def load(self, store_data=True):
if not self.article_css_selector or not self.article_link_selector:
# print "No CSS selector information supplied, cannot load"
return None
article_links = SiteLinkLoader.get_elements_by_css(self.url, self.article_link_selector)
new_tags = DataTag.objects.filter(name='new')
new_tag = DataTag.objects.get(name='new')
# get all the new data objects
new_datas = []
unprocessed_datas = RawData.objects.filter(source=self.source_node, tags__in=new_tags)
if unprocessed_datas:
new_datas.extend(unprocessed_datas)
if article_links:
for article_link in article_links:
article_title = strip_tags(article_link.html())
article_url = article_link.href
if article_url.startswith('/'):
article_url = article_url[1:] # get rid of the starting slash
if not article_url.lower().startswith('http://') and not article_url.lower().startswith(self.hostname):
article_url = 'http://%s/%s' % (self.hostname, article_url)
article_content = ""
article_pieces = SiteLinkLoader.get_elements_by_css(article_url, self.article_css_selector)
for article_piece in article_pieces:
article_content += strip_tags(article_piece.html())
# create a new raw data if none exists
similar_raw_datas = RawData.objects.filter(title=article_title)
if not similar_raw_datas:
# print " + Saving article: %s" % article_title
new_data = RawData()
new_data.title = article_title
new_data.data = article_content
new_data.link = article_url
new_data.source = self.source_node
new_data.data_id = article_url
if store_data:
# save it
new_data.save()
new_data.tags.add(new_tag)
new_data.save()
new_datas.append(new_data)
if new_datas:
ba = BasicAgent()
ba.search(raw_data_set=new_datas)
def fetch_data(self): """ Override fetch data of parent Uses the feedparser library to fetch the contents of an RSS feed from a given url. The entries in the result are then used to create a new RawData object.""" data = feedparser.parse(self.url) # print data # add data to 'RawData table' for entry in data.entries: rd = RawData() rd.title = entry.title rd.data = entry rd.source = self.source_node rd.save() return data
def load_feed(self, store_data=True):
working_url = self._get_working_url('feed')
#print 'loading feed from: %s' % working_url
# try:
f = urllib2.urlopen(working_url)
#print f.info()
data = f.read()
# print data
# try to decode the json string
json_obj = json.loads(data)
# print '\n\n--------------------\n%s' % json_obj['data'][0]
if store_data:
# store the data in the raw_data model
new_tag = DataTag.objects.get(name='new')
new_datas = []
unprocessed_datas = RawData.objects.filter(source=self.source_node, tags__in=[new_tag])
if unprocessed_datas:
new_datas.extend(unprocessed_datas)
for data in json_obj['data']:
fblinktype = data.get('type',None)
new_data = RawData()
title = data.get('name',data['id'])
created_at = data.get('created_time',str(datetime.datetime.now()))
if fblinktype:
if fblinktype == "link" and data.get('description',None):
title = data.get('description')
elif data.get('message',None):
title = data.get('message')
if len(title) > 100:
title = title[:100] + "..."
new_data.title = title
new_data.data_id = data['id']
new_data.source = self.data_src
new_data.data = json.dumps(data)
new_data.link = data.get('link',None)
# try and parse the date
try:
dt = parser.parse(created_at)
except ValueError:
dt = datetime.datetime.now()
new_data.occurred_at = dt
# make sure that the raw data does not exist
if not new_data.exists():
new_data.save()
new_data.tags.add(new_tag)
new_data.save()
new_datas.append(new_data)
if new_datas:
fba = FacebookAgent()
fba.search(raw_data_set = new_datas)
def load(self, store_data = True, date_limit=None, run_agent=False):
for data_src in self.data_sources:
print "Loading data from: %s" % data_src
# init variables from the data source
url = data_src.src_id
source_node = data_src
parameters = data_src.get_parameters()
username = parameters.get('username','*****@*****.**')
psw = parameters.get('password','choirpassword')
article_css_selector = parameters.get('article-css-selector','')
fetch_limit = parameters.get('fetch-limit',None)
auth = ClientAuthMethod(username,psw)
reader = GoogleReader(auth)
if reader.buildSubscriptionList():
feeds = reader.getSubscriptionList()
new_tag = DataTag.objects.get(name='new')
new_datas = []
fetch_count = 0
# loop through and store feeds we already have RawData for
for feed in feeds:
if not fetch_limit:
fetch_limit = feed.unread
read_items = []
print "Reading " + feed.title + " (%s unread)" % feed.unread
print "===================================================="
print
print "Loading items"
print
feed.loadItems()
print "Loaded %s items" % (len(feed.items),)
print
index = 0
for item in feed.items:
# make sure it doesn't already exist
title = item.title
url = item.url
index+=1
if index + 1 >= len(feed.items) and fetch_count < fetch_limit:
print "Loading more items...."
print
feed.loadMoreItems()
f = urllib.urlopen(url)
html = f.read()
doc = leaf.parse(html)
elements = doc(article_css_selector)
for element in elements:
# print
article_html = element.html()
new_data = RawData()
new_data.title = title
new_data.source = source_node
new_data.data = strip_tags(article_html)
new_data.data_id = item.id
new_data.link = item.url
try:
new_data.occurred_at = datetime.datetime.fromtimestamp(feed.lastUpdated)
except ValueError:
# print "Error, could not parse timestamp: %s" % feed.lastUpdated
new_data.occurred_at = datetime.datetime.now()
# patching in date limit thing Parris wanted --------------------------
# if date_limit is None:
# date_limit = datetime.date.today() - datetime.timedelta(week=1)
#
# if new_data.occured_at < date_limit:
# # we should skip this item .... it is too old
# continue
#
# end patch -----------------------------------------------------------
# Abandonning this idea for now ... I think it's best to patch the map view and not mess with this for now
# if it is not new... save it
if not new_data.exists():
print " + Saving article: %s" % new_data.title
new_data.save()
new_data.tags.add(new_tag)
new_datas.append(new_data)
fetch_count +=1
read_items.append(item)
# print "All done.\n %s items fetched, our limit is %s. There are %s feeds. We stopped at index %s" % (fetch_count, self.fetch_limit, len(feed.items),index)
if new_datas and run_agent:
gra = GoogleReaderAgent()
gra.search(raw_data_set = new_datas)
return new_datas
return None
