fix: add timeout to get_url http request

commonknowledge · Oct 3, 2024 · 43ab4e3 · 43ab4e3
1 parent d50d5e5
commit 43ab4e3
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 1 deletion.
diff --git a/wagtaillinkchecker/scanner.py b/wagtaillinkchecker/scanner.py
@@ -64,7 +64,7 @@ def get_url(url, page, site):
     }
     response = None
     try:
-        response = requests.get(url, verify=True)
+        response = requests.get(url, verify=True, timeout=30)
         data["response"] = response
     except (requests.exceptions.InvalidSchema, requests.exceptions.MissingSchema):
         data["invalid_schema"] = True

diff --git a/wagtaillinkchecker/tasks.py b/wagtaillinkchecker/tasks.py
@@ -22,8 +22,12 @@ def check_link(
 def check_link_sync(link_pk, verbosity=1):
     link = ScanLink.objects.get(pk=link_pk)
     site = link.scan.site
+    if verbosity > 1:
+        print(f"Checking {link.url}")
     url = get_url(link.url, link.page, site)
     link.status_code = url.get("status_code")
+    if verbosity > 1:
+        print(f"Link is {'broken' if url['error'] else 'OK'}")
 
     if url["error"]:
         link.broken = True