def crawl_friends(self, targeturl, callback):
count = 0
# load the friends page
friendsurl = join_url(targeturl, page_references.get('friends_page'))
self.load(friendsurl)
while True:
all_friends = self.driver.find_elements_by_css_selector(css_selectors.get('friends_selector'))
# break if no more friends
if len(all_friends) <= count:
break
for friend in all_friends[count:]:
if self.stop_request:
return count
count += 1
friend_info = friend.find_element_by_xpath(xpath_selectors.get('friend_info'))
name = friend_info.text
url = friend_info.get_attribute('href')
imgurl = friend.find_element_by_css_selector(css_selectors.get('friend_image')).get_attribute('src')
callback(name, url, imgurl, count)
self.scroll_to_bottom(wait=True)
return count
def login(self, user, password):
try:
self.load('https://www.facebook.com/login.php')
self.js("document.querySelector('{}').value = '{}';".format(css_selectors.get('email_field'), user))
self.js("document.querySelector('{}').value = '{}';".format(css_selectors.get('password_field'), password))
self.js("document.querySelector('{}').submit();".format(css_selectors.get('login_form')))
self.delay()
return 'login' not in self.driver.current_url
except WebDriverException:
log.error('Couldn\'t load page. Are you connected to the internet?')
return False
def _grab_post_content(self, p, with_translation=True):
translation = ''
# if there's original content grab it
# sometimes facebook automatically shows the translated text and not
# the original.
try:
so = p.find_element_by_link_text(text_content['see_original'])
# split the post text into the original post and translation
parts = so.find_elements_by_xpath(xpath_selectors.get('trans_splitter'))
translation = parts[1].text
if self.force_click(p, so):
log.info('Found and clicked "See original"')
# remove the translation from the post text
post_text = p.text.replace(text_content['hide_original'], text_content['see_original'])
post_text = post_text.replace(translation, '')
if with_translation:
return post_text, translation
else:
return post_text, ''
except (NoSuchElementException):
pass
# grab the text normally
post_text = p.text
# expand the see more links if it exists
try:
# clicking the see more link can be quite unpredictable
# sometimes it takes a while for it to load andetimes clicking it seems
# to do nothing. so wrap it in a while loop.
sm = p.find_element_by_css_selector(css_selectors.get('see_more'))
if self.force_click(p, sm):
log.info('Found and expanded "See more"')
post_text = p.text
except (NoSuchElementException):
pass
# grab the translation if it exists and is needed
if with_translation:
try:
st = p.find_element_by_link_text(text_content.get('see_translation_text'))
st_parent = st.find_element_by_xpath('../..')
if self.force_click(p, st):
log.info('Found post translation')
translation = st_parent.find_element_by_css_selector(css_selectors.get('translation')).text
except (NoSuchElementException):
pass
return post_text, translation
def crawl_about(self, targeturl, callback):
self.load(join_url(targeturl, page_references.get('about_page')))
about_links = self.driver.find_elements_by_css_selector(css_selectors.get('about_links'))
count = 0
for l in about_links:
if self.stop_request:
return count
l.click()
self.delay()
title = l.get_attribute('title')
main_pane = self.driver.find_element_by_css_selector(css_selectors.get('about_main'))
callback(title, main_pane.text)
count += 1
return count
def crawl_one_album(self, albumurl, callback):
"""Callback format: photo_source_url, photo_post_permalink, count
"""
self.load(albumurl)
count = 0
while True:
all_photos = self.driver.find_elements_by_css_selector(css_selectors.get('album_photo'))
# break if no more photos
if len(all_photos) <= count:
break
for p in all_photos[count:]:
if self.stop_request:
return count
# url of the image
try:
img = p.find_element_by_css_selector('img')
except NoSuchElementException:
# this is probably a video in the video folder
img = p.find_element_by_css_selector('span > div')
photo_source_url = self.get_bg_img_url(img)
photo_post_permalink = p.get_attribute('ajaxify')
count += 1
callback(photo_source_url, photo_post_permalink, count)
self.scroll_to_bottom(wait=True)
return count
def is_valid_target(self, targeturl):
"""Returns True if <targeturl> is a valid profile.
Does this by checking if a certain error message text is contained inside the page body.
"""
try:
self.load(targeturl)
header_text = self.driver.find_element_by_css_selector(css_selectors.get('error_header')).text
return text_content.get('error_header_text').lower() not in header_text.lower()
except NoSuchElementException:
return True
def crawl_event_guests(self, url, callback, guest_filter=None):
if guest_filter is None:
guest_filter = ['interested', 'going', 'invited']
self.load(url, scroll=False)
total = 0
# open guests list
guest_list = self.driver.find_element_by_css_selector(css_selectors.get('event_guests'))
guest_list.click()
self.delay(1.5)
buttons = self.driver.find_elements_by_css_selector(css_selectors.get('guest_buttons'))
dialog = buttons[0].find_element_by_xpath('../../..')
for b in buttons:
# check to see if we want to scrape these guests
label = b.text.strip().split(' ')[0].lower()
if label not in guest_filter:
continue
# we want to scrape these guests
count = 0
b.click()
self.delay()
scroller = dialog.find_element_by_css_selector(css_selectors.get('guest_scroller'))
while True:
results = dialog.find_elements_by_css_selector(css_selectors.get('guest_list'))
if len(results) <= count:
break
for friend in results[count:]:
if self.stop_request:
return total
friend_info = friend.find_element_by_xpath(xpath_selectors.get('event_friend_info'))
name = friend_info.text
url = friend_info.get_attribute('href')
imgurl = friend.find_element_by_css_selector(css_selectors.get('friend_image')).get_attribute('src')
count += 1
total += 1
callback(label, name, url, imgurl, total)
self.js('a = arguments[0]; a.scrollTo(0, a.scrollHeight);', scroller)
self.delay()
return total
def crawl_albums(self, targeturl, callback):
# scrape all albums
self.load(join_url(targeturl, page_references.get('albums')))
albums = self.driver.find_elements_by_css_selector(css_selectors.get('indiv_albums'))
count = 0
for album_name, album_url in [(a.text, a.get_attribute('href')) for a in albums]:
if self.stop_request:
return count
count += 1
callback(album_name, album_url, count)
return count
def crawl_search_results(self, url, callback, limit=0):
"""Accepts a callback method which has search result's name, url, imageurl,
as well as the current search result count.
Limit is the maximum number of results to return. A limit of zero is unlimited.
"""
self.load(url)
count = 0
while True:
results = self.driver.find_elements_by_css_selector(css_selectors.get('search_results'))
if len(results) <= count:
break
for r in results[count:]:
if self.stop_request or limit > 0 and count >= limit:
return count
imageurl = r.find_element_by_css_selector(css_selectors.get('search_pics')).get_attribute('src')
url = r.find_elements_by_css_selector(css_selectors.get('search_link'))[1]
name = url.text
count += 1
callback(name, url.get_attribute('href'), imageurl, count)
self.scroll_to_bottom(wait=True)
return count
def crawl_checkins(self, targeturl, callback):
"""Callback format: check_in_name, check_in_url, count
"""
self.load(join_url(targeturl, page_references.get('checkins')))
count = 0
while True:
# get groups, break if no more groups
checkins = self.driver.find_elements_by_css_selector(css_selectors.get('checkins'))
if len(checkins) <= count:
break
# extract check in info
for p in checkins:
if self.stop_request:
return count
count += 1
callback(p.text, p.get_attribute('href'), count)
self.scroll_to_bottom(wait=True)
return count
def crawl_groups(self, targeturl, callback):
"""Callback format: group_name, group_url, count
"""
self.load(join_url(targeturl, page_references.get('groups_page')))
count = 0
while True:
# get groups, break if no more groups
groups = self.driver.find_elements_by_css_selector(css_selectors.get('groups'))
if len(groups) <= count:
break
# extract group info
for g in groups:
if self.stop_request:
return count
count += 1
callback(g.text, g.get_attribute('href'), count)
self.scroll_to_bottom(wait=True)
return count
def crawl_photos(self, targeturl, callback):
"""Callback format: photo_source_url, photo_description, photo_post_permalink, count
The description is not actually the description that was posted with the photo but an
automatically generated description by Facebook's algorithms. It can detect objects that
might be contained within the image for example: "trees, person, smiling" could be a description.
"""
albumurl = join_url(targeturl, page_references.get('photos_page'))
self.load(albumurl)
count = 0
while True:
all_photos = self.driver.find_elements_by_css_selector(css_selectors.get('photo_selector'))
# break if no more photos
if len(all_photos) <= count:
break
for p in all_photos[count:]:
if self.stop_request:
return count
# url of the image
photo_source_url = p.get_attribute('data-starred-src')
# get the metadata and store that too
link = p.find_element_by_css_selector('a')
photo_description = link.get_attribute('aria-label')
# get the permalink url of the image post
photo_post_permalink = link.get_attribute('href')
count += 1
callback(photo_source_url, photo_description, photo_post_permalink, count)
self.scroll_to_bottom(wait=True)
return count
