Merge pull request #747 from tcely/patch-2

Add and use `refresh_formats`

tubesync/sync/filtering.py

@@ -130,14 +130,7 @@ def filter_max_cap(instance: Media):
         return False
 
     max_cap_age = instance.source.download_cap_date
-    if not max_cap_age:
-        log.debug(
-            f"Media: {instance.source} / {instance} has not max_cap_age "
-            f"so not skipping based on max_cap_age"
-        )
-        return False
-
-    if instance.published <= max_cap_age:
+    if max_cap_age and instance.published <= max_cap_age:
         # log new media instances, not every media instance every time
         if not instance.skip:
             log.info(

tubesync/sync/models.py

@@ -1037,6 +1037,40 @@ class Media(models.Model):
         except Exception as e:
             return {}
 
+    @property
+    def refresh_formats(self):
+        data = self.loaded_metadata
+        metadata_seconds = data.get('epoch', None)
+        if not metadata_seconds:
+            self.metadata = None
+            return False
+
+        now = timezone.now()
+        formats_seconds = data.get('formats_epoch', metadata_seconds)
+        metadata_dt = self.metadata_published(formats_seconds)
+        if (now - metadata_dt) < timedelta(seconds=self.source.index_schedule):
+            return False
+
+        self.skip = False
+        metadata = self.index_metadata()
+        if self.skip:
+            return False
+ 
+        response = metadata
+        if getattr(settings, 'SHRINK_NEW_MEDIA_METADATA', False):
+            response = filter_response(metadata, True)
+
+        field = self.get_metadata_field('formats')
+        data[field] = response.get(field, [])
+        if data.get('availability', 'public') != response.get('availability', 'public'):
+            data['availability'] = response.get('availability', 'public')
+        data['formats_epoch'] = response.get('epoch', formats_seconds)
+
+        from common.utils import json_serial
+        compact_json = json.dumps(data, separators=(',', ':'), default=json_serial)
+        self.metadata = compact_json
+        return True
+
     @property
     def url(self):
         url = self.URLS.get(self.source.source_type, '')

tubesync/sync/tasks.py

@@ -553,6 +553,8 @@ def download_media(media_id):
         err = (f'Failed to download media: {media} (UUID: {media.pk}) to disk, '
                f'expected outfile does not exist: {filepath}')
         log.error(err)
+        # Try refreshing formats
+        media.refresh_formats
         # Raising an error here triggers the task to be re-attempted (or fail)
         raise DownloadFailedException(err)
 
@@ -587,10 +589,25 @@ def save_all_media_for_source(source_id):
         log.error(f'Task save_all_media_for_source(pk={source_id}) called but no '
                   f'source exists with ID: {source_id}')
         return
+
+    already_saved = set()
+    mqs = Media.objects.filter(source=source)
+    refresh_qs = mqs.filter(
+        can_download=False,
+        skip=False,
+        manual_skip=False,
+        downloaded=False,
+    )
+    for media in refresh_qs:
+        media.refresh_formats
+        media.save()
+        already_saved.add(media.uuid)
+
     # Trigger the post_save signal for each media item linked to this source as various
     # flags may need to be recalculated
-    for media in Media.objects.filter(source=source):
-        media.save()
+    for media in mqs:
+        if media.uuid not in already_saved:
+            media.save()
 
 
 @background(schedule=0, remove_existing_tasks=True)

tubesync/sync/youtube.py

@@ -85,6 +85,7 @@ def get_channel_image_info(url):
         'simulate': True,
         'logger': log,
         'extract_flat': True,  # Change to False to get detailed info
+        'check_formats': False,
     })
 
     with yt_dlp.YoutubeDL(opts) as y:
@@ -143,7 +144,11 @@ def get_media_info(url):
         'simulate': True,
         'logger': log,
         'extract_flat': True,
-        'extractor_args': {'youtubetab': {'approximate_date': ['true']}},
+        'check_formats': True,
+        'extractor_args': {
+            'youtube': {'formats': ['missing_pot']},
+            'youtubetab': {'approximate_date': ['true']},
+        },
     })
     response = {}
     with yt_dlp.YoutubeDL(opts) as y:
@@ -248,7 +253,7 @@ def download_media(
         'writeautomaticsub': auto_subtitles,
         'subtitleslangs': sub_langs.split(','),
         'writethumbnail': embed_thumbnail,
-        'check_formats': False,
+        'check_formats': None,
         'overwrites': None,
         'sleep_interval': 10 + int(settings.DOWNLOAD_MEDIA_DELAY / 20),
         'max_sleep_interval': settings.DOWNLOAD_MEDIA_DELAY,