def create_caption_dict():
"""
Create a dictionary with influencer ids, list of posts, and list of captions.
Args: None
Output:
captions (dict): Dcitionary with influencer id, list of posts, and list of captions.s
"""
client, collection = setup_mongo_client('instascrape', 'test')
captions = {}
cursor = collection.find({})
for x in cursor:
user_id = x['node']['owner']['id']
if user_id in captions:
add_posts_captions(captions, user_id, x)
else:
captions[user_id] = {'posts': [], 'caption': []}
add_posts_captions(captions, user_id, x)
# Remove profiles that have been deleted since initial scraping
del captions['4018066784']
client.close()
return captions
def followscrape(num_requests):
"""
Scrape instagram followers
Args:
num_requests (int): Number of influencers to be scraped
Output: None
"""
client, collection = setup_mongo_client('instascrape', 'followers')
driver = webdriver.Firefox()
selenium_instagram_login(driver, 'instagram_credentials.json')
init_url_search = "https://www.instagram.com/graphql/query/?query_id=17851374694183129&variables={{%22id%22:%22{}%22,%22first%22:20}}"
base_url_search = "https://www.instagram.com/graphql/query/?query_id=17851374694183129&variables={{%22id%22:%22{}%22,%22first%22:500,%22after%22:%22{}%22}}"
for i in range(num_requests):
influencer_id = find_next_influencer('data/ordered_influencers.txt',
'data/scraped_influencers.txt')
# Initial search for followers
driver.get(init_url_search.format(influencer_id))
data_as_json = load_json_from_html(driver)
insert_data(data_as_json, influencer_id, collection)
page_info = get_page_info(data_as_json)
page_counter = 1
print "Finished scraping {} pages for influencer {}".format(
page_counter, influencer_id)
# Keep searching while followers still exist
while page_info['has_next_page']:
driver.get(
base_url_search.format(influencer_id,
str(page_info['end_cursor'])))
data_as_json = load_json_from_html(driver)
insert_data(data_as_json, influencer_id, collection)
page_info = get_page_info(data_as_json)
page_counter += 1
print "Finished scraping {} pages for influencer {}".format(
page_counter, influencer_id)
time.sleep(np.random.uniform(7, 10))
write_text(influencer_id, 'data/scraped_influencers.txt')
time.sleep(np.random.uniform(7, 10))
print "Finished scraping {} influencers of {}".format(
i + 1, num_requests)
client.close()
print "\n Finished scraping {} influencers' followers".format(num_requests)
return None
def instascrape(page_info_filepath, num_requests):
"""
Scrape instagram hashtag search
Args:
page_info_filepath (str): Filepath to text file with page_info dicts
num_requests (int): Number of pages to be scraped
Action: saves influencer node information to pymongo database
Output: None
"""
client, collection = setup_mongo_client('instascrape', 'test')
page_info = load_last_line(page_info_filepath)
base_url_search = "https://www.instagram.com/graphql/query/?query_id=17875800862117404&variables={{%22tag_name%22:%22womenwhoclimb%22,%22first%22:{},%22after%22:%22{}%22}}"
for i in range(num_requests):
if page_info['has_next_page']:
response = requests.get(
base_url_search.format('12', str(page_info['end_cursor'])))
if response.status_code == 200:
insert_edge(response, collection)
page_info = get_page_info(response)
add_new_line(page_info, page_info_filepath)
else:
print "Status Code = " + str(response.status_code)
return None
time.sleep(np.random.uniform(15, 45))
print "Finished scraping {} pages of {}".format(i + 1, num_requests)
client.close()
print "\n Finished scraping {} pages of 12 influencers each".format(
num_requests)
return None
def create_influencer_dict(filepath_json, return_dict=False):
"""
Create a dictionary with influencer id and posts.
Args:
filepath_json (str): Filepath where influencer dictionary will be saved as json.
Action: Saves influencer dictionary to filepath.
influencers (dict)
Keys: id
Values (dict):
Keys: 'posts' (list), 'followers' (list)
Output: None
"""
client, collection = setup_mongo_client('instascrape', 'test')
# Retrieve shortcodes and ids from influencers in MongoDB
shortcodes_ids = []
cursor = collection.find({})
for x in cursor:
shortcodes_ids.append(
(str(x['node']['shortcode']), str(x['node']['owner']['id'])))
# Create influencers dictionary
influencers = {}
for sc_id_tuple in shortcodes_ids:
if sc_id_tuple[1] in influencers:
influencers[sc_id_tuple[1]]['posts'].append(sc_id_tuple[0])
else:
influencers[sc_id_tuple[1]] = {'posts': [sc_id_tuple[0]]}
# Remove profiles that have been deleted since initial scraping
del influencers['4018066784']
write_json(influencers, filepath_json)
client.close()
if return_dict:
return influencers
def create_hashtag_likes_dict(filepath_json, return_dict=False):
"""
Create a dictionary with influencer id, hashtags, posts, and
number of likes for each post.
Args:
filepath_json (str): Filepath where dictionary will be saved as json.
Action: Saves influencer dictionary to filepath.
influencers (dict)
Keys: id
Values (dict):
Keys: 'hashtags' (list), posts (list), 'likes' (list)
Output: None
"""
client, collection = setup_mongo_client('instascrape', 'test')
influencers = {}
cursor = collection.find({})
for x in cursor:
user_id = x['node']['owner']['id']
if user_id in influencers:
add_hashtag_posts_likes(influencers, user_id, x)
else:
influencers[user_id] = {'hashtags': [], 'posts': [], 'likes': []}
add_hashtag_posts_likes(influencers, user_id, x)
# Remove profiles that have been deleted since initial scraping
del influencers['4018066784']
client.close()
write_json(influencers, filepath_json)
if return_dict:
return influencers
