bluesky.from_as1: if post text gets truncated, add an external embed …

…to the original post

for snarfed/bridgy-fed#986

README.md

@@ -305,7 +305,7 @@ Changelog
   * `from_as1`:
     * Add hashtag, mention, block, and flag support. Interpret `tags` with missing `objectType` as hashtags.
     * Guess missing indices in facets based on content text. Otherwise, if we still don't know a facet's indices, discard it.
-    * If an output string value is longer than its ``maxGraphemes`` or ``maxLength`` in its lexicon, it's truncated with an ``…`` ellipsis character at the end in order to fit.
+    * If an output string value is longer than its ``maxGraphemes`` or ``maxLength`` in its lexicon, truncate it with an ``…`` ellipsis character at the end in order to fit. If this happens to post text, include a link embed pointing to the original post.
     * Populate `reply.root` properly in reply posts ([snarfed/bridgy#1696](https://github.com/snarfed/bridgy/issues/1696)).
     * Support `lexrpc.Client` as well as `Bluesky` for `client` kwarg.
   * `from_as1_to_strong_ref`:

granary/bluesky.py

@@ -549,8 +549,9 @@ def from_as1(obj, out_type=None, blobs=None, client=None):
     # convert text from HTML and truncate
     src = Bluesky('unused')
     content = obj.get('content')
-    text = obj.get('summary') or content or obj.get('name') or ''
-    text = src.truncate(html_to_text(text), None, OMIT_LINK)
+    full_text = html_to_text(obj.get('summary') or content or obj.get('name') or '')
+    text = src.truncate(full_text, None, OMIT_LINK)
+    truncated = text != full_text
 
     facets = []
     # convert index-based tags to facets
@@ -667,7 +668,17 @@ def from_as1(obj, out_type=None, blobs=None, client=None):
     # article/note attachments
     record_embed = record_record_embed = external_embed = external_record_embed = None
 
-    for att in util.get_list(obj, 'attachments'):
+    attachments = util.get_list(obj, 'attachments')
+    if truncated:
+      if url := as1.get_url(obj):
+        # override attachments, use one link to original post instead
+        attachments = [{
+          'objectType': 'link',
+          'url': url,
+          'displayName': f'Original post on {util.domain_from_link(url)}',
+        }]
+
+    for att in attachments:
       if not att.get('objectType') in ('article', 'link', 'note'):
         continue
 

granary/tests/test_bluesky.py

@@ -750,6 +750,27 @@ def test_from_as1_post_without_tag_indices(self):
     # no facet
     self.assert_equals(POST_BSKY, from_as1(post_as))
 
+  @patch.object(Bluesky, 'TRUNCATE_TEXT_LENGTH', 12)
+  def test_from_as1_post_truncate_adds_link_embed(self):
+    self.assert_equals({
+      '$type': 'app.bsky.feed.post',
+      'text': 'more than…',
+      'createdAt': '2022-01-02T03:04:05.000Z',
+      'embed': {
+        '$type': 'app.bsky.embed.external',
+        'external': {
+          '$type': 'app.bsky.embed.external#external',
+          'description': '',
+          'title': 'Original post on my.inst',
+          'uri': 'http://my.inst/post',
+        },
+      },
+    }, from_as1({
+      'objectType': 'note',
+      'url': 'http://my.inst/post',
+      'content': 'more than ten chars long',
+    }))
+
   def test_from_as1_tag_without_url(self):
     self.assert_equals(POST_BSKY, from_as1({
       **POST_AS,