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

Refactor to tweetId variable

Browse Source
This commit is contained in:
JustAnotherArchivist
2022-04-15 03:14:29 +00:00
parent 5fc67f2bcf
commit 247bd82d79
1 changed files with 5 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
snscrape/modules/twitter.py
View File

@@ -798,8 +798,9 @@ class _TwitterAPIScraper(snscrape.base.Scraper):
return tweet['id'] if 'id' in tweet else int(tweet['id_str']) return tweet['id'] if 'id' in tweet else int(tweet['id_str'])
def _make_tweet(self, tweet, user, retweetedTweet = None, quotedTweet = None, card = None): def _make_tweet(self, tweet, user, retweetedTweet = None, quotedTweet = None, card = None):
tweetId = self._get_tweet_id(tweet)
kwargs = {} kwargs = {}
kwargs['id'] = self._get_tweet_id(tweet) kwargs['id'] = tweetId
kwargs['content'] = tweet['full_text'] kwargs['content'] = tweet['full_text']
kwargs['renderedContent'] = self._render_text_with_urls(tweet['full_text'], tweet['entities'].get('urls')) kwargs['renderedContent'] = self._render_text_with_urls(tweet['full_text'], tweet['entities'].get('urls'))
kwargs['user'] = user kwargs['user'] = user
@@ -807,7 +808,7 @@ class _TwitterAPIScraper(snscrape.base.Scraper):
if tweet['entities'].get('urls'): if tweet['entities'].get('urls'):
kwargs['outlinks'] = [u['expanded_url'] for u in tweet['entities']['urls']] kwargs['outlinks'] = [u['expanded_url'] for u in tweet['entities']['urls']]
kwargs['tcooutlinks'] = [u['url'] for u in tweet['entities']['urls']] kwargs['tcooutlinks'] = [u['url'] for u in tweet['entities']['urls']]
kwargs['url'] = f'https://twitter.com/{user.username}/status/{kwargs["id"]}' kwargs['url'] = f'https://twitter.com/{user.username}/status/{tweetId}'
kwargs['replyCount'] = tweet['reply_count'] kwargs['replyCount'] = tweet['reply_count']
kwargs['retweetCount'] = tweet['retweet_count'] kwargs['retweetCount'] = tweet['retweet_count']
kwargs['likeCount'] = tweet['favorite_count'] kwargs['likeCount'] = tweet['favorite_count']
@@ -822,7 +823,7 @@ class _TwitterAPIScraper(snscrape.base.Scraper):
if 'extended_entities' in tweet and 'media' in tweet['extended_entities']: if 'extended_entities' in tweet and 'media' in tweet['extended_entities']:
media = [] media = []
for medium in tweet['extended_entities']['media']: for medium in tweet['extended_entities']['media']:
if (mediumO := self._make_medium(medium, kwargs['id'])): if (mediumO := self._make_medium(medium, tweetId)):
media.append(mediumO) media.append(mediumO)
if media: if media:
kwargs['media'] = media kwargs['media'] = media
@@ -869,7 +870,7 @@ class _TwitterAPIScraper(snscrape.base.Scraper):
try: try:
i = kwargs['tcooutlinks'].index(card.url) i = kwargs['tcooutlinks'].index(card.url)
except ValueError: except ValueError:
_logger.warning(f'Could not find card URL in tcooutlinks on tweet {kwargs["id"]}') _logger.warning(f'Could not find card URL in tcooutlinks on tweet {tweetId}')
else: else:
card.url = kwargs['outlinks'][i] card.url = kwargs['outlinks'][i]
return Tweet(**kwargs) return Tweet(**kwargs)