(tmp_path, body, status, code): log = _ErrorLog() From 06193a3566fc7b5e7d954276572538fdf7d366cb Mon Sep 17 00:00:00 2001 From: chodeus <190988615+chodeus@users.noreply.github.com> Date: Thu, 19 Nov 2026 07:27:20 +0800 Subject: [PATCH 6/6] fix(media): collection create checks instance_name's type An object or blank instance_name passed the truthiness check and failed at SQLite binding with a 500; a missing key keeps MISSING_INSTANCE, a present but invalid one answers INVALID_FIELDS. --- backend/api/media_api.py | 10 ++++++++-- tests/test_media_api_queries.py | 7 ++++++- 2 files changed, 14 insertions(+), 3 deletions(-) diff --git a/backend/api/media_api.py b/backend/api/media_api.py index 6bd5a4084..98ad3864c 100644 --- a/backend/api/media_api.py +++ b/backend/api/media_api.py @@ -844,13 +844,19 @@ async def create_collection( "Field 'title' is required", code="MISSING_TITLE", status_code=400 ) - instance_name = payload.get("instance_name") - if not instance_name: + if "instance_name" not in payload: return error( "Field 'instance_name' is required", code="MISSING_INSTANCE", status_code=400, ) + instance_name = payload["instance_name"] + if not _is_text(instance_name): + return error( + "Field 'instance_name' must be a non-empty string", + code="INVALID_FIELDS", + status_code=400, + ) if "library_name" not in payload: return error( diff --git a/tests/test_media_api_queries.py b/tests/test_media_api_queries.py index 1dcf641be..b131e023e 100644 --- a/tests/test_media_api_queries.py +++ b/tests/test_media_api_queries.py @@ -491,8 +491,9 @@ def test_create_collection_stores_the_title_columns_a_sync_stores(tmp_path, titl [ ({"instance_name": "plex1", "library_name": "Movies"}, "MISSING_TITLE"), ({"title": "Marvel", "instance_name": "plex1"}, "MISSING_LIBRARY"), + ({"title": "Marvel", "library_name": "Movies"}, "MISSING_INSTANCE"), ], - ids=["no-title", "no-library"], + ids=["no-title", "no-library", "no-instance"], ) def test_create_collection_refuses_a_body_missing_a_key_field(db, body, code): resp = _client(db).post("/api/media/collections", json=body) @@ -511,6 +512,8 @@ def test_create_collection_refuses_a_body_missing_a_key_field(db, body, code): {"title": " "}, {"title": "\U0001F600"}, {"library_name": ""}, + {"instance_name": {"name": "plex1"}}, + {"instance_name": " "}, {"year": [2008]}, {"year": True}, {"tmdb_id": 2**63}, @@ -526,6 +529,8 @@ def test_create_collection_refuses_a_body_missing_a_key_field(db, body, code): "entity-blank", "emoji-only", "blank-library", + "object-instance", + "blank-instance", "list-year", "bool-year", "huge-id",