atom: bug fix for de-duping images in attachments

README.md

@@ -311,6 +311,8 @@ Changelog
 ### 2.2 - unreleased
 * Facebook:
   * Add `get_activities(scrape=True)` for scraping HTML from [m.facebook.com](https://m.facebook.com/). Requires `c_user` and `xs` cookies from a logged in session. ([snarfed/bridgy#886](https://github.com/snarfed/bridgy/issues/886)
+* Atom:
+  * Bug fix for de-duping images in attachments.
 
 ### 2.1 - 2019-09-04
 * Convert AS2 `Mention` tags to AS1 `objectType` `mention` (non-standard) and vice versa ([snarfed/bridgy-fed#46](https://github.com/snarfed/bridgy-fed/issues/46)).

granary/atom.py

@@ -417,7 +417,8 @@ def _prepare_activity(a, reader=True):
     parsed = urllib.parse.urlparse(url)
     rest = urllib.parse.urlunparse(('', '') + parsed[2:])
     img_src_re = re.compile(r"""src *= *['"] *((https?:)?//%s)?%s *['"]""" %
-                            (re.escape(parsed.netloc), re.escape(rest)))
+                            (re.escape(parsed.netloc),
+                             _encode_ampersands(re.escape(rest))))
     if (url and url not in image_urls_seen and
         not img_src_re.search(obj['rendered_content'])):
       children.append(microformats2.img(url))

granary/tests/test_atom.py

@@ -750,16 +750,15 @@ def test_image_duplicated_in_content(self):
 
     https://github.com/snarfed/granary/issues/113
     """
-    for url in 'http://pics/1.jpg?foo', '/1.jpg?foo':
-      activity = {
-        'object': {
-          'content': 'foo <img src="%s"> bar' % url,
-          'image': [
-            {"url": "http://pics/1.jpg?foo"},
-            {"url": "http://pics/2.jpg"},
-          ],
-        },
-      }
+    activity = {
+      'object': {
+        'content': 'foo <img src="/1.jpg?foo"> bar',
+        'image': [
+          {"url": "http://pics/1.jpg?foo"},
+          {"url": "http://pics/2.jpg"},
+        ],
+      },
+    }
 
     got = atom.activities_to_atom([activity], {})
     self.assertNotIn('<img class="u-photo" src="http://pics/1.jpg?foo" alt="" />', got)
@@ -778,23 +777,23 @@ def test_image_duplicated_in_attachment(self):
       'object': {
         'content': 'foo bar',
         'image': [
-          {'url': 'http://pics/1.jpg'},
+          {'url': 'http://pics/1.jpg?x&y'},
           {'url': 'http://pics/2.jpg'},
         ],
         'attachments': [{
           'objectType': 'note',
           'image': {'url': 'http://pics/2.jpg'},
         }, {
           'objectType': 'image',
-          'image': {'url': 'http://pics/1.jpg'},
+          'image': {'url': 'http://pics/1.jpg?x&y'},
         }],
       },
     }
 
     got = atom.activities_to_atom([activity], {})
-    self.assertEqual(1, got.count('<img class="u-photo" src="http://pics/1.jpg" alt="" />'), got)
+    self.assertEqual(1, got.count('<img class="u-photo" src="http://pics/1.jpg?x&amp;y" alt="" />'), got)
     self.assert_multiline_in("""
-<link rel="enclosure" href="http://pics/1.jpg" type="image/jpeg" />
+<link rel="enclosure" href="http://pics/1.jpg?x&amp;y" type="" />
 """, got)
     self.assertNotIn('<img class="u-photo" src="http://pics/2.jpg" alt="" />', got, got)