[videos] Fixing imdb lookups and making it more modular

vrobbler/apps/music/constants.py

@@ -19,6 +19,8 @@ JELLYFIN_POST_KEYS = {
     "ALBUM_NAME": "Album",
     "ARTIST_NAME": "Artist",
     "STATUS": "Status",
+    "VIDEO_TITLE": "Name",
+    "IMDB_ID": "Provider_imdb",
 }
 
 MOPIDY_POST_KEYS = {

vrobbler/apps/scrobbles/models.py

@@ -769,6 +769,30 @@ class Scrobble(TimeStampedModel):
         if not padding_seconds:
             return is_in_progress
 
+        if not self.media_obj:
+            logger.info(
+                "[scrobbling] scrobble has no media obj",
+                extra={
+                    "media_id": self.media_obj,
+                    "scrobble_id": self.id,
+                    "media_type": self.media_type,
+                    "probably_still_in_progress": is_in_progress,
+                },
+            )
+            return is_in_progress
+
+        if not self.media_obj.run_time_seconds:
+            logger.info(
+                "[scrobbling] media has no run time seconds, cannot calculate end",
+                extra={
+                    "media_id": self.media_obj.id,
+                    "scrobble_id": self.id,
+                    "media_type": self.media_type,
+                    "probably_still_in_progress": is_in_progress,
+                },
+            )
+            return is_in_progress
+
         expected_end = self.timestamp + datetime.timedelta(
             seconds=self.media_obj.run_time_seconds
         )

vrobbler/apps/scrobbles/scrobblers.py

@@ -65,7 +65,7 @@ def jellyfin_scrobble_media(
         media_type = Scrobble.MediaType.TRACK
 
     logger.info(
-        "[jellyfin_scrobble_track] called",
+        "[jellyfin_scrobble_media] called",
         extra={
             "user_id": user_id,
             "post_data": post_data,
@@ -81,32 +81,45 @@ def jellyfin_scrobble_media(
     # Jellyfin has some race conditions with it's webhooks, these hacks fix some of them
     if null_position_on_progress:
         logger.info(
-            "[jellyfin_scrobble_track] no playback position tick, aborting",
+            "[jellyfin_scrobble_media] no playback position tick, aborting",
             extra={"post_data": post_data},
         )
         return
 
+    timestamp = parse(
+        post_data.get(JELLYFIN_POST_KEYS.get("TIMESTAMP"))
+    ).replace(tzinfo=pytz.utc)
+
     if media_type == Scrobble.MediaType.VIDEO:
         media_obj = Video.find_or_create(post_data)
+        playback_position_seconds = (timezone.now() - timestamp).seconds
     else:
         media_obj = get_or_create_track(
             post_data, post_keys=JELLYFIN_POST_KEYS
         )
+        playback_position_seconds = 0
+
+    if not media_obj:
+        logger.info(
+            "[jellyfin_scrobble_media] no video found from POST data",
+            extra={"post_data": post_data},
+        )
+        return
 
-    timestamp = parse(
-        post_data.get(JELLYFIN_POST_KEYS.get("TIMESTAMP"))
-    ).replace(tzinfo=pytz.utc)
     playback_status = "resumed"
     if post_data.get("IsPaused"):
         playback_status = "paused"
     elif post_data.get("NotificationType") == "PlaybackStop":
         playback_status = "stopped"
 
-    # TODO Add some logging here, maybe?
+    logger.info(
+        "[jellyfin_scrobble_media] no playback position tick, aborting",
+        extra={"post_data": post_data, "playback_status": playback_status},
+    )
 
     return media_obj.scrobble_for_user(
         user_id,
-        playback_position_seconds=(timezone.now() - timestamp).seconds,
+        playback_position_seconds=playback_position_seconds,
         status=playback_status,
     )
 
@@ -150,6 +163,17 @@ def manual_scrobble_video_game(hltb_id: str, user_id: int):
     game = VideoGame.objects.filter(hltb_id=hltb_id).first()
     if not game:
         data_dict = lookup_game_from_hltb(hltb_id)
+        if not data_dict:
+            logger.info(
+                "[manual_scrobble_video_game] game not found on hltb",
+                extra={
+                    "hltb_id": hltb_id,
+                    "user_id": user_id,
+                    "media_type": Scrobble.MediaType.VIDEO_GAME,
+                },
+            )
+            return
+
         game = VideoGame.find_or_create(data_dict)
 
     scrobble_dict = {
@@ -295,3 +319,49 @@ def gpslogger_scrobble_location(data_dict: dict, user_id: int) -> Scrobble:
     )
 
     return scrobble
+
+
+def web_scrobbler_scrobble_video_or_song(
+    data_dict: dict, user_id: Optional[int]
+) -> Scrobble:
+    # We're not going to create music tracks, because the only time
+    # we'd hit this is if we're listening to a concert or something.
+    artist_name = data_dict.get("artist")
+    track_name = data_dict.get("track")
+    tracks = Track.objects.filter(
+        artist__name=data_dict.get("artist"), title=data_dict.get("track")
+    )
+    if tracks.count() > 1:
+        logger.warning(
+            "Multiple tracks found for Web Scrobbler",
+            extra={"artist": artist_name, "track": track_name},
+        )
+    track = tracks.first()
+
+    # No track found, create a Video
+    if not track:
+        Video.find_or_create(data_dict)
+
+    # Now we run off a scrobble
+    mopidy_data = {
+        "user_id": user_id,
+        "timestamp": timezone.now(),
+        "playback_position_seconds": data_dict.get("playback_time_ticks"),
+        "source": "Mopidy",
+        "mopidy_status": data_dict.get("status"),
+    }
+
+    logger.info(
+        "[scrobblers] webhook mopidy scrobble request received",
+        extra={
+            "episode_id": episode.id if episode else None,
+            "user_id": user_id,
+            "scrobble_dict": mopidy_data,
+            "media_type": Scrobble.MediaType.PODCAST_EPISODE,
+        },
+    )
+
+    scrobble = None
+    if episode:
+        scrobble = Scrobble.create_or_update(episode, user_id, mopidy_data)
+    return scrobble

vrobbler/apps/videos/imdb.py

@@ -3,19 +3,21 @@ from django.utils import timezone
 
 from imdb import Cinemagoer, helpers
 from imdb.Character import IMDbParserError
+from scrobbles.dataclasses import VideoLogData
 
 imdb_client = Cinemagoer()
 
 logger = logging.getLogger(__name__)
 
 
-def lookup_video_from_imdb(name_or_id: str, kind: str = "movie") -> dict:
+def lookup_video_from_imdb(
+    name_or_id: str, kind: str = "movie"
+) -> VideoLogData:
 
     # Very few video titles start with tt, but IMDB IDs often come in with it
     if name_or_id.startswith("tt"):
         name_or_id = name_or_id[2:]
 
-    video_dict = {}
     imdb_id = None
 
     try:
@@ -23,27 +25,64 @@ def lookup_video_from_imdb(name_or_id: str, kind: str = "movie") -> dict:
     except ValueError:
         pass
 
+    video_metadata = None
     if imdb_id:
         imdb_result = imdb_client.get_movie(name_or_id)
-        video_dict = imdb_result
+        imdb_client.update(imdb_result, info=["plot", "synopsis", "taglines"])
+        video_metadata = imdb_result
 
-    if not video_dict:
+    if not video_metadata:
         imdb_results = imdb_client.search_movie(name_or_id)
         if len(imdb_results) > 1:
             for result in imdb_results:
                 if result["kind"] == kind:
-                    video_dict = result
+                    video_metadata = result
                     break
 
         if len(imdb_results) == 1:
-            video_dict = imdb_results[0]
+            video_metadata = imdb_results[0]
+        imdb_client.update(
+            video_metadata, info=["plot", "synopsis", "taglines"]
+        )
 
-    if not video_dict:
+    if not video_metadata:
-        logger.warn(f"No video found for key {name_or_id}")
+        logger.info(
-        return video_dict
+            f"[lookup_video_from_imdb] no video found on imdb",
-    imdb_client.update(video_dict)
+            extra={"name_or_id": name_or_id},
+        )
+        return video_metadata
 
-    cover_url = video_dict.get("cover url")
+    imdb_client.update(video_metadata)
+
+    cover_url = video_metadata.get("cover url")
     if cover_url:
-        video_dict["cover url"] = helpers.resizeImage(cover_url, width=800)
+        cover_url = helpers.resizeImage(cover_url, width=800)
-    return video_dict
+
+    from videos.models import Video
+
+    video_type = Video.VideoType.MOVIE
+    series_name = None
+    if video_metadata.get("kind") == "episode":
+        series_name = video_metadata.get("episode of", None).data.get(
+            "title", None
+        )
+        video_type = Video.VideoType.TV_EPISODE
+
+    run_time_seconds = 0
+    if video_metadata.get("runtimes"):
+        run_time_seconds = int(video_metadata.get("runtimes")[0]) * 60
+
+    return {
+        "title": video_metadata.get("title"),
+        "imdb_id": video_metadata.get("imdbID"),
+        "video_type": video_type,
+        "run_time_seconds": run_time_seconds,
+        "episode_number": video_metadata.get("episode", None),
+        "season_number": video_metadata.get("season", None),
+        "next_imdb_id": video_metadata.get("next episode", None),
+        "year": video_metadata.get("year", None),
+        "series_name": series_name,
+        "plot": video_metadata.get("plot outline"),
+        "imdb_rating": video_metadata.get("rating"),
+        "cover_url": cover_url,
+    }

vrobbler/apps/videos/models.py

@@ -11,6 +11,7 @@ from django.utils.translation import gettext_lazy as _
 from django_extensions.db.models import TimeStampedModel
 from imagekit.models import ImageSpecField
 from imagekit.processors import ResizeToFit
+from music.constants import JELLYFIN_POST_KEYS
 from scrobbles.mixins import ObjectWithGenres, ScrobblableMixin
 from taggit.managers import TaggableManager
 from videos.imdb import lookup_video_from_imdb
@@ -218,8 +219,17 @@ class Video(ScrobblableMixin):
         if genres := imdb_dict.data.get("genres"):
             self.genre.add(*genres)
 
+    def scrape_cover_from_url(
+        self, cover_url: str, force_update: bool = False
+    ):
+        if not self.cover_image or force_update:
+            r = requests.get(cover_url)
+            if r.status_code == 200:
+                fname = f"{self.title}_{self.uuid}.jpg"
+                self.cover_image.save(fname, ContentFile(r.content), save=True)
+
     @classmethod
-    def find_or_create(cls, data_dict: Dict) -> "Video":
+    def find_or_create(cls, data_dict: Dict) -> Optional["Video"]:
         """Given a data dict from Jellyfin, does the heavy lifting of looking up
         the video and, if need, TV Series, creating both if they don't yet
         exist.
@@ -230,16 +240,9 @@ class Video(ScrobblableMixin):
             get_or_create_video_from_jellyfin,
         )
 
-        if "NotificationType" not in data_dict.keys():
+        video = get_or_create_video(data_dict, JELLYFIN_POST_KEYS)
-            name_or_id = data_dict.get("imdb_id") or data_dict.get("title")
-            video = get_or_create_video(name_or_id)
-            return video
 
-        if not data_dict.get("Provider_imdb"):
+        if not video:
-            title = data_dict.get("Name", "")
-            logger.warn(
-                f"No IMDB ID from Jellyfin, check metadata for {title}"
-            )
             return
 
         return get_or_create_video_from_jellyfin(data_dict)

vrobbler/apps/videos/utils.py

@@ -7,55 +7,61 @@ from scrobbles.utils import convert_to_seconds
 logger = logging.getLogger(__name__)
 
 
-def get_or_create_video(name_or_id: str, force_update=False):
+def get_or_create_video(data_dict: dict, post_keys: dict, force_update=False):
-    imdb_dict = lookup_video_from_imdb(name_or_id)
+    name_or_id = data_dict.get(post_keys.get("IMDB_ID"), "") or data_dict.get(
+        post_keys.get("VIDEO_TITLE"), ""
+    )
+    imdb_metadata = lookup_video_from_imdb(name_or_id)
+    # skatevideosite_metadata = lookup_video_from_skatevideosite(name_or_id)
+    # youtube_metadata = lookup_vide_from_youtube(name_or_id)
 
-    if not imdb_dict:
+    video_dict = imdb_metadata
+    # video_metadata = imdb_metadata or skatevideosite_metadata or youtube_metadata
+    if not video_dict:
+        logger.info(
+            "No video found on imdb, skatevideosite or youtube, cannot scrobble",
+            extra={"name_or_id": name_or_id},
+        )
         return
 
     video, video_created = Video.objects.get_or_create(
-        imdb_id=imdb_dict.get("imdbID"), title=imdb_dict.get("title")
+        imdb_id=video_dict.get("imdb_id"),
+        title=video_dict.get("title"),
     )
-
     if video_created or force_update:
-        video_type = Video.VideoType.MOVIE
         series = None
-        if imdb_dict.get("kind") == "episode":
+        if video_dict.get("video_type") == Video.VideoType.TV_EPISODE:
-            series_name = imdb_dict.get("episode of").data.get("title")
+
+            series_name = video_dict.pop("series_name")
             series, series_created = Series.objects.get_or_create(
                 name=series_name
             )
-            video_type = Video.VideoType.TV_EPISODE
             if series_created:
                 series.fix_metadata()
+            video_dict["tv_series_id"] = series.id
+
+        if genres := video_dict.pop("genres", None):
+            video.genre.add(*genres)
+
+        if cover_url := video_dict.pop("cover_url", None):
+            video.scrape_cover_from_url(cover_url)
 
-        run_time_seconds = 0
-        if imdb_dict.get("runtimes"):
-            run_time_seconds = int(imdb_dict.get("runtimes")[0]) * 60
-        video_dict = {
-            "video_type": video_type,
-            "run_time_seconds": run_time_seconds,
-            "episode_number": imdb_dict.get("episode", None),
-            "season_number": imdb_dict.get("season", None),
-            "next_imdb_id": imdb_dict.get("next episode", None),
-            "tv_series_id": series.id if series else None,
-        }
         Video.objects.filter(pk=video.id).update(**video_dict)
         video.refresh_from_db()
-
-        video.fix_metadata()
-
     return video
 
 
+def get_or_create_video_from_skatevideosite(title: str, force_update=True):
+    ...
+
+
 def get_or_create_video_from_jellyfin(jellyfin_data: dict, force_update=True):
     """Given a Jellyfin webhook payload as a dictionary, lookup the video or
     create a new one.
 
     """
-
     video, video_created = Video.objects.get_or_create(
-        imdb_id=jellyfin_data.get("Provider_imdb").replace("tt", ""),
+        imdb_id=jellyfin_data.get("Provider_imdb", "").replace("tt", ""),
         title=jellyfin_data.get("Name"),
     )