def check(self):
from urllib import urlopen
from time import mktime
import logging, feedparser
kwargs = {'page': 1}
while True:
logger = logging.getLogger('bambu.megaphone')
try:
feed = feedparser.parse(urlopen(self.url % kwargs))
except IOError:
logger.error('IO error when looking for feed items')
break
added = False
for entry in feed.get('entries', []):
url = helpers.fix_url(entry.link)
if self.feed_ptr.items.filter(url = url).count() == 0:
if self.feed_ptr.include_item(entry.title):
logger.debug('Adding item %s' % url)
content = entry.get('content', [])
if len(content) > 0:
content = content[0].get('value') or entry.description
else:
content = entry.get('description') or None
try:
Item.objects.create_item(
feed = self.feed_ptr,
primary_text = entry.title,
secondary_text = content,
url = url,
date = datetime.fromtimestamp(
mktime(entry.updated_parsed)
).replace(tzinfo = utc),
data = {},
links = [url]
)
added = True
except Exception, ex:
logger.error('Error getting item: %s' % unicode(ex))
else:
logger.info('Ignoring item %s' % entry.link)
if added:
kwargs['page'] += 1
logger.debug('Moving to page %(page)d' % kwargs)
else:
break
def get_tweet_items(self, access_token, latest_item, **kwargs):
from bambu.megaphone.helpers import fix_url
from bambu.megaphone.models import ServiceFeed
from urllib import urlopen
from os import path
page = 1
logger = logging.getLogger('bambu.megaphone')
kwargs = {}
if latest_item:
matches = TWEET_URL_REGEX.match(latest_item.url)
if not matches or len(matches.groups()) == 0:
raise Exception('Latest tweet URL doesn\'t match expected pattern')
kwargs['since_id'] = matches.groups()[0]
feed = ServiceFeed.objects.get(
service__access_token = access_token
)
while True:
try:
data = self.get_url(self.get_messages_url, access_token,
count = 200, page = page, include_entities = 1, **kwargs
)
except HTTPException:
logger.error('Got bad HTTP response when looking for tweets')
break
except IOError:
logger.error('IO error when looking for tweets')
break
try:
items = simplejson.load(data)
except Exception, ex:
return
if len(items) == 0 or not isinstance(items, (list, tuple)):
break
for item in items:
user = item.pop('user')
place = item.pop('coordinates')
text = item.pop('text')
attachments = []
links = []
if place:
longitude, latitude = place.get('coordinates')
else:
latitude, longitude = None, None
entities = item.pop('entities', {})
existing_urls = False
if any(entities):
for url in entities.get('urls', []):
fixed = fix_url(url.get('expanded_url'))
if feed.items.filter(url__startswith = url):
existing_urls = True
break
text = text.replace(
url.get('url'), fix_url(fixed)
)
links.append(fixed)
if existing_urls:
logger.debug('Ignoring item as a repost')
continue
for url in entities.get('media', []):
url = url.get('media_url')
attachments.append(
{
'file': (urlopen, url),
'title': url[:100],
'url': url,
'extension': path.splitext(url)[-1]
}
)
yield {
'date': (
parse(item.get('created_at'))
).replace(tzinfo = utc),
'primary_text': text,
'data': item,
'url': 'http://twitter.com/%s/statuses/%s' % (
user.get('id'), item.get('id')
),
'latitude': latitude,
'longitude': longitude,
'attachments': attachments,
'links': links
}
if len(items) >= 100:
page += 1
logger.debug('Moving to page %d' % page)
else:
break
