[BUGFIX] Fix source metadata getting fetched when yt-dlp breaks

This commit is contained in:
Jesse Bannon 2024-01-09 17:39:28 -08:00
parent be0f6f2d2d
commit 1ffbb814e8

View file

@ -216,15 +216,12 @@ class YTDLP:
**kwargs
arguments passed directory to YoutubeDL extract_info
"""
parent_dict: Dict = {}
try:
with cls._listen_and_log_downloaded_info_json(
working_directory=working_directory, log_prefix=log_prefix_on_info_json_dl
):
# TODO: Fetch parent_dict if it breaks
parent_dict = cls.extract_info(
ytdl_options_overrides=ytdl_options_overrides, **kwargs
)
cls.extract_info(ytdl_options_overrides=ytdl_options_overrides, **kwargs)
except RejectedVideoReached:
cls.logger.debug(
"RejectedVideoReached, stopping additional downloads "
@ -238,23 +235,32 @@ class YTDLP:
except MaxDownloadsReached:
cls.logger.info("MaxDownloadsReached, stopping additional downloads.")
# For YouTube playlists in particular, channel metadata is not fetched. Attempt to get
# channel metadata via grabbing uploader_url info json a max of 3 times
if isinstance(parent_dict, dict):
current_iter = 0
url = kwargs.get("url")
uploader_url = parent_dict.get("uploader_url")
while current_iter < 3 and uploader_url and url != uploader_url:
cls.logger.debug("Attempting to get parent metadata from URL %s", uploader_url)
parent_dicts: List[Dict] = []
entry_dicts = cls._get_entry_dicts_from_info_json_files(working_directory=working_directory)
entry_ids = {entry_dict.get("id") for entry_dict in entry_dicts}
# Try to get additional uploader (source) metadata that yt-dlp does not fetch
# in a single request
for entry_dict in entry_dicts:
if not (uploader_id := entry_dict.get("uploader_id")):
continue
if uploader_id in entry_ids or not (uploader_url := entry_dict.get("uploader_url")):
continue
cls.logger.debug("Attempting to get parent metadata from URL %s", uploader_url)
try:
parent_dict = cls.extract_info(
ytdl_options_overrides=ytdl_options_overrides | {"playlist_items": "0:0"},
url=uploader_url,
)
except Exception: # pylint: disable=broad-except
# Do not try this uploader_id again
entry_ids.add(uploader_id)
break
if not isinstance(parent_dict, dict):
break
current_iter += 1
url = uploader_url
uploader_url = parent_dict.get("uploader_url")
if isinstance(parent_dict, dict):
parent_dicts.append(parent_dict)
entry_ids |= {uploader_id, parent_dict.get("id")}
return cls._get_entry_dicts_from_info_json_files(working_directory=working_directory)
return entry_dicts + parent_dicts