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 crash on banner-less profiles and nested descriptionUrls

Browse Source
This commit is contained in:
JustAnotherArchivist
2020-09-24 01:58:38 +00:00
parent 1a09f9b9a3
commit b78bf3e642
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
snscrape/modules/twitter.py
View File

@@ -362,7 +362,7 @@ class TwitterAPIScraper(snscrape.base.Scraper):
kwargs['id'] = user['id'] if 'id' in user else int(user['id_str']) kwargs['id'] = user['id'] if 'id' in user else int(user['id_str'])
kwargs['description'] = self._render_text_with_urls(user['description'], user['entities']['description']['urls']) kwargs['description'] = self._render_text_with_urls(user['description'], user['entities']['description']['urls'])
kwargs['rawDescription'] = user['description'] kwargs['rawDescription'] = user['description']
kwargs['descriptionUrls'] = [{'text': x['display_url'], 'url': x['expanded_url'], 'tcourl': x['url'], 'indices': tuple(x['indices'])} for x in user['entities']['description']['urls']], kwargs['descriptionUrls'] = [{'text': x['display_url'], 'url': x['expanded_url'], 'tcourl': x['url'], 'indices': tuple(x['indices'])} for x in user['entities']['description']['urls']]
kwargs['verified'] = user['verified'] kwargs['verified'] = user['verified']
kwargs['created'] = email.utils.parsedate_to_datetime(user['created_at']) kwargs['created'] = email.utils.parsedate_to_datetime(user['created_at'])
kwargs['followersCount'] = user['followers_count'] kwargs['followersCount'] = user['followers_count']
@@ -371,7 +371,7 @@ class TwitterAPIScraper(snscrape.base.Scraper):
kwargs['linkUrl'] = user['entities']['url']['urls'][0]['expanded_url'] if 'url' in user['entities'] else None kwargs['linkUrl'] = user['entities']['url']['urls'][0]['expanded_url'] if 'url' in user['entities'] else None
kwargs['linkTcourl'] = user.get('url') kwargs['linkTcourl'] = user.get('url')
kwargs['profileImageUrl'] = user['profile_image_url_https'] kwargs['profileImageUrl'] = user['profile_image_url_https']
kwargs['profileBannerUrl'] = user['profile_banner_url'] kwargs['profileBannerUrl'] = user.get('profile_banner_url')
return User(**kwargs) return User(**kwargs)