def peatix_selenium_run():
'''
Spider for peatix based on selenium
'''
# starting url for scrawling
base_url = "http://peatix.com/search/?country=SG&p="
search_range = 10
event_list = []
# load existing url for peatix
existing_urls = load_existing_urls('peatix')
# start web driver
browser = webdriver.Firefox()
for i in range(search_range):
url = base_url + str(i+1)
logger.info('Start to parse url %s', url)
browser.get(url)
# wait until certain element to ensure web page rendering finished
WebDriverWait(browser, 10).until(
EC.presence_of_element_located((By.CSS_SELECTOR, "li.pod-summary-block.ng-scope"))
)
browser.implicitly_wait(2)
# extract all events element
events = browser.find_elements_by_xpath('//li[@class="event-summary-block ng-scope"]')
logger.info('Found %d events in url %s', len(events), url)
if len(events) > 0:
# open another web driver for event detail page parsing
sub_browser = webdriver.Firefox()
for event in events:
try:
title_element = event.find_element_by_xpath('.//div[@class="detail-block"]//h3[@class="event-name ng-binding"]')
title = title_element.text
# parse event detail page url
url_element = event.find_element_by_xpath('.//a[@class="event-summary"]')
url = url_element.get_attribute('href')
# filtering out already parsed urls
if url not in existing_urls:
logger.info('Start to parse event %s with url %s', title, url)
event_info = parse_one_event(url, sub_browser)
# save parsed event to event list
if event_info:
event_list.append(event_info)
else:
logger.info('Event %s with url %s has already been parsed.', title, url)
except Exception, e:
logger.error(e.message, exc_info=True)
sub_browser.quit()
def facebook_selenium_run():
'''
Spider for facebook. The logic is to get all event ids of https://www.facebook.com/events/upcoming
through selenium. Then, retrieve the event detail through facebook api.
'''
event_info_list = []
# extract facebook user going event list
event_info_list.extend(_locate_event_list())
time.sleep(5)
# extract facebook user nearby suggest event list
event_info_list.extend(_locate_event_list(suggest=True))
logger.info('Aggrerate %d for detail parsing through facebook api', len(event_info_list))
valid_event_list = []
# load existing urls for facebook
existing_urls = load_existing_urls('facebook')
# parse each event detail through facebook graph api
for event in event_info_list:
# filtering out already exist urls
if event['source_url'] not in existing_urls:
try:
json_postdata = _render_to_json(_create_post_url(event['id'], APP_ID, APP_SECRET))
# title is necessary, raise exception if no title
event['title'] = json_postdata['name']
# start time is necessary, raise exception if no start time
event['start_time'] = _process_start_time(json_postdata['start_time'])
event['detail'] = json_postdata.get('description', None)
event['location'] = _process_location(json_postdata.get('place', None))
event['category'] = json_postdata.get('category', None)
event['w0'] = json_postdata.get('attending_count', None)
event['w1'] = json_postdata.get('interested_count', None)
event['w2'] = json_postdata.get('maybe_count', None)
valid_event_list.append(event)
except:
logger.error('Error for event %d', int(event['id']), exc_info=True)
else:
logger.info('Event with url %s has already been parsed.', event['source_url'])
logger.info('Successfully get %d events detail through facebook graph api', len(valid_event_list))
# save all valid events to database
if len(valid_event_list) > 0:
save_events_to_db(valid_event_list)
def meetup_selenium_run():
"""Spider for meetup. The logic is as follows:
1. Find all groups on start page
2. For each group, extract its upcoming events
3. For each upcoming event, go into detail page
"""
# start url
base_url = "http://www.meetup.com/"
group_list = []
try:
# start web driver
browser = webdriver.Firefox()
# load existing urls for meetup
existing_urls = load_existing_urls("meetup")
# load base url
browser.get(base_url)
# find all the groups in the base url
groups = browser.find_elements_by_xpath('//li[@class="groupCard tileGrid-tile"]')
logger.info("Find %d groups on meetup home page.", len(groups))
# start parse url for each group
for group in groups:
group_url = group.find_element_by_xpath('.//a[@class="display-none"]')
if group_url:
# add valid group url into list
group_list.append(group_url.get_attribute("href"))
browser.quit()
logger.info("Find %d groups with url on meetup home page.", len(group_list))
# start to parse each individual group
if len(group_list) > 0:
for group_url in group_list:
logger.info("Start to parse group with url %s", group_url)
meetup_parse_one_group(group_url, existing_urls)
# wait for 2 seconds
time.sleep(2)
except Exception, e:
logging.error(e.message, exc_info=True)
