Triplehx3/contentdb
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Link checker: display original href rather than resolved URL

Browse Source
This commit is contained in:
rubenwardy
2025-10-05 16:29:53 +01:00
parent a604b3cd09
commit 8bafaed671
2 changed files with 8 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
app/markdown/__init__.py
View File

@@ -15,7 +15,6 @@
# along with this program. If not, see <https://www.gnu.org/licenses/>.
from typing import Sequence
from urllib.parse import urljoin
from bs4 import BeautifulSoup
from jinja2.utils import markupsafe
from markdown_it import MarkdownIt
@@ -107,7 +106,7 @@ def get_user_mentions(html: str) -> set:
return set([x.get("data-username") for x in links])
def get_links(html: str, url: str) -> set:
def get_links(html: str) -> set:
soup = BeautifulSoup(html, "html.parser")
links = soup.select("a[href]")
return set([urljoin(url, x.get("href")) for x in links])
return set([x.get("href") for x in links])