def test_no_duplicate_links(self):
"""Make sure that a link found by both original-post-discovery and
posse-post-discovery will not result in two webmentions being sent.
"""
source = self.sources[0]
source.domain_urls = ['http://target1']
activity = self.activities[0]
activity['object'].update({
'content': 'with a backlink http://target1/post/url',
'url': 'https://fa.ke/post/url',
})
original = 'http://target1/post/url'
syndicated = 'https://fa.ke/post/url'
self.expect_requests_get('http://target1', """
<html class="h-feed">
<div class="h-entry">
<a class="u-url" href="%s"></a>
</div>
</html>""" % original)
self.expect_requests_get(original, """
<div class="h-entry">
<a class="u-url" href="%s"></a>
<a class="u-syndication" href="%s"></a>
</div>""" % (original, syndicated))
self.mox.ReplayAll()
logging.debug('Original post discovery %s -> %s', source, activity)
wmtargets = tasks.get_webmention_targets(source, activity)
self.assertEquals([original], activity['object']['upstreamDuplicates'])
self.assertEquals(set([original]), wmtargets)
def test_no_duplicate_links(self):
"""Make sure that a link found by both original-post-discovery and
posse-post-discovery will not result in two webmentions being sent.
"""
self.activity['object']['content'] = 'with a link http://author/post/url'
original = 'http://author/post/url'
self.expect_requests_get('http://author', """
<html class="h-feed">
<div class="h-entry">
<a class="u-url" href="%s"></a>
</div>
</html>""" % original)
self.expect_requests_get(original, """
<div class="h-entry">
<a class="u-url" href="%s"></a>
<a class="u-syndication" href="%s"></a>
</div>""" % (original, 'https://fa.ke/post/url'))
self.mox.ReplayAll()
logging.debug('Original post discovery %s -> %s', self.source, self.activity)
wmtargets = tasks.get_webmention_targets(self.source, self.activity)
self.assertEquals([original], self.activity['object']['upstreamDuplicates'])
self.assertEquals([original], wmtargets)
def on_data(self, raw_data):
try:
# logging.debug('Received streaming message: %s...', raw_data[:100])
data = json.loads(raw_data)
if data.get('event') == 'favorite':
response = self.source.as_source.streaming_event_to_object(data)
if not response:
logging.warning('Discarding malformed favorite event: %s', raw_data)
return True
tweet = data.get('target_object')
activity = self.source.as_source.tweet_to_activity(tweet)
elif (data.get('retweeted_status', {}).get('user', {}).get('screen_name') ==
self.source.key.string_id()):
response = self.source.as_source.retweet_to_object(data)
activity = self.source.as_source.tweet_to_activity(data['retweeted_status'])
# not handling replies right now. i wish i could, but we only get the
# individual tweet that it's replying too, not the original root tweet that
# started the chain, which is the one that will have the original post
# links and should be used as the activity.
#
# worse, we store this as a response, even though it has the wrong
# activity, and so when the poll task later finds it and has the correct
# root tweet, it sees that the response has already been saved to the
# datastore and propagated, so it drops the good one on the floor.
#
# sigh. oh well.
#
# elif ('in_reply_to_status_id_str' in data and
# data.get('in_reply_to_screen_name') == self.source.key.string_id()):
# response = self.source.as_source.tweet_to_object(data)
# activity = self.source.as_source.get_activities(
# activity_id=data['in_reply_to_status_id_str'])[0]
else:
# logging.debug("Discarding message we don't handle: %s", data)
return True
targets = tasks.get_webmention_targets(self.source, activity)
models.Response(id=response['id'],
source=self.source.key,
activities_json=[json.dumps(activity)],
response_json=json.dumps(response),
type = models.Response.get_type(response),
unsent=list(targets),
).get_or_save()
except:
logging.warning('Error processing message: %s', raw_data, exc_info=True)
return True
