From deadfb2bf887e5b1bc1ca7db548a17e62afa9ad4 Mon Sep 17 00:00:00 2001
From: Toorero <22551563+Toorero@users.noreply.github.com>
Date: Sun, 31 Oct 2021 22:32:11 +0100
Subject: [PATCH] Simplified default regex

---
 PFERD/crawl/kit_ipd_crawler.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/PFERD/crawl/kit_ipd_crawler.py b/PFERD/crawl/kit_ipd_crawler.py
index 2e47190..689cca5 100644
--- a/PFERD/crawl/kit_ipd_crawler.py
+++ b/PFERD/crawl/kit_ipd_crawler.py
@@ -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"))