simon
/
wp-spider
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

improved image dedection from lightbox gallery a tag

This commit is contained in:
simon 2021-03-29 12:02:07 +07:00
parent b70f1e6cd6
commit 794a28398a
1 changed files with 7 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
src/parse_html.py
View File

@ -56,10 +56,16 @@ def get_images(soup, config):
img_url_set.add(url) img_url_set.add(url)
except: except:
continue continue
# gallery item from a tag
all_a_tag = soup.find_all('a')
for a_tag in all_a_tag:
a_href = a_tag['href']
if upload_folder in a_href:
img_url_set.add(a_href)
# external # external
all_external_css = soup.find_all("link", {"rel": "stylesheet"}) all_external_css = soup.find_all("link", {"rel": "stylesheet"})
for css_file in all_external_css: for css_file in all_external_css:
remote_file = all_external_css[0]["href"] remote_file = css_file["href"]
try: try:
remote_css = requests.get(remote_file).text remote_css = requests.get(remote_file).text
except ConnectionError: except ConnectionError: