extract atom rendering out into its own module

activitystreams.py

@@ -29,11 +29,11 @@
 import json
 import logging
 import re
-import os
 import urllib
 from webob import exc
 
 import appengine_config
+import atom
 import facebook
 import instagram
 import microformats2
@@ -43,7 +43,6 @@
 import source
 import twitter
 
-from google.appengine.ext.webapp import template
 import webapp2
 
 # maps app id to source class
@@ -57,7 +56,6 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <response>%s</response>
 """
-ATOM_TEMPLATE_FILE = os.path.join(os.path.dirname(__file__), 'templates', 'user_feed.atom')
 ITEMS_PER_PAGE = 100
 
 # default values for each part of the API request path, e.g. /@me/@self/@all/...
@@ -120,37 +118,25 @@ def get(self):
     response = source.get_activities_response(*args, **paging_params)
     activities = response['items']
 
-    if format == 'atom':
-      # strip the access token from the request URL before returning
-      params = dict(self.request.GET.items())
-      for key in 'access_token', 'access_token_key', 'access_token_secret':
-        if key in params:
-          del params[key]
-      request_url = '%s?%s' % (self.request.path_url, urllib.urlencode(params))
-      actor = source.get_actor(user_id)
-      response.update({
-          'host_url': self.request.host_url + "/",
-          'request_url': request_url,
-          'title': 'User feed for ' + actor.get('displayName', 'unknown'),
-          'updated': activities[0]['object'].get('published') if activities else '',
-          'actor': actor,
-          })
-      # TODO: render activity contents as html. i'd need to allow markup
-      # through, though, which iirc has broken before when posts have included
-      # HTML entities.
-      # for activity in response.get('items', []):
-      #   obj = activity['object']
-      #   obj['content'] = microformats2.render_content(obj)
-
     # encode and write response
     self.response.headers['Access-Control-Allow-Origin'] = '*'
     if format == 'json':
       self.response.headers['Content-Type'] = 'application/json'
       self.response.out.write(json.dumps(response, indent=2))
     elif format == 'atom':
+      # TODO: render activity contents as html. i'd need to allow markup
+      # through, though, which iirc has broken before when posts have included
+      # HTML entities.
+      # for activity in response.get('items', []):
+      #   obj = activity['object']
+      #   obj['content'] = microformats2.render_content(obj)
+      actor = source.get_actor(user_id)
+      request_url = '%s?%s' % (self.request.path_url,
+                               urllib.urlencode(dict(self.request.GET.items())))
       self.response.headers['Content-Type'] = 'text/xml'
-      self.preprocess_for_atom(response)
-      self.response.out.write(template.render(ATOM_TEMPLATE_FILE, response))
+      self.response.out.write(atom.activities_to_atom(
+          activities, actor, host_url=self.request.host_url + '/',
+          request_url=request_url))
     elif format == 'xml':
       self.response.headers['Content-Type'] = 'text/xml'
       self.response.out.write(XML_TEMPLATE % util.to_xml(response))
@@ -201,22 +187,6 @@ def get_positive_int(self, param):
       raise exc.HTTPBadRequest('Invalid %s: %s (should be positive int)' %
                                (param, val))
 
-  def preprocess_for_atom(self, response):
-    """Prepares data to be rendered as Atom.
-
-    Right now, just populates each activity's title field, since Atom <entry>
-    requires the title element.
-
-    Args:
-      response: ActivityStreams response dict
-    """
-    for activity in response.get('items', []):
-      if not activity.get('title'):
-        obj = activity.get('object', {})
-        activity['title'] = util.ellipsize(
-          activity.get('displayName') or activity.get('content') or
-          obj.get('title') or obj.get('displayName') or obj.get('content'))
-
 
 application = webapp2.WSGIApplication([('.*', Handler)],
                                       debug=appengine_config.DEBUG)

activitystreams_test.py

@@ -127,9 +127,10 @@ def test_atom_format(self):
       # include access_token param to check that it gets stripped
       resp = self.get_response('?format=atom&access_token=foo&a=b')
       self.assertEquals(200, resp.status_int)
-      request_url = 'http://localhost?a=b&format=atom'
-      self.assert_multiline_equals(test_module.ATOM % {'request_url': request_url},
-                                   resp.body)
+      self.assert_multiline_equals(
+        test_module.ATOM % {'request_url': 'http://localhost',
+                            'host_url': 'http://localhost/'},
+        resp.body)
 
   def test_unknown_format(self):
     resp = activitystreams.application.get_response('?format=bad')

atom.py

@@ -0,0 +1,56 @@
+"""Convert ActivityStreams to Atom.
+
+Atom spec: http://atomenabled.org/developers/syndication/
+"""
+
+import logging
+import os
+import string
+import urlparse
+
+from google.appengine.ext.webapp import template
+from oauth_dropins.webutil import util
+
+ATOM_TEMPLATE_FILE = os.path.join(os.path.dirname(__file__), 'templates', 'user_feed.atom')
+
+
+def activities_to_atom(activities, actor, request_url=None, host_url=None):
+  """Converts ActivityStreams activites to an Atom feed.
+
+  Args:
+    activities: list of ActivityStreams activity dicts
+    actor: ActivityStreams actor dict, the author of the feed
+    request_url: the URL of this Atom feed, if any. Used in a link rel="self".
+    host_url: the home URL for this Atom feed, if any. Used in the top-level
+      feed <id> element.
+
+  Returns: unicode string with Atom XML
+  """
+  # Strip query params from URLs so that we don't include access tokens, etc
+  host_url = (_remove_query_params(host_url) if host_url
+              else 'https://github.com/snarfed/activitystreams-unofficial')
+  request_url = _remove_query_params(request_url) if request_url else host_url
+
+  # Make sure every activity has the title field, since Atom <entry> requires
+  # the title element.
+  for a in activities:
+    if not a.get('title'):
+      obj = a.get('object', {})
+      a['title'] = util.ellipsize(
+        a.get('displayName') or a.get('content') or
+        obj.get('title') or obj.get('displayName') or obj.get('content'))
+
+  return template.render(ATOM_TEMPLATE_FILE, {
+    'items': activities,
+    'host_url': host_url,
+    'request_url': request_url,
+    'title': 'User feed for ' + actor.get('displayName', 'unknown'),
+    'updated': activities[0]['object'].get('published') if activities else '',
+    'actor': actor,
+    })
+
+
+def _remove_query_params(url):
+  parsed = list(urlparse.urlparse(url))
+  parsed[4] = ''
+  return urlparse.urlunparse(parsed)

atom_test.py

@@ -0,0 +1,26 @@
+"""Unit tests for atom.py."""
+
+import copy
+
+import atom
+import facebook_test
+import instagram_test
+from oauth_dropins.webutil import testutil
+import twitter_test
+
+
+class AtomTest(testutil.HandlerTest):
+
+  def test_activities_to_atom(self):
+    for test_module in facebook_test, instagram_test, twitter_test:
+      self.assert_multiline_equals(
+        test_module.ATOM % {'request_url': 'http://request/url',
+                            'host_url': 'http://host/url',
+                            },
+        atom.activities_to_atom(
+          [copy.deepcopy(test_module.ACTIVITY)],
+          test_module.ACTOR,
+          request_url='http://request/url?access_token=foo',
+          host_url='http://host/url',
+          ))
+

facebook_test.py

@@ -469,7 +469,7 @@ def tag_uri(name):
       >
 <generator uri="https://github.com/snarfed/activitystreams-unofficial" version="0.1">
   activitystreams-unofficial</generator>
-<id>http://localhost/</id>
+<id>%(host_url)s</id>
 <title>User feed for Ryan Barrett</title>
 
 <subtitle>something about me</subtitle>

instagram_test.py

@@ -244,7 +244,7 @@ def tag_uri(name):
       >
 <generator uri="https://github.com/snarfed/activitystreams-unofficial" version="0.1">
   activitystreams-unofficial</generator>
-<id>http://localhost/</id>
+<id>%(host_url)s</id>
 <title>User feed for Ryan B</title>
 
 <subtitle>foo</subtitle>

twitter_test.py

@@ -450,7 +450,7 @@ def tag_uri(name):
       >
 <generator uri="https://github.com/snarfed/activitystreams-unofficial" version="0.1">
   activitystreams-unofficial</generator>
-<id>http://localhost/</id>
+<id>%(host_url)s</id>
 <title>User feed for Ryan Barrett</title>
 
 <subtitle>my description</subtitle>