tubearchivist/tubearchivist
1
0
Fork 0
mirror of https://github.com/tubearchivist/tubearchivist.git synced 2024-12-25 03:00:13 +00:00
Code Issues Projects Releases Wiki Activity

fix channel extraction to catch all alerts, #312

Browse Source
This commit is contained in:
simon 2022-09-06 23:12:49 +07:00
parent 3c92de8e24
commit a6dd927860
No known key found for this signature in database
GPG Key ID: 2C15AA5E89985DD4
1 changed files with 8 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
tubearchivist/home/src/index/channel.py
View File

@ -73,14 +73,15 @@ class ChannelScraper:
def _is_deactivated(self): def _is_deactivated(self):
"""check if channel is deactivated""" """check if channel is deactivated"""
alert_text = "This channel does not exist."
alerts = self.yt_json.get("alerts") alerts = self.yt_json.get("alerts")
if alerts and alert_text in str(alerts): if not alerts:
print(f"{self.channel_id}: {alert_text}")
return True
return False return False
for alert in alerts:
alert_text = alert["alertRenderer"]["text"]["simpleText"]
print(f"{self.channel_id}: failed to extract, {alert_text}")
return True
def _parse_channel_main(self): def _parse_channel_main(self):
"""extract maintab values from scraped channel json data""" """extract maintab values from scraped channel json data"""
main_tab = self.yt_json["header"]["c4TabbedHeaderRenderer"] main_tab = self.yt_json["header"]["c4TabbedHeaderRenderer"]
@ -226,8 +227,8 @@ class YoutubeChannel(YouTubeItem):
self.json_data.update( self.json_data.update(
{ {
"channel_subs": content["channel_follower_count"], "channel_subs": content.get("channel_follower_count", 0),
"channel_description": content["description"], "channel_description": content.get("description", False),
} }
) )