From: shamoon <4887959+shamoon@users.noreply.github.com>
Date: Sat, 1 Nov 2025 16:37:49 +0000 (-0700)
Subject: Chore: add headers for wikipedia CI tests (#11253)
X-Git-Tag: v2.19.4~13
X-Git-Url: http://git.ipfire.org/gitweb.cgi?a=commitdiff_plain;h=cffb9c34f0fd66f0ae0acfc678b3447f54b75435;p=thirdparty%2Fpaperless-ngx.git

Chore: add headers for wikipedia CI tests (#11253)
---

diff --git a/src/paperless_mail/tests/test_parsers_live.py b/src/paperless_mail/tests/test_parsers_live.py
index e1febb1e5d..2c3ef262aa 100644
--- a/src/paperless_mail/tests/test_parsers_live.py
+++ b/src/paperless_mail/tests/test_parsers_live.py
@@ -53,6 +53,15 @@ class TestUrlCanary:
     Verify certain URLs are still available so testing is valid still
     """
 
+    # Wikimedia rejects requests without a browser-like User-Agent header and returns 403.
+    _WIKIMEDIA_HEADERS = {
+        "User-Agent": (
+            "Mozilla/5.0 (X11; Linux x86_64) "
+            "AppleWebKit/537.36 (KHTML, like Gecko) "
+            "Chrome/123.0.0.0 Safari/537.36"
+        ),
+    }
+
     def test_online_image_exception_on_not_available(self):
         """
         GIVEN:
@@ -70,6 +79,7 @@ class TestUrlCanary:
         with pytest.raises(httpx.HTTPStatusError) as exec_info:
             resp = httpx.get(
                 "https://upload.wikimedia.org/wikipedia/en/f/f7/nonexistent.png",
+                headers=self._WIKIMEDIA_HEADERS,
             )
             resp.raise_for_status()
 
@@ -90,7 +100,10 @@ class TestUrlCanary:
         """
 
         # Now check the URL used in samples/sample.html
-        resp = httpx.get("https://upload.wikimedia.org/wikipedia/en/f/f7/RickRoll.png")
+        resp = httpx.get(
+            "https://upload.wikimedia.org/wikipedia/en/f/f7/RickRoll.png",
+            headers=self._WIKIMEDIA_HEADERS,
+        )
         resp.raise_for_status()