Web.fetch: handle representative h-card URL with and without trailing…

… slash
snarfed · Oct 25, 2024 · 05b6e33 · 05b6e33
1 parent 6e59248
commit 05b6e33
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 3 deletions.
diff --git a/tests/test_web.py b/tests/test_web.py
@@ -2598,6 +2598,19 @@ def test_verify_both_work(self, mock_get, _):
             ],
         })
 
+    def test_verify_representative_hcard_url_without_trailing_slash(self, mock_get, _):
+        hcard = requests_response("""
+<html><body class="h-card">
+  <a class="u-url" href="https://user.com"></a>
+</body></html>""",
+            url='https://user.com/',
+        )
+        mock_get.side_effect = [FULL_REDIR, hcard]
+        self._test_verify(True, True, {
+            'objectType': 'person',
+            'url': 'https://user.com',
+        })
+
     def test_verify_www_redirect(self, mock_get, _):
         www_user = self.make_user('www.user.com', cls=Web)
 

diff --git a/web.py b/web.py
@@ -500,7 +500,7 @@ def fetch(cls, obj, gateway=False, check_backlink=False,
                             else None)
         if metaformats is None:
             # default to only for homepages
-            metaformats = urlparse(url).path in ('', '/')
+            metaformats = is_homepage
 
         try:
             parsed = util.fetch_mf2(url, gateway=gateway, metaformats=metaformats,
@@ -515,9 +515,12 @@ def fetch(cls, obj, gateway=False, check_backlink=False,
         # find mf2 item
         if is_homepage:
             logger.info(f"{url} is user's web url")
-            entry = mf2util.representative_hcard(parsed, parsed['url'])
+            parsed_url = (parsed['url'] or '').rstrip('/')
+            # try both with and without trailing slash
+            entry = (mf2util.representative_hcard(parsed, parsed_url)
+                     or mf2util.representative_hcard(parsed, parsed_url + '/'))
             if not entry:
-                error(f"Couldn't find a representative h-card (http://microformats.org/wiki/representative-hcard-parsing) on {parsed['url']}")
+                error(f"Couldn't find a representative h-card (http://microformats.org/wiki/representative-h-card-parsing) on {parsed['url']}")
             logger.info(f'Found representative h-card')
         else:
             entry = mf2util.find_first_entry(parsed, ['h-entry'])