def crawl_tweets_for_event(event_id):
r = Redis()
p = HTMLParser()
total_tweets = 0
event_title = r.get("festival:%s:title" % event_id).decode("utf-8", errors="ignore")
event_title = strip_accents(event_title)
event_title = p.unescape(event_title)
event_title = remove_stopwords(event_title)
artists = r.get("festival:%s:artists" % event_id)
for k, v in eval(artists).items():
if type(v) == list:
for artist in v:
print F, "searching tweets for %s %s" % (k, artist)
total_tweets += search_term(artist)
elif type(v) == str:
print F, "searching tweets for %s %s" % (k, v)
total_tweets += search_term(v)
r.incr("festival:%s:crawled_times" % event_id)
print F, "searching tweets for festival title: %s" % event_title
total_tweets += search_term(event_title, event_id) # newsid
print F, "total tweets: %d" % total_tweets
def enrich_news(redis):
keys = redis.keys('news:*:id')
if len(keys) == 0:
return
news_tweets = []
pages = []
for key in keys:
locale = redis.get('news:%s:locale' % key.split(':')[1])
if locale == 'en_us':
lang = 'english'
else:
lang = 'spanish'
id = key.split(':')[1]
terms = get_search_terms_news(redis, id, lang)
for term in terms:
tweets, pages = twitter.search_term(term)
for tweet in tweets:
tweet.event_id = id
news_tweets.extend(tweets)
pages.extend(pages)
print tag, "got", len(news_tweets), 'tweets for news'
return (news_tweets, pages)
def enrich_festivals(redis):
import datetime
keys = redis.keys('festival:*:startDate')
to_search_keys = []
# solo buscar en los festivales que estan pasando ahora
for key in keys:
startDate = redis.get(key)
id = key.split(':')[1]
startDate = datetime.datetime.strptime(startDate, '%a, %d %b %Y %H:%M:%S')
if datetime.datetime.today() >= startDate:
to_search_keys.append(key)
if len(keys) == 0:
return
festivals_tweets = []
pages = []
for key in to_search_keys:
id = key.split(':')[1]
terms = get_search_terms_festivals(redis, id)
for term in terms:
tweets, pages = twitter.search_term(term)
for tweet in tweets:
tweet.event_id = id
festivals_tweets.extend(tweets)
pages.extend(pages)
print tag, "got", len(festivals_tweets), 'tweets for festivals'
return (festivals_tweets, pages)
def crawl_week_later():
gn = GoogleNews()
r = Redis()
all_news = r.keys('page:*:title')
i = 1
total = 0
print F, "total news: %d" % len(all_news)
for key_news_title in all_news:
newsid = key_news_title.split(':')[1]
if r.get('page:%s:crawled_week' % newsid) is None and r.get('page:%s:crawled_day' % newsid) is not None:
i += 1
if r.get('page:%s:locale' % newsid) == 'es_cl':
lang = 'spanish'
else:
lang = 'english'
news_title_stopwords = p.unescape(strip_accents(r.get(key_news_title).decode('utf-8', errors='ignore')))
news_title = remove_stopwords(news_title_stopwords, lang=lang)
print F, "searching tweets for news (w/ sw): \"%s\"" % news_title_stopwords
print F, "searching tweets for news (w/o sw): \"%s\"" % news_title
print F, "searching tweets for news: %s" % news_title
total += search_term(news_title, newsid) # search by title
print F, "total news searched: %d" % i
print F, "total tweets crawled: %d" % total
def enrich_event(redis_key):
tag = '[events_enricher]'
redis = Redis()
queries = list(set(redis.lrange(redis_key, 0, -1)))
tweets = []
for query in queries:
#print tag, 'searching "%s"' % query.decode('utf-8', errors='ignore')
tweets.extend(search_term(query.decode('utf-8', errors='ignore')))
print tag, 'got', len(tweets), 'tweets'
event_id = redis_key.split(':')[1]
save_tweets(tweets, event_id)
def crawl_current_day():
gn = GoogleNews()
r = Redis()
gn.get_topnews()
all_news = r.keys('page:*:title')
p = HTMLParser()
i = 1
total = 0
print F, "total pages: %d" % len(all_news)
for key_news_title in all_news:
newsid = key_news_title.split(':')[1]
# only interested in news here
if r.get('page:%s:type' % newsid) != 'news':
continue
# and pages not already crawled in its first day
if r.get('page:%s:crawled_day' % newsid) is None:
i += 1
# lang for stopwords remove
if r.get('page:%s:locale' % newsid) == 'es_cl':
lang = 'spanish'
else:
lang = 'english'
news_title_stopwords = p.unescape(strip_accents(r.get(key_news_title).decode('utf-8', errors='ignore')))
news_title = remove_stopwords(news_title_stopwords, lang=lang)
print F, "searching tweets for news (w/ sw): \"%s\"" % news_title_stopwords
print F, "searching tweets for news (w/o sw): \"%s\"" % news_title
# mark its news' first day as searched
r.incr('page:%s:crawled_day' % newsid)
# search by title in twitter
total += search_term(news_title, newsid)
print F, "total news searched: %d" % i
print F, "total tweets crawled: %d" % total
