Render JSON Feed entries’ content with render_content

granary/jsonfeed.py

@@ -2,7 +2,7 @@
 
 JSON Feed spec: https://jsonfeed.org/version/1
 """
-from __future__ import unicode_literals
+from __future__ import absolute_import, unicode_literals
 from builtins import str
 from past.builtins import basestring
 
@@ -11,12 +11,11 @@
 import mf2util
 from oauth_dropins.webutil import util
 
+from . import microformats2
+
 # allowed ActivityStreams objectTypes for attachments
 ATTACHMENT_TYPES = {'image', 'audio', 'video'}
 
-# format string for inserting a JSON Feed entry's image into its content_html
-HTML_IMAGE_TEMPLATE = '<p><img src="{}"/></p>'
-
 
 def activities_to_jsonfeed(activities, actor=None, title=None, feed_url=None,
                            home_page_url=None):
@@ -55,18 +54,8 @@ def actor_name(obj):
     if obj.get('objectType') == 'person':
       continue
     author = obj.get('author', {})
-
-    content = obj.get('content')
-    # The JSON Feed spec (https://jsonfeed.org/version/1#items) says that the
-    # URL from the "image" property may also appear in "content_html", in which
-    # case it should be interpreted as the "main, featured image" of the
-    # post. It does not specify the behavior or semantics in the case that the
-    # image does *not* appear in "content_html", but currently at least one
-    # feed reader (Feedbin) will not display the image as part of the post
-    # content unless it is explicitly included in "content_html".
-    if content and image_url(obj):
-      content += HTML_IMAGE_TEMPLATE.format(image_url(obj))
-
+    content = microformats2.render_content(
+            obj, include_location=True, render_attachments=True)
     obj_title = obj.get('title') or obj.get('displayName')
     item = {
       'id': obj.get('id') or obj.get('url'),
@@ -152,26 +141,11 @@ def attachment(jf):
       as1['url'] = url
     return as1
 
-  def content(item):
-    if item.get('content_html'):
-      content = item.get('content_html')
-
-      # If the content_html ends with an <img> tag that could have been generated
-      # by Granary (and which is redundant with the entry's "image" property),
-      # remove that tag.
-      if item.get('image'):
-        image_html = HTML_IMAGE_TEMPLATE.format(item.get('image'))
-        if content.endswith(image_html):
-          content = content[:-len(image_html)]
-    else:
-      content = item.get('content_text')
-    return content
-
   activities = [{'object': {
     'objectType': 'article' if item.get('title') else 'note',
     'title': item.get('title'),
     'summary': item.get('summary'),
-    'content': content(item),
+    'content': item.get('content_html') or item.get('content_text'),
     'id': str(item.get('id') or ''),
     'published': item.get('date_published'),
     'updated': item.get('date_modified'),

granary/test/test_jsonfeed.py

@@ -57,7 +57,7 @@ def test_activities_to_jsonfeed_name_is_not_title(self):
 
   def test_activities_to_jsonfeed_image_attachment(self):
       self.assert_equals([{
-        'content_text': '',
+        'content_html': '\n<p>\n<img class="u-photo" src="http://pict/ure.jpg" alt="" />\n</p>',
         'attachments': [{
           'url': 'http://pict/ure.jpg',
           'mime_type': 'image/jpeg',
@@ -67,7 +67,9 @@ def test_activities_to_jsonfeed_image_attachment(self):
       }], {})['items'])
 
   def test_activities_to_jsonfeed_ignore_other_attachment_types(self):
-      self.assert_equals([{'content_text': ''}], activities_to_jsonfeed([{
+      self.assert_equals([{
+        'content_html': '\n<p>\n<a class="link" href="http://some/one">\n</a>\n</p>'
+      }], activities_to_jsonfeed([{
         'attachments': [{
           'url': 'http://quoted/tweet',
           'objectType': 'note',

granary/test/testdata/article_with_attachment_image_list.feed-from-as.json

@@ -8,7 +8,7 @@
     "avatar": "http://example.com/martin/image"
   },
   "items": [{
-    "content_text": "",
+    "content_html": "\n<p>\n</p>",
     "attachments": [{
       "mime_type": "image/gif",
       "url": "http://a/pic.gif"

granary/test/testdata/article_with_attachment_stream_list.feed-from-as.json

@@ -8,7 +8,7 @@
     "avatar": "http://example.com/martin/image"
   },
   "items": [{
-    "content_text": "",
+    "content_html": "\n<p>\n</p>",
     "attachments": [{
       "mime_type": "video/quicktime",
       "url": "http://a/vidjo.mov"

granary/test/testdata/article_with_audio_video.feed.json → granary/test/testdata/article_with_audio_video.feed-from-as.json

@@ -9,7 +9,7 @@
   },
   "items": [{
     "title": "i'm ready to speak",
-    "content_html": "something",
+    "content_html": "something\n<p><audio class=\"u-audio\" src=\"http://a/podcast.mp3\" controls=\"controls\">Your browser does not support the audio tag. <a href=\"http://a/podcast.mp3\">Click here to listen directly.</a></audio>\n</p>\n<p><video class=\"u-video\" src=\"http://a/vidjo.mov\" controls=\"controls\" poster=\"\">Your browser does not support the video tag. <a href=\"http://a/vidjo.mov\">Click here to view directly. </a></video>\n</p>",
     "attachments": [{
       "mime_type": "audio/mpeg",
       "url": "http://a/podcast.mp3"

granary/test/testdata/note.feed.json → granary/test/testdata/note.feed-from-as.json

@@ -12,7 +12,7 @@
     "url": "http://example.com/blog-post-123",
     "image": "http://example.com/blog-post-123/image",
     "summary": "too cool to summarize",
-    "content_html": "A note. link too<p><img src=\"http://example.com/blog-post-123/image\"/></p>",
+    "content_html": "A note. <a href=\"http://my/link\">link</a> too\n<p>  <span class=\"p-location h-card\">\n    <data class=\"p-uid\" value=\"31cb9e7ed29dbe52\"></data>\n    <a class=\"p-name u-url\" href=\"https://maps.google.com/maps?q=32.4004416,-98.9852672\">Carcassonne, Aude</a>\n    \n    \n  </span>\n</p>",
     "date_published": "2012-02-22T20:26:41",
     "date_modified": "2013-10-25T10:31:30+00:00",
     "author": {