Also update element lookup cache on mass extract collections with errors

- to avoid direct DB access
- to allow for getter & setter functions
- makes consistency checks on each lookup unneccesary
- propose async consistency checks (TODO)

server/.vscode/settings.json

@@ -0,0 +1,21 @@
+{
+    "python.analysis.typeCheckingMode": "off",
+    "python.linting.enabled": false,
+    "python.linting.mypyArgs": [
+        "--follow-imports=silent",
+        "--ignore-missing-imports",
+        "--show-column-numbers",
+        "--no-pretty",
+        "--warn-return-any",
+        "--disallow-untyped-calls",
+        "--disallow-untyped-defs",
+        "--disallow-incomplete-defs",
+        "--no-implicit-optional",
+    ],
+    "python.linting.mypyEnabled": false,
+    "python.testing.pytestArgs": [
+        "tests",
+    ],
+    "python.testing.pytestEnabled": true,
+    "python.testing.unittestEnabled": false,
+}

server/app.py

@@ -1,3 +1,7 @@
+# TODO check that some queries check DB integrity like:
+# SELECT e.uri FROM mediaelment e LEFT JOIN mediaurimapping m ON e.uri = m.uri WHERE m.uri IS NULL;
+# SELECT e.uri FROM mediacollection e LEFT JOIN collectionurimapping m ON e.uri = m.uri WHERE m.uri IS NULL;
+
 ####
 # Imports
 ####
@@ -64,7 +68,10 @@ from entertainment_decider.extractors.media import (
     media_extract_uri,
     media_update,
 )
-from entertainment_decider.extras import remove_common_trails
+from entertainment_decider.extras import (
+    gen_api_error,
+    remove_common_trails,
+)
 
 
 T = TypeVar("T")
@@ -80,12 +87,29 @@ def adapt_score_list(
     return o
 
 
+def preference_from_base64(in_data: str) -> PreferenceScore[Tag]:
+    return PreferenceScore.from_base64(
+        in_data=in_data,
+        get_tag=lambda i: Tag[i],
+    )
+
+
 ####
 # Logging Config
 ####
 
 
-logging.basicConfig(format="%(asctime)s === %(message)s", level=logging.DEBUG)
+logging.basicConfig(
+    format="%(asctime)s - %(name)s === %(levelname)s: %(message)s",
+    level=logging.DEBUG,
+)
+_LOG_LEVELS = {
+    "httpcore": logging.WARNING,  # >INFO to hide outgoing requests
+    "httpx": logging.WARNING,  # >DEBUG to hide outgoing SSL infos
+    "werkzeug": logging.INFO,  # INFO so incoming requests show up
+}
+for module_name, log_level in  _LOG_LEVELS.items():
+    logging.getLogger(module_name).setLevel(log_level)
 
 
 ####
@@ -636,7 +660,7 @@ def recommend_adaptive() -> ResponseReturnValue:
     preferences = request.cookies.get(
         key=PREFERENCES_SCORE_NAME,
         default=PreferenceScore(),
-        type=PreferenceScore.from_base64,
+        type=preference_from_base64,
     ) * (1 if score_adapt > 0 else -1 if score_adapt < 0 else 0)
     if "max_length" not in request.args:
         # ask for max length before calculating to save time
@@ -684,7 +708,7 @@ def cookies_rating(negative: bool) -> ResponseReturnValue:
     preferences = request.cookies.get(
         key=PREFERENCES_SCORE_NAME,
         default=PreferenceScore(),
-        type=PreferenceScore.from_base64,
+        type=preference_from_base64,
     ).adapt_score(element, score=3 if negative else -3)
     resp = redirect_back_or_okay()
     resp.set_cookie(PREFERENCES_SCORE_NAME, preferences.to_base64())
@@ -730,11 +754,9 @@ def refresh_collections() -> ResponseReturnValue:
                     "collection": {
                         "id": coll.id,
                         "title": coll.title,
-                        "uri": coll.uri,
-                    },
-                    "error": {
-                        "args": repr(e.args),
+                        "uri": coll.primary_uri,
                     },
+                    "error": gen_api_error(e),
                 },
             )
     update_element_lookup_cache(changed_colls)
@@ -914,12 +936,11 @@ def api_collection_extract_mass() -> ResponseReturnValue:
             errors.append(
                 {
                     "uri": u,
-                    "error": {
-                        "type": repr(type(e)),
-                        "args": repr(e.args),
-                    },
+                    "error": gen_api_error(e),
                 }
             )
+    if coll_ids:
+        update_element_lookup_cache(coll_ids)
     if errors:
         return {
             "status": False,
@@ -929,8 +950,6 @@ def api_collection_extract_mass() -> ResponseReturnValue:
                 "data": errors,
             },
         }, 501
-    if coll_ids:
-        update_element_lookup_cache(coll_ids)
     if coll_ids and environ_bool(data.get("redirect_to_overview", False)):
         return redirect(
             "/collection/overview?ids=" + ",".join(str(i) for i in coll_ids)
@@ -1098,10 +1117,7 @@ def api_media_extract_mass() -> ResponseReturnValue:
             errors.append(
                 {
                     "uri": u,
-                    "error": {
-                        "type": repr(type(e)),
-                        "args": repr(e.args),
-                    },
+                    "error": gen_api_error(e),
                 }
             )
     if errors:

server/entertainment_decider/common.py

@@ -5,15 +5,24 @@ import sys
 from typing import (
     IO,
     Iterable,
+    Iterator,
     List,
     Literal,
     Optional,
     Sequence,
+    Tuple,
     TypeVar,
     Union,
+    overload,
 )
 
 
+def all_same(iterable: Iterable) -> bool:
+    it = iter(iterable)
+    first = next(it)
+    return all(first == elem for elem in it)
+
+
 def call(
     args: Sequence[str],
     check: bool = True,
@@ -78,5 +87,75 @@ def limit_iter(iter: Iterable[T], limit: int) -> List[T]:
     return list(itertools.islice(iter, limit))
 
 
+class _IterFixer(Iterator[T]):
+    __it: Iterator[T]
+
+    def __init__(self, it: Iterator[T]) -> None:
+        super().__init__()
+        self.__it = it
+
+    def __iter__(self) -> Iterator[T]:
+        return self
+
+    def __next__(self) -> T:
+        return next(self.__it)
+
+
+def fix_iter(iterable: Iterable[T]) -> Iterable[T]:
+    return _IterFixer(iter(iterable))
+
+
+@overload
+def iter_lookahead(
+    iterable: Iterable[T],
+    get_first: Literal[False] = False,
+    get_last: Literal[False] = False,
+) -> Iterable[Tuple[T, T]]:
+    ...
+
+
+@overload
+def iter_lookahead(
+    iterable: Iterable[T],
+    get_first: Literal[True],
+    get_last: Literal[False] = False,
+) -> Iterable[Tuple[None, T] | Tuple[T, T]]:
+    ...
+
+
+@overload
+def iter_lookahead(
+    iterable: Iterable[T],
+    get_first: Literal[False] = False,
+    get_last: Literal[True] = True,  #  <- default only to satisfy python
+) -> Iterable[Tuple[T, T] | Tuple[T, None]]:
+    ...
+
+
+@overload
+def iter_lookahead(
+    iterable: Iterable[T],
+    get_first: Literal[True],
+    get_last: Literal[True],
+) -> Iterable[Tuple[None, T] | Tuple[T, T] | Tuple[T, None]]:
+    ...
+
+
+def iter_lookahead(
+    iterable: Iterable[T],
+    get_first: bool = False,
+    get_last: bool = False,
+) -> Iterable[Tuple[None, T] | Tuple[T, T] | Tuple[T, None]]:
+    it = iter(iterable)
+    last = next(it)
+    if get_first:
+        yield None, last
+    for cur in it:
+        yield last, cur
+        last = cur
+    if get_last:
+        yield last, None
+
+
 def date_to_datetime(d: date) -> datetime:
     return datetime(d.year, d.month, d.day)

server/entertainment_decider/extractors/all/tvmaze.py

@@ -189,7 +189,9 @@ WEB_CHANNEL_PREFIX = f"{EXTRACTOR_KEY}/web_channel"
 def get_show_tags(show: TvmazeShow) -> Iterable[Tag]:
     yield predefined_series_tag()
     yield get_show_type_tag(show["type"])
-    yield get_language_tag(show["language"])
+    language = show["language"]
+    if language is not None:
+        yield get_language_tag(language)
     for genre in show["genres"]:
         yield get_genre_tag(genre)
     network = show["network"]

server/entertainment_decider/extractors/collection/__init__.py

@@ -43,7 +43,7 @@ def collection_update(
     collection: MediaCollection,
     check_cache_expired: bool = True,
 ) -> ChangedReport:
-    ex = collection_expect_extractor(collection.uri)
+    ex = collection_expect_extractor(collection.primary_uri)
     return ex.update_object(
         object=collection,
         check_cache_expired=check_cache_expired,

server/entertainment_decider/extractors/collection/aggregated.py

@@ -48,7 +48,7 @@ class AggregatedCollectionExtractor(CollectionExtractor[DataType]):
         return True
 
     def _cache_expired(self, object: MediaCollection) -> bool:
-        colls = self.__get_collections(object.uri)
+        colls = self.__get_collections(object.primary_uri)
         for c in colls:
             if c.last_updated is None or object.last_updated <= c.last_updated:
                 return True
@@ -86,7 +86,7 @@ class AggregatedCollectionExtractor(CollectionExtractor[DataType]):
         data: DataType,
     ) -> ChangedReport:
         if object.title is None or "[aggregated]" not in object.title:
-            object.title = f"[aggregated] {object.uri}"
+            object.title = f"[aggregated] {object.primary_uri}"
         object.creator = None
         object.set_watch_in_order_auto(True)
         all_links: Set[int] = set(
@@ -101,6 +101,5 @@ class AggregatedCollectionExtractor(CollectionExtractor[DataType]):
                     episode=episode + 1,
                 )
         orm.delete(link for link in object.media_links if link.element.id in all_links)
-        for uri_link in list(object.uris):
-            uri_link.delete()
+        object.set_as_only_uri(object.primary_uri)
         return ChangedReport.ChangedSome  # TODO improve

server/entertainment_decider/extractors/collection/base.py

@@ -37,14 +37,6 @@ class CollectionExtractor(GeneralExtractor[MediaCollection, T]):
         mapping: CollectionUriMapping = CollectionUriMapping.get(uri=uri)
         if mapping:
             return mapping.element
-        elem: MediaCollection = MediaCollection.get(uri=uri)
-        if elem:
-            logging.warning(
-                f"Add missing URI mapping entry for uri {uri!r}, "
-                + "this should not happen at this point and is considered a bug"
-            )
-            elem.add_single_uri(uri)
-            return elem
         return None
 
     @staticmethod
@@ -116,7 +108,7 @@ class CollectionExtractor(GeneralExtractor[MediaCollection, T]):
             element = extractor.inject_object(data)
         except ExtractionError:
             logging.warning(
-                f"Failed while extracting media {data.object_uri!r} while injecting from {collection.uri!r}",
+                f"Failed while extracting media {data.object_uri!r} while injecting from {collection.primary_uri!r}",
                 exc_info=True,
             )
             return None

server/entertainment_decider/extractors/collection/rss.py

@@ -76,7 +76,7 @@ class RssCollectionExtractor(CollectionExtractor[RSSFeed]):
         object.description = data.description
         object.set_watch_in_order_auto(True)
         object.add_single_uri(
-            self.__get_uri(object.uri)
+            self.__get_uri(object.primary_uri)
         )  # add url without prefix if required
         for item in data.feed:
             element = self._add_episode(

server/entertainment_decider/extractors/collection/tt_rss.py

@@ -80,11 +80,11 @@ class TtRssCollectionExtractor(CollectionExtractor[HeadlineList]):
         data: HeadlineList,
     ) -> ChangedReport:
         if not object.title:
-            object.title = object.uri
+            object.title = object.primary_uri
         object.creator = None
         object.set_watch_in_order_auto(True)
         logging.debug(f"Got {len(data)} headlines")
-        rss_uri = self.__decode_uri(object.uri)
+        rss_uri = self.__decode_uri(object.primary_uri)
         readed_headlines = list[int]()
         for headline in data:
             elem = self._add_episode(collection=object, uri=headline.url)

server/entertainment_decider/extractors/collection/tvmaze.py

@@ -120,7 +120,7 @@ class TvmazeCollectionExtractor(CollectionExtractor[TvmazeShowEmbedded]):
         object: MediaCollection,
         data: TvmazeShowEmbedded,
     ) -> ChangedReport:
-        object.title = f"[tvmaze] {data['name']}"
+        object.title = f"[{self.name}] {data['name']}"
         object.description = data.get("summary", "")
         object.release_date = datetime.strptime(data["premiered"], "%Y-%m-%d")
         object.set_watch_in_order_auto(True)

server/entertainment_decider/extractors/generic.py

@@ -77,14 +77,14 @@ class ExtractedDataLight:
     object_key: str
 
     def create_media(self) -> MediaElement:
-        return MediaElement(
+        return MediaElement.new(
             uri=self.object_uri,
             extractor_name=self.extractor_name,
             extractor_key=self.object_key,
         )
 
     def create_collection(self) -> MediaCollection:
-        return MediaCollection(
+        return MediaCollection.new(
             uri=self.object_uri,
             extractor_name=self.extractor_name,
             extractor_key=self.object_key,
@@ -219,7 +219,7 @@ class GeneralExtractor(Generic[E, T]):
         return self._extract_online(data.object_uri)
 
     def _update_object(self, object: E, data: ExtractedDataOnline[T]) -> ChangedReport:
-        object.uri = data.object_uri
+        object.primary_uri = data.object_uri
         object.tag_list.add(self._get_extractor_tag())
         self._update_object_raw(object, data.data)
         self._update_hook(object, data)
@@ -240,7 +240,7 @@ class GeneralExtractor(Generic[E, T]):
                 f"Skip info for element as already extracted and cache valid: {object.title!r}"
             )
             return ChangedReport.StayedSame
-        data = self._extract_online(object.uri)
+        data = self._extract_online(object.primary_uri)
         logging.debug(f"Updating info for media: {data!r}")
         return self._update_object(object, data)
 

server/entertainment_decider/extractors/media/__init__.py

@@ -28,7 +28,7 @@ def media_expect_extractor(uri: str) -> MediaExtractor:
 
 
 def media_update(element: MediaElement, check_cache_expired: bool = True) -> None:
-    ex = media_expect_extractor(element.uri)
+    ex = media_expect_extractor(element.primary_uri)
     ex.update_object(
         object=element,
         check_cache_expired=check_cache_expired,

server/entertainment_decider/extractors/media/base.py

@@ -1,6 +1,5 @@
 from __future__ import annotations
 
-import logging
 from typing import Optional, TypeVar
 
 from ...models import MediaCollection, MediaElement, MediaUriMapping
@@ -17,7 +16,6 @@ T = TypeVar("T")
 
 
 class MediaExtractor(GeneralExtractor[MediaElement, T]):
-
     # abstract
 
     def _get_author_data(self, data: T) -> Optional[AuthorExtractedData]:
@@ -30,14 +28,6 @@ class MediaExtractor(GeneralExtractor[MediaElement, T]):
         mapping: MediaUriMapping = MediaUriMapping.get(uri=uri)
         if mapping:
             return mapping.element
-        elem: MediaElement = MediaElement.get(uri=uri)
-        if elem:
-            logging.warning(
-                f"Add missing URI mapping entry for uri {uri!r}, "
-                + "this should not happen at this point and is considered a bug"
-            )
-            elem.add_single_uri(uri)
-            return elem
         return None
 
     def _create_object(self, data: ExtractedDataOffline[T]) -> MediaElement:

server/entertainment_decider/extractors/media/tmdb.py

@@ -81,7 +81,7 @@ class TmdbMovieMediaExtractor(MediaExtractor[TmdbMovieData]):
         # sanity check
         if not data.was_released:
             raise ExtractionError(
-                f"Could not extract {object.uri!r} because of missing data probably due to not being released yet"
+                f"Could not extract {object.primary_uri!r} because of missing data probably due to not being released yet"
             )
         # extract data
         object.title = data.title
@@ -98,6 +98,6 @@ class TmdbMovieMediaExtractor(MediaExtractor[TmdbMovieData]):
                 data.tmdb_short_uri,
             )
         )
-        for tag in data.get_tags(data):
+        for tag in data.get_tags():
             object.tag_list.add(tag)
         return ChangedReport.ChangedSome  # TODO improve

server/entertainment_decider/extractors/media/tvmaze.py

@@ -106,7 +106,7 @@ class TvmazeMediaExtractor(MediaExtractor[TvmazeEpisodeEmbedded]):
         airstamp = data.get("airstamp")
         if airstamp is None:  # not released yet
             raise ExtractionError(
-                f"Could not extract {object.uri!r} because of missing data probably due to not being released yet"
+                f"Could not extract {object.primary_uri!r} because of missing data probably due to not being released yet"
             )
         # extract data
         show = data["_embedded"]["show"]

server/entertainment_decider/extractors/media/youtube.py

@@ -112,7 +112,7 @@ class YoutubeMediaExtractor(MediaExtractor[YoutubeVideoData]):
         object.length = int(data["duration"]["secondsText"])
         for tag in get_video_tags(data):
             object.tag_list.add(tag)
-        object.uri = f"https://www.youtube.com/watch?v={data['id']}"
+        object.primary_uri = f"https://www.youtube.com/watch?v={data['id']}"
         object.add_uris(
             (
                 f"https://youtu.be/{data['id']}",

server/entertainment_decider/extras/__init__.py

@@ -1,4 +1,5 @@
 from .chain import Chain
+from .errors import gen_api_error
 from .strings import remove_common_trails
 from .typing import LazyValue
 
@@ -6,5 +7,6 @@ from .typing import LazyValue
 __all__ = [
     "Chain",
     "LazyValue",
+    "gen_api_error",
     "remove_common_trails",
 ]

server/entertainment_decider/extras/errors.py

@@ -0,0 +1,12 @@
+from __future__ import annotations
+
+from traceback import format_exception
+from typing import Dict
+
+
+def gen_api_error(exc: Exception) -> Dict:
+    return {
+        "type": repr(type(exc)),
+        "args": repr(exc.args),
+        "traceback": list(format_exception(exc)),
+    }

server/entertainment_decider/models/entities.py

@@ -189,8 +189,8 @@ class Tag(db.Entity, Tagable, TagProto["Tag"]):
     super_tag_list: Iterable[Tag] = orm.Set(lambda: Tag, reverse="sub_tag_list")
     sub_tag_list: Iterable[Tag] = orm.Set(lambda: Tag, reverse="super_tag_list")
 
-    _collection_list: Iterable[MediaCollection] = orm.Set(lambda: MediaCollection)
-    _media_list: Iterable[MediaElement] = orm.Set(lambda: MediaElement)
+    collection_list: Iterable[MediaCollection] = orm.Set(lambda: MediaCollection)
+    media_list: Iterable[MediaElement] = orm.Set(lambda: MediaElement)
 
     @property
     def orm_assigned_tags(self) -> Query[Tag]:
@@ -202,7 +202,6 @@ class Tag(db.Entity, Tagable, TagProto["Tag"]):
 
 
 class TagKey(db.Entity):
-
     num_id: int = orm.PrimaryKey(int, auto=True)
     tag_key: str = orm.Required(str, unique=True)
     """Format: <domain>[/<kind>][/<id>]
@@ -330,8 +329,9 @@ class MediaElement(db.Entity, UriHolder, Tagable):
         int,
         auto=True,
     )
-    uri: str = orm.Required(
+    __uri: str = orm.Required(
         str,
+        column="uri",
         unique=True,
     )
 
@@ -377,7 +377,7 @@ class MediaElement(db.Entity, UriHolder, Tagable):
     tag_list: Iterable[Tag] = orm.Set(
         lambda: Tag,
     )
-    uris: Iterable[MediaUriMapping] = orm.Set(
+    __uri_list: Iterable[MediaUriMapping] = orm.Set(
         lambda: MediaUriMapping,
     )
     collection_links: Iterable[MediaCollectionLink] = orm.Set(
@@ -393,21 +393,35 @@ class MediaElement(db.Entity, UriHolder, Tagable):
         reverse="blocked_by",
     )
 
+    @classmethod
+    def new(
+        cls,
+        *,
+        extractor_name: str,
+        extractor_key: str,
+        uri: str,
+    ) -> MediaElement:
+        return cls(
+            extractor_name=extractor_name,
+            extractor_key=extractor_key,
+            _MediaElement__uri=uri,  # manual mangling for MediaElement
+        )
+
     ### for UriHolder
 
     @property
     def _primary_uri(self) -> str:
-        return self.uri
+        return self.__uri
 
     def _set_primary_uri(self, uri: str) -> None:
-        self.uri = uri
+        self.__uri = uri
 
     @property
-    def _get_uri_set(self) -> Set[str]:
-        return {m.uri for m in self.uris}
+    def _uri_set(self) -> Set[str]:
+        return {m.uri for m in self.__uri_list}
 
     def _set_uri_set(self, uri_set: Set[str]) -> None:
-        self.uris = set()
+        self.__uri_list = set()
         self.add_uris(uri_set)
 
     ### for Tagable
@@ -530,7 +544,7 @@ class MediaElement(db.Entity, UriHolder, Tagable):
             other.ignored = True
         if self.progress >= 0 and other.progress <= 0:
             other.progress = self.progress
-        for uri_map in self.uris:
+        for uri_map in self.__uri_list:
             uri_map.element = other
         for link in self.collection_links:
             if not MediaCollectionLink.get(collection=link.collection, element=other):
@@ -557,7 +571,7 @@ class MediaElement(db.Entity, UriHolder, Tagable):
         self.before_update()
 
     def before_update(self) -> None:
-        self.add_single_uri(self.uri)
+        self.add_single_uri(self.__uri)
 
 
 class MediaThumbnail(db.Entity):
@@ -640,8 +654,9 @@ class MediaCollection(db.Entity, UriHolder, Tagable):
         int,
         auto=True,
     )
-    uri: str = orm.Required(
+    __uri: str = orm.Required(
         str,
+        column="uri",
         unique=True,
     )
 
@@ -697,7 +712,7 @@ class MediaCollection(db.Entity, UriHolder, Tagable):
     tag_list: Iterable[Tag] = orm.Set(
         lambda: Tag,
     )
-    uris: Iterable[CollectionUriMapping] = orm.Set(
+    __uri_set: Iterable[CollectionUriMapping] = orm.Set(
         lambda: CollectionUriMapping,
     )
     media_links: Iterable[MediaCollectionLink] = orm.Set(
@@ -707,21 +722,35 @@ class MediaCollection(db.Entity, UriHolder, Tagable):
         lambda: MediaCollection,
     )
 
+    @classmethod
+    def new(
+        cls,
+        *,
+        extractor_name: str,
+        extractor_key: str,
+        uri: str,
+    ) -> MediaCollection:
+        return cls(
+            extractor_name=extractor_name,
+            extractor_key=extractor_key,
+            _MediaCollection__uri=uri,  # manual mangling for MediaCollection
+        )
+
     ### for UriHolder
 
     @property
     def _primary_uri(self) -> str:
-        return self.uri
+        return self.__uri
 
     def _set_primary_uri(self, uri: str) -> None:
-        self.uri = uri
+        self.__uri = uri
 
     @property
-    def _get_uri_set(self) -> Set[str]:
-        return {m.uri for m in self.uris}
+    def _uri_set(self) -> Set[str]:
+        return {m.uri for m in self.__uri_set}
 
     def _set_uri_set(self, uri_set: Set[str]) -> None:
-        self.uris = set()
+        self.__uri_set = set()
         self.add_uris(uri_set)
 
     ### for Tagable
@@ -917,7 +946,7 @@ class MediaCollection(db.Entity, UriHolder, Tagable):
         self.before_update()
 
     def before_update(self) -> None:
-        self.add_single_uri(self.uri)
+        self.add_single_uri(self.__uri)
 
 
 class CollectionUriMapping(db.Entity):

server/entertainment_decider/models/extras/uris.py

@@ -5,7 +5,6 @@ from typing import Iterable, Optional, Set
 
 
 class UriHolder:
-
     ### abstracted
 
     @abstractproperty
@@ -17,7 +16,7 @@ class UriHolder:
         """Sets the primary uri of this object in a naive way."""
 
     @abstractproperty
-    def _get_uri_set(self) -> Set[str]:
+    def _uri_set(self) -> Set[str]:
         """Returns the uri set of this object in a naive way."""
 
     @abstractmethod
@@ -45,6 +44,17 @@ class UriHolder:
         """Returns the current primary uri of this object."""
         return self._primary_uri
 
+    @primary_uri.setter
+    def primary_uri(self, uri: str) -> None:
+        self.set_primary_uri(uri)
+
+    @property
+    def uri_set(self) -> Set[str]:
+        return self._uri_set
+
+    # uri_set has no setter due to the problem which uri then becomes primary
+    # instead, set_as_only_uri & add_uris should be used so the primary becomes obvious
+
     def is_primary_uri(self, compare_uri: str) -> bool:
         """Returns True if the given uri is equal to the current primary uri."""
         return self.primary_uri == compare_uri
@@ -54,6 +64,8 @@ class UriHolder:
 
         It will also add the uri to the uri set.
         Returns True if the uri was not in the uri set before.
+
+        You may also just write the primary_uri property if you do not need the return value.
         """
         ret = self._add_uri_to_set(uri)  # might fail, so try first
         self._set_primary_uri(uri)
@@ -68,3 +80,9 @@ class UriHolder:
 
     def add_uris(self, uri_list: Iterable[Optional[str]]) -> bool:
         return any([self.add_single_uri(uri) for uri in set(uri_list) if uri])
+
+    def remove_single_uri(self, uri: str) -> bool:
+        return self._remove_uri_from_set(uri)
+
+    def remove_uris(self, uri_list: Iterable[Optional[str]]) -> bool:
+        return any([self.remove_single_uri(uri) for uri in set(uri_list) if uri])

server/entertainment_decider/preferences/__init__.py

@@ -8,3 +8,12 @@ from .tag_scoring import (
     PreferenceScoreCompatible,
     PreferenceScoreSuper,
 )
+
+
+__all__ = [
+    "PreferenceScore",
+    "PreferenceScoreAppender",
+    "PreferenceScoreCompatible",
+    "PreferenceScoreSuper",
+    "generate_preference_list",
+]

server/templates/collection_element.htm

@@ -88,8 +88,8 @@
         <pre>{{ collection.notes or "" }}</pre>
         <h2>Links</h2>
         <ul>
-            {% for link in collection.uris|sort(attribute="uri") %}
-                <li>{{ link.uri | as_link }} {% if collection.uri == link.uri %}*{% endif %}</li>
+            {% for link in collection.uri_set|sort %}
+                <li>{{ link | as_link }} {% if collection.is_primary_uri(link) %}*{% endif %}</li>
             {% endfor %}
         </ul>
         {% if collection.created_collections %}

server/templates/macros.htm

@@ -358,7 +358,7 @@
     {# TODO do not hardcode certain extractors here #}
     {% if element.extractor_name in ["ytdl", "youtube"] %}
         {%- set opts = {
-            "video_uri": element.uri,
+            "video_uri": element.primary_uri,
             "start": element.progress,
         } -%}
         <a class="button play_button" href="entertainment-decider:///player/play?{{ opts | encode_options }}">{{ symbol | safe }}</a>

server/templates/media_element.htm

@@ -77,8 +77,8 @@
                 </ul>
                 <h2>Links</h2>
                 <ul>
-                    {% for link in element.uris|sort(attribute="uri") %}
-                        <li>{{ link.uri | as_link }} {% if element.uri == link.uri %}*{% endif %}</li>
+                    {% for link in element.uri_set|sort %}
+                        <li>{{ link | as_link }} {% if element.is_primary_uri(link) %}*{% endif %}</li>
                     {% endfor %}
                 </ul>
                 <h2>Blocked By</h2>