diff --git a/atproto.py b/atproto.py
index bc19ebfb..c790ea82 100644
--- a/atproto.py
+++ b/atproto.py
@@ -879,9 +879,17 @@ def fetch_blob(url, blob_field, name, check_size=True, check_type=True):
for facet in ret.get('facets', []):
if feats := facet.get('features'):
if feats[0]['$type'] == 'app.bsky.richtext.facet#link':
- if link := web.Web.load(feats[0]['uri'], metaformats=True,
+ try:
+ link = web.Web.load(feats[0]['uri'], metaformats=True,
authorship_fetch_mf2=False,
- raise_=False):
+ raise_=False)
+ except AssertionError as e:
+ # we probably have an Object already stored for this URL
+ # with source_protocol that's not web
+ logger.warning(e)
+ continue
+
+ if link:
if img := util.get_url(link.as1, 'image'):
props = appview.defs['app.bsky.embed.external#external']['properties']
fetch_blob(img, props, name='thumb',
diff --git a/tests/test_atproto.py b/tests/test_atproto.py
index 85c2ba94..87dad606 100644
--- a/tests/test_atproto.py
+++ b/tests/test_atproto.py
@@ -1535,6 +1535,49 @@ def test_send_note_first_link_preview_embed(self, _, __):
},
}, repo.get_record('app.bsky.feed.post', last_tid), ignore=['facets'])
+ @patch.object(tasks_client, 'create_task', return_value=Task(name='my task'))
+ @patch('requests.get', side_effect=[
+ requests_response(f'A poast',
+ url='http://orig.co/inal'),
+ ])
+ def test_send_note_link_preview_non_web_url(self, mock_get, mock_create_task):
+ user = self.make_user_and_repo()
+
+ obj = Object(id='fake:post', source_protocol='fake', our_as1={
+ **NOTE_AS,
+ 'content': 'My original post',
+ })
+ self.assertTrue(ATProto.send(obj, 'https://bsky.brid.gy'))
+
+ # check repo, record
+ did = user.key.get().get_copy(ATProto)
+ repo = self.storage.load_repo(did)
+ last_tid = arroba.util.int_to_tid(arroba.util._tid_ts_last)
+ self.assert_equals({
+ **NOTE_BSKY,
+ 'bridgyOriginalText': 'My original post',
+ }, repo.get_record('app.bsky.feed.post', last_tid), ignore=['facets'])
+
+ @patch.object(tasks_client, 'create_task', return_value=Task(name='my task'))
+ @patch('requests.get')
+ def test_send_note_link_preview_blocklisted_domain(self, mock_get, __):
+ user = self.make_user_and_repo()
+
+ obj = Object(id='fake:post', source_protocol='fake', our_as1={
+ **NOTE_AS,
+ 'content': 'My original post',
+ })
+ self.assertTrue(ATProto.send(obj, 'https://bsky.brid.gy'))
+
+ # check repo, record
+ did = user.key.get().get_copy(ATProto)
+ repo = self.storage.load_repo(did)
+ last_tid = arroba.util.int_to_tid(arroba.util._tid_ts_last)
+ self.assert_equals({
+ **NOTE_BSKY,
+ 'bridgyOriginalText': 'My original post',
+ }, repo.get_record('app.bsky.feed.post', last_tid), ignore=['facets'])
+
@patch.object(tasks_client, 'create_task', return_value=Task(name='my task'))
@patch('requests.get', side_effect=[
requests_response(f"""\