bellingcat/snscrape
1
0
Fork 0
mirror of https://github.com/bellingcat/snscrape.git synced 2026-06-11 03:48:29 +03:00
Code Issues Packages Projects Releases Wiki Activity

fix(vkontakte): update photo detection

Browse Source
This commit is contained in:
hgrsd
2022-07-16 18:27:02 +01:00
parent d72b51953f
commit 279d1cf4a1
1 changed files with 4 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
snscrape/modules/vkontakte.py
View File

@@ -117,6 +117,9 @@ class VKontakteUserScraper(snscrape.base.Scraper):
return urllib.parse.unquote(a['href'][13 : end])
return None
def is_photo(self, a):
return 'aria-label' in a.attrs and a.attrs['aria-label'].startswith('photo')
def _date_span_to_date(self, dateSpan):
if not dateSpan:
return None
@@ -172,7 +175,7 @@ class VKontakteUserScraper(snscrape.base.Scraper):
not (not isCopy and thumbsDiv.parent.name == 'div' and 'class' in thumbsDiv.parent.attrs and 'copy_quote' in thumbsDiv.parent.attrs['class']): # Skip post quotes
photos = []
for a in thumbsDiv.find_all('a', class_ = 'page_post_thumb_wrap'):
if 'data-photo-id' not in a.attrs and 'data-video' not in a.attrs:
if not self.is_photo(a) and 'data-video' not in a.attrs:
_logger.warning(f'Skipping non-photo and non-video thumb wrap on {url}')
continue
if 'data-video' in a.attrs: