bellingcat/snscrape
1
0
Fork 0
mirror of https://github.com/bellingcat/snscrape.git synced 2026-06-08 02:28:29 +03:00
Code Issues Packages Projects Releases Wiki Activity

Fix t.co card URL replacement on retweets

Browse Source 
Fixes #411
This commit is contained in:
JustAnotherArchivist
2022-04-15 03:18:45 +00:00
parent 247bd82d79
commit 5103a33afa
1 changed files with 7 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
snscrape/modules/twitter.py
View File

@@ -865,14 +865,15 @@ class _TwitterAPIScraper(snscrape.base.Scraper):
kwargs['cashtags'] = [o['text'] for o in tweet['entities']['symbols']]
if card:
kwargs['card'] = card
if hasattr(card, 'url') and '//t.co/' in card.url and 'tcooutlinks' in kwargs:
if hasattr(card, 'url') and '//t.co/' in card.url:
# Try to convert the URL to the non-shortened/t.co one
try:
i = kwargs['tcooutlinks'].index(card.url)
except ValueError:
_logger.warning(f'Could not find card URL in tcooutlinks on tweet {tweetId}')
# Retweets inherit the card but not the outlinks; try to get them from the retweeted tweet instead in that case.
if 'tcooutlinks' in kwargs and card.url in kwargs['tcooutlinks']:
card.url = kwargs['outlinks'][kwargs['tcooutlinks'].index(card.url)]
elif retweetedTweet and retweetedTweet.tcooutlinks and card.url in retweetedTweet.tcooutlinks:
card.url = retweetedTweet.outlinks[retweetedTweet.tcooutlinks.index(card.url)]
else:
card.url = kwargs['outlinks'][i]
_logger.warning(f'Could not translate t.co card URL on tweet {tweetId}')
return Tweet(**kwargs)
def _make_medium(self, medium, tweetId):