Garmelon/PFERD
1
0
Fork 0
mirror of https://github.com/Garmelon/PFERD.git synced 2026-04-12 15:35:05 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

Simplified default regex

Browse source
This commit is contained in:
Toorero 2021-10-31 22:32:11 +01:00
parent 13d4835d70
commit deadfb2bf8
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
PFERD/crawl/kit_ipd_crawler.py
View file

@ -125,7 +125,7 @@ class KitIpdCrawler(HttpCrawler):
def _fetch_file_regex(self) -> re.Pattern:
if "link_regex" in self._config:
return re.compile(self._config["link_regex"])
return re.compile(".*\/[^\/]*\.(?:(?:pdf)|(?:zip)|(?:c)|(?:java))")
return re.compile(".*\/[^\/]*\.(?:pdf|zip|c|java)")
def _abs_url_from_link(self, link_tag: Tag) -> str:
return urljoin(self._url, link_tag.get("href"))