def like_loop(self, what: str, base_link: str, amount: int,
users_validated: False) -> Interactions:
try_again = 0
sc_rolled = 0
scroll_nap = 1.5
already_interacted_links = []
interactions = Interactions()
likes = 0
try:
while likes in range(0, amount):
if self.aborting or (self.until_time
and datetime.now() > self.until_time):
break
if self.jumps.check_likes():
self.logger.warning(
"Like quotient reached its peak, leaving Like By {} activity"
.format(what))
self.quotient_breach = True
# reset jump counter after a breach report
self.jumps.likes = 0
break
if sc_rolled > 100:
try_again += 1
if try_again > 2:
self.logger.info("'{}' possibly has less images than "
"desired ({}), found: {}".format(
what, amount,
len(already_interacted_links)))
break
delay_random = random.randint(400, 600)
self.logger.info(
"Scrolled too much. Sleeping {} minutes and {} seconds".
format(int(delay_random / 60), delay_random % 60))
sleep(delay_random)
sc_rolled = 0
main_elem = self.browser.find_element_by_tag_name("main")
posts = nf_get_all_posts_on_element(main_elem)
# Interact with links
for post in posts:
link = post.get_attribute("href")
if link not in already_interacted_links:
sleep(1)
nf_scroll_into_view(self, post)
sleep(1)
nf_click_center_of_element(self, post, link)
msg, post_interactions = interact_with_post(
self, link, users_validated)
interactions += post_interactions
if post_interactions.liked_img > 0:
likes += 1
self.logger.info("[{}] - Like [{}/{}]".format(
what, likes, amount))
if msg == "block on likes":
raise SoftBlockedException(msg)
else:
sleep(1)
nf_find_and_press_back(self, base_link)
already_interacted_links.append(link)
break
else:
# For loop ended means all posts in screen has been interacted with
# will scroll the screen a bit and reload
for i in range(3):
self.browser.execute_script(JS.SCROLL_SCREEN)
self.quota_supervisor.add_server_call()
sc_rolled += 1
sleep(scroll_nap)
except SoftBlockedException:
sleep_while_blocked(self)
except Exception as err:
self.logger.error("Unexpected Exception: {}".format(err))
finally:
return interactions
def follow_user_follow(
self,
follow: str,
usernames: List[str],
amount: int = 10,
randomize: bool = False
):
if self.aborting:
return self
valid = {"followers", "followings"}
if follow not in valid:
raise ValueError(
"nf_follow_user_follow: follow must be one of %r." % valid)
self.logger.info("Starting to follow user {}".format(follow))
for index, username in enumerate(usernames):
state = {
'liked_img': 0,
'already_liked': 0,
'inap_img': 0,
'commented': 0,
'followed': 0,
'not_valid_users': 0,
}
self.logger.info("User [{}/{}]".format(index + 1, len(usernames)))
self.logger.info("--> {}".format(username.encode("utf-8")))
nf_go_to_user_page(self, username)
sleep(1)
# TODO: get follow count
follow_count = 10
actual_amount = amount
if follow_count < amount:
actual_amount = follow_count
self.logger.info("About to go to {} page".format(follow))
nf_go_to_follow_page(self, follow, username)
sleep(2)
sc_rolled = 0
scroll_nap = 1.5
already_interacted_links = []
random_chance = 50
try:
while state['followed'] in range(0, actual_amount):
if self.quotient_breach:
self.logger.warning(
"--> Follow quotient reached its peak!"
"\t~leaving Follow-User-Follow_ activity\n"
)
break
if sc_rolled > 100:
self.logger.info("Scrolled too much! ~ sleeping 10 minutes")
sleep(600)
sc_rolled = 0
users = nf_get_all_users_on_element(self)
# Interact with links instead of just storing them
for user in users:
link = user.get_attribute("href")
if link not in already_interacted_links:
msg = ""
try:
self.logger.info("about to scroll to user")
sleep(1)
nf_scroll_into_view(self, user)
self.logger.info("about to click to user")
sleep(1)
nf_click_center_of_element(self, user)
sleep(2)
valid = False
if (
user.text not in self.dont_include
and not follow_restriction(
"read", user.text, self.follow_times, self.logger
and random.randint(0, 100) <= random_chance
)
):
valid, details = nf_validate_user_call(self, user.text)
self.logger.info("Valid User: {}, details: {}".format(valid, details))
if valid:
self.logger.info("about to follow user")
follow_state, msg = follow_user(
self.browser,
"profile",
self.username,
user.text,
None,
self.blacklist,
self.logger,
self.logfolder,
)
if follow_state is True:
state['followed'] += 1
self.logger.info("user followed")
else:
self.logger.info("--> Not following")
sleep(1)
if random.randint(0, 100) <= self.user_interact_percentage:
self.logger.info(
"--> User gonna be interacted: '{}'".format(
user.text
)
)
# disable re-validating user in like_by_users
like_by_users(
self,
[user.text],
None,
True,
)
else:
state["not_valid_users"] += 1
finally:
sleep(5)
user_link = "https://www.instagram.com/{}".format(username)
follow_link = "https://www.instagram.com/{}/{}".format(username, follow)
nf_find_and_press_back(self, follow_link)
sleep(3)
if check_if_in_correct_page(self, user_link):
nf_go_to_follow_page(self, follow, username)
already_interacted_links.append(link)
if msg == "block on follow":
pass # TODO deal with block on follow
break
else:
# For loop ended means all users in screen has been interacted with
scrolled_to_bottom = self.browser.execute_script(
"return window.scrollMaxY == window.scrollY"
)
if scrolled_to_bottom and randomize and random_chance < 100:
random_chance += 25
self.browser.execute_script(
"window.scrollTo(0, 0);"
)
update_activity(self.browser, state=None)
sc_rolled += 1
sleep(scroll_nap)
elif scrolled_to_bottom:
# already followed all possibles users
break
# will scroll the screen a bit and reload
for i in range(3):
self.browser.execute_script(
"window.scrollTo(0, document.body.scrollHeight);"
)
update_activity(self.browser, state=None)
sc_rolled += 1
sleep(scroll_nap)
except Exception:
raise
sleep(4)
self.logger.info("User [{}/{}]".format(index + 1, len(usernames)))
self.logger.info("Liked: {}".format(state['liked_img']))
self.logger.info("Already Liked: {}".format(state['already_liked']))
self.logger.info("Commented: {}".format(state['commented']))
self.logger.info("Followed: {}".format(state['followed']))
self.logger.info("Inappropriate: {}".format(state['inap_img']))
self.logger.info("Not valid users: {}\n".format(state['not_valid_users']))
self.liked_img += state['liked_img']
self.already_liked += state['already_liked']
self.commented += state['commented']
self.followed += state['followed']
self.inap_img += state['inap_img']
self.not_valid_users += state['not_valid_users']
return self
def db_store_comments(self, posts: List[Post], post_link: str):
"""Stores all comments of open post then goes back to post page"""
try:
comments_button = self.browser.find_elements_by_xpath(
'//article//div[2]/div[1]//a[contains(@href,"comments")]')
if comments_button:
nf_scroll_into_view(self, comments_button[0])
nf_click_center_of_element(self, comments_button[0])
sleep(2)
comments_link = post_link + 'comments/'
if not check_if_in_correct_page(self, comments_link):
self.logger.error(
"Failed to go to comments page, navigating there")
# TODO: retry to get there naturally
web_address_navigator(self.browser, comments_link)
more_comments = self.browser.find_elements_by_xpath(
'//span[@aria-label="Load more comments"]')
counter = 1
while more_comments and counter <= 10:
self.logger.info("Loading comments ({}/10)...".format(counter))
nf_scroll_into_view(self, more_comments[0])
self.browser.execute_script("arguments[0].click();",
more_comments[0])
more_comments = self.browser.find_elements_by_xpath(
'//span[@aria-label="Load more comments"]')
counter += 1
comments = self.browser.find_elements_by_xpath(
'/html/body/div[1]/section/main/div/ul/ul[@class="Mr508"]')
for comment in comments:
inner_container = comment.find_element_by_xpath(
'.//div[@class="C4VMK"]')
username = inner_container.find_element_by_xpath(
'.//h3/div/a').text
text, _ = deform_emojis(
inner_container.find_element_by_xpath('.//span').text)
post_date = inner_container.find_element_by_xpath(
'.//time').get_attribute('datetime')
post_date = datetime.fromisoformat(post_date[:-1])
user = db_get_or_create_user(self, username)
self.db.session.add(user)
self.db.session.commit()
for post in posts:
comment = Comment(
date_posted=post_date,
text=text,
user=user,
post=post,
)
self.db.session.add(comment)
self.db.session.commit()
else:
self.logger.error("No comments found")
except SQLAlchemyError:
self.db.session.rollback()
raise
finally:
self.db.session.commit()
nf_find_and_press_back(self, post_link)
def like_by_users(
self,
usernames: List[str],
amount: int = None,
users_validated: bool = False
):
"""Likes some amounts of images for each usernames"""
if self.aborting:
return self
amount = amount or self.user_interact_amount
usernames = usernames or []
self.quotient_breach = False
for index, username in enumerate(usernames):
if self.quotient_breach:
break
state = {
'liked_img': 0,
'already_liked': 0,
'inap_img': 0,
'commented': 0,
'followed': 0,
'not_valid_users': 0,
}
self.logger.info(
"Username [{}/{}]".format(index + 1, len(usernames))
)
self.logger.info("--> {}".format(username.encode("utf-8")))
if len(usernames) == 1 and users_validated:
nf_go_from_post_to_profile(self, username)
else:
nf_go_to_user_page(self, username)
if not users_validated:
validation, details = nf_validate_user_call(self, username)
if not validation:
self.logger.info(
"--> Not a valid user: {}".format(details)
)
state["not_valid_users"] += 1
continue
try_again = 0
sc_rolled = 0
scroll_nap = 1.5
already_interacted_links = []
try:
while state['liked_img'] in range(0, amount):
if self.jumps["consequent"]["likes"] >= self.jumps["limit"]["likes"]:
self.logger.warning(
"--> Like quotient reached its peak!\t~leaving "
"Like-By-Users activity\n"
)
self.quotient_breach = True
# reset jump counter after a breach report
self.jumps["consequent"]["likes"] = 0
break
if sc_rolled > 100:
try_again += 1
if try_again > 2: # you can try again as much as you want by changing this number
self.logger.info(
"'{}' user POSSIBLY has less valid images than "
"desired:{} found:{}...".format(
username, amount, len(already_interacted_links))
)
break
self.logger.info(
"Scrolled too much! ~ sleeping 10 minutes")
sleep(600)
sc_rolled = 0
main_elem = self.browser.find_element_by_tag_name("main")
# feed = main_elem.find_elements_by_xpath('//div[@class=" _2z6nI"]')
posts = nf_get_all_posts_on_element(main_elem)
# Interact with links instead of just storing them
for post in posts:
link = post.get_attribute("href")
if link not in already_interacted_links:
self.logger.info("about to scroll to post")
sleep(1)
nf_scroll_into_view(self, post)
self.logger.info("about to click to post")
sleep(1)
nf_click_center_of_element(self, post)
success, msg, state = nf_interact_with_post(
self,
link,
amount,
state,
users_validated,
)
self.logger.info(
"Returned from liking, should still be in post page")
sleep(5)
nf_find_and_press_back(
self,
"https://www.instagram.com/{}/".format(username)
)
already_interacted_links.append(link)
if success:
break
if msg == "block on likes":
# TODO deal with block on likes
break
else:
# For loop ended means all posts in screen has been interacted with
# will scroll the screen a bit and reload
for i in range(3):
self.browser.execute_script(
"window.scrollTo(0, document.body.scrollHeight);"
)
update_activity(self.browser, state=None)
sc_rolled += 1
sleep(scroll_nap)
except Exception:
raise
sleep(4)
self.logger.info("Username [{}/{}]".format(index + 1, len(usernames)))
self.logger.info("--> {} ended".format(username.encode("utf-8")))
self.logger.info("Liked: {}".format(state['liked_img']))
self.logger.info("Already Liked: {}".format(state['already_liked']))
self.logger.info("Commented: {}".format(state['commented']))
self.logger.info("Followed: {}".format(state['followed']))
self.logger.info("Inappropriate: {}".format(state['inap_img']))
self.logger.info("Not valid users: {}\n".format(state['not_valid_users']))
self.liked_img += state['liked_img']
self.already_liked += state['already_liked']
self.commented += state['commented']
self.followed += state['followed']
self.inap_img += state['inap_img']
self.not_valid_users += state['not_valid_users']
return self
def like_by_tags(
self,
tags: List[str] = None,
amount: int = 50,
skip_top_posts: bool = True,
use_smart_hashtags: bool = False,
use_smart_location_hashtags: bool = False,
):
"""Likes (default) 50 images per given tag"""
if self.aborting:
return self
# if smart hashtag is enabled
if use_smart_hashtags is True and self.smart_hashtags != []:
self.logger.info("Using smart hashtags")
tags = self.smart_hashtags
elif use_smart_location_hashtags is True and self.smart_location_hashtags != []:
self.logger.info("Using smart location hashtags")
tags = self.smart_location_hashtags
# deletes white spaces in tags
tags = [tag.strip() for tag in tags]
tags = tags or []
self.quotient_breach = False
for index, tag in enumerate(tags):
if self.quotient_breach:
break
state = {
'liked_img': 0,
'already_liked': 0,
'inap_img': 0,
'commented': 0,
'followed': 0,
'not_valid_users': 0,
}
self.logger.info("Tag [{}/{}]".format(index + 1, len(tags)))
self.logger.info("--> {}".format(tag.encode("utf-8")))
tag = tag[1:] if tag[:1] == "#" else tag
nf_go_to_tag_page(self, tag)
# get amount of post with this hashtag
try:
possible_posts = self.browser.execute_script(
"return window._sharedData.entry_data."
"TagPage[0].graphql.hashtag.edge_hashtag_to_media.count"
)
except WebDriverException:
try:
possible_posts = self.browser.find_element_by_xpath(
read_xpath("get_links_for_tag", "possible_post")
).text
if possible_posts:
possible_posts = format_number(possible_posts)
else:
self.logger.info(
"Failed to get the amount of possible posts in '{}' tag "
"~empty string".format(tag)
)
possible_posts = None
except NoSuchElementException:
self.logger.info(
"Failed to get the amount of possible posts in {} tag".format(tag)
)
possible_posts = None
self.logger.info(
"desired amount: {} | top posts [{}] | possible posts: "
"{}".format(
amount,
"enabled" if not skip_top_posts else "disabled",
possible_posts,
)
)
if possible_posts is not None:
amount = possible_posts if amount > possible_posts else amount
# sometimes pages do not have the correct amount of posts as it is
# written there, it may be cos of some posts is deleted but still keeps
# counted for the tag
sleep(1)
try_again = 0
sc_rolled = 0
scroll_nap = 1.5
already_interacted_links = []
try:
while state['liked_img'] in range(0, amount):
if sc_rolled > 100:
try_again += 1
if try_again > 2:
self.logger.info(
"'{}' tag POSSIBLY has less images than "
"desired:{} found:{}...".format(
tag,
amount,
len(already_interacted_links)
)
)
break
self.logger.info("Scrolled too much! ~ sleeping 10 minutes")
sleep(600)
sc_rolled = 0
main_elem = self.browser.find_element_by_tag_name("main")
posts = nf_get_all_posts_on_element(main_elem)
# Interact with links instead of just storing them
for post in posts:
link = post.get_attribute("href")
if link not in already_interacted_links:
self.logger.info("about to scroll to post")
sleep(1)
nf_scroll_into_view(self, post)
self.logger.info("about to click to post")
sleep(1)
nf_click_center_of_element(self, post)
success, msg, state = nf_interact_with_post(
self,
link,
amount,
state,
)
self.logger.info("Returned from liking, should still be in post page")
sleep(2)
nf_find_and_press_back(self, "https://www.instagram.com/explore/tags/{}/".format(tag))
already_interacted_links.append(link)
if success:
break
if msg == "block on likes":
# TODO deal with block on likes
break
else:
# For loop ended means all posts in screen has been interacted with
# will scroll the screen a bit and reload
for i in range(3):
self.browser.execute_script(
"window.scrollTo(0, document.body.scrollHeight);"
)
update_activity(self.browser, state=None)
sc_rolled += 1
sleep(scroll_nap)
except Exception:
raise
sleep(2)
self.logger.info("Tag [{}/{}]".format(index + 1, len(tags)))
self.logger.info("--> {} ended".format(tag.encode("utf-8")))
self.logger.info("Liked: {}".format(state['liked_img']))
self.logger.info("Already Liked: {}".format(state['already_liked']))
self.logger.info("Commented: {}".format(state['commented']))
self.logger.info("Followed: {}".format(state['followed']))
self.logger.info("Inappropriate: {}".format(state['inap_img']))
self.logger.info("Not valid users: {}\n".format(state['not_valid_users']))
self.liked_img += state['liked_img']
self.already_liked += state['already_liked']
self.commented += state['commented']
self.followed += state['followed']
self.inap_img += state['inap_img']
self.not_valid_users += state['not_valid_users']
return self
def follow_user_follow(self,
relation: str,
usernames: List[str],
amount: int = 10,
randomize: bool = False,
random_chance: int = 50):
"""
Follows 'amount' users of 'relation' ("following" or "followers") list of each user in usernames
:param relation: what list to use, "following" or "followers"
:param usernames: list of usernames to follow relations of
:param amount: amount of users to follow for each user in 'usernames'
:param randomize: if the bot will include a random factor to choose who to follow or
follow the first 'amount' of usernames on the list
:param random_chance: chance a user will be followed if using 'randomize'
"""
if self.aborting:
return self
valid = {"followers", "following"}
if relation not in valid:
self.logger.info(
'{} is not a valid relation, using "followers"'.format(
relation))
relation = "followers"
self.logger.info("Starting to follow users {}".format(relation))
# for each username
for index, username in enumerate(usernames):
# if aborting or quota was breached or its past time according to settings break the loop
if self.aborting or self.quotient_breach or (
self.until_time and datetime.now() > self.until_time):
break
interactions = Interactions()
self.logger.info("Follow User {} [{}/{}]: {} - started".format(
relation, index + 1, len(usernames), username))
user_link = "https://www.instagram.com/{}".format(username)
follow_link = "https://www.instagram.com/{}/{}".format(
username, relation)
# navigate to user page
if not check_if_in_correct_page(self, user_link):
nf_go_to_user_page(self, username)
sleep(1)
# get followers & following counts and change amount if less than desired
followers_count, following_count = get_relationship_counts(
self, username)
follow_count = following_count if relation == "following" else followers_count
follow_count = follow_count if follow_count else 0
actual_amount = amount
if follow_count < amount:
actual_amount = follow_count
# go to relation page
nf_go_to_follow_page(self, relation, username)
sleep(2)
# follow users
sc_rolled = 0
scroll_nap = 1.5
already_interacted_links = []
while interactions.followed in range(actual_amount):
# if aborting or quota was breached or its past time according to settings break the loop
if self.aborting or (self.until_time
and datetime.now() > self.until_time):
break
# if quotient was breached break the loop
if self.jumps.check_follows():
self.logger.warning(
"Follow quotient reached its peak, leaving Follow User {} activity"
.format(relation))
# reset jump counter before breaking the loop
self.jumps.follows = 0
self.quotient_breach = True
break
# if scrolled too much sleep for 5-10 minutes
if sc_rolled > 100:
delay_random = random.randint(300, 600)
self.logger.info(
"Scrolled too much, sleeping {} minutes and {} seconds"
.format(int(delay_random / 60), delay_random % 60))
sleep(delay_random)
sc_rolled = 0
# get loaded usernames
users = nf_get_all_users_on_element(self)
# if no users were grabbed try to go back and load the relation page again
while len(users) == 0:
nf_find_and_press_back(self, user_link)
in_user_page = check_if_in_correct_page(self, user_link)
if not in_user_page:
nf_go_to_user_page(self, username)
nf_go_to_follow_page(self, relation, username)
# get loaded usernames
users = nf_get_all_users_on_element(self)
# If after rechecking we are in the correct page there still no are users
# the bot is most surely soft blocked from seeing relations, that block doesnt last long usually.
# sleep for 5-10 minutes
if len(users) == 0:
delay_random = random.randint(300, 600)
self.logger.info(
"Soft block on see followers, "
"sleeping {} minutes and {} seconds".format(
int(delay_random / 60), delay_random % 60))
sleep(delay_random)
self.logger.debug("Grabbed {} usernames".format(len(users)))
# first one in the list is un-clickable by bad design on browser instagram, its behind the top bar
for user in users[1:]:
link = user.get_attribute("href")
# try to follow first not already interacted user
if link not in already_interacted_links:
msg = ""
try:
user_text = user.text
user_link2 = "https://www.instagram.com/{}".format(
user_text)
self.logger.info("Followed [{}/{}]".format(
interactions.followed, actual_amount))
# Go to user page
self.logger.info(
"Trying user {}".format(user_text))
nf_scroll_into_view(self, user)
sleep(1)
nf_click_center_of_element(self, user, user_link2)
sleep(2)
# validate user
valid = False
if (user_text not in self.settings.dont_include and
not is_follow_restricted(self, user_text)
and
random.randint(0, 100) <= random_chance):
valid, details = nf_validate_user_call(
self, user_text,
self.quota_supervisor.FOLLOW)
self.logger.info(
"Valid User: {}, details: {}".format(
valid, details))
# follow user
if valid:
follow_state, msg = follow_user(
self, "profile", user_text)
if follow_state is True:
interactions.followed += 1
elif msg == "already followed":
interactions.already_followed += 1
elif msg == "jumped":
# will break the loop after certain consecutive jumps
self.jumps.follows += 1
# interact with user
if (self.settings.do_like
and random.randint(0, 100) <=
self.settings.user_interact_percentage
):
self.logger.info(
"Interacting with user '{}'".format(
user_text))
if not check_if_in_correct_page(
self, user_link2):
nf_go_from_post_to_profile(
self, user_text)
interactions += like_loop(
self, "Interact with user '{}'".format(
user_text), user_link2,
self.settings.user_interact_amount,
True)
else:
interactions.not_valid_users += 1
except Exception as e:
self.logger.error(e)
finally:
# go back to relation page and start the loop again
sleep(1)
nf_find_and_press_back(self, follow_link)
in_follow_page = check_if_in_correct_page(
self, follow_link)
if not in_follow_page:
in_user_page = check_if_in_correct_page(
self, user_link)
if not in_user_page:
nf_go_to_user_page(self, username)
nf_go_to_follow_page(self, relation, username)
already_interacted_links.append(link)
if msg == "block on follow":
# raise SoftBlockedException(msg)
pass # TODO: deal with block on follow
break
else:
# For loop ended means all users in screen has been interacted with
scrolled_to_bottom = self.browser.execute_script(
JS.SCROLLED_TO_BOTTOM)
# even if we are at the bottom if we were using randomize some users were ignored
# so the bot can go back and look again with a higher random chance to foollow the users
if scrolled_to_bottom and randomize and random_chance < 100:
random_chance += 25
self.browser.execute_script(JS.SCROLL_TO_TOP)
self.quota_supervisor.add_server_call()
sc_rolled += 1
sleep(scroll_nap)
elif scrolled_to_bottom:
# already followed all possibles users
break
# if not at the bottom of the list
# will scroll the screen a bit and look again
for i in range(3):
self.browser.execute_script(JS.SCROLL_SCREEN)
self.quota_supervisor.add_server_call()
sc_rolled += 1
sleep(scroll_nap)
sleep(3)
self.logger.info("Follow User {} [{}/{}] - ended".format(
relation, index + 1, len(usernames)))
self.logger.info(str(interactions))
self.interactions += interactions
return self
def get_follow(
self, username: str, follow: str
) -> Set[str]: # set of followers or following of given username
valid = {"followers", "following"}
if follow not in valid:
raise ValueError("get_follow: follow must be one of %r." % valid)
user_link = "https://www.instagram.com/{}/".format(username)
if not check_if_in_correct_page(self, user_link):
if self.username == username:
go_to_bot_user_page(self)
else:
nf_go_to_user_page(self, username)
if follow == 'followers':
query = self.instauser.followers.all()
else:
query = self.instauser.following.all()
usernames = set([instauser.username for instauser in query])
followers_count, following_count = get_relationship_counts(self, username)
count = followers_count if follow == 'followers' else following_count
if count == len(usernames):
return usernames
sleep(2)
nf_go_to_follow_page(self, follow, username)
sleep(2)
sc_rolled = 0
scroll_nap = 1.5
seen_usernames = []
try:
while True:
if self.aborting or (self.until_time
and datetime.now() > self.until_time):
break
if sc_rolled > 100:
delay_random = random.randint(400, 600)
self.logger.info(
"Scrolled too much, sleeping {} minutes and {} seconds".
format(int(delay_random / 60), delay_random % 60))
sleep(delay_random)
sc_rolled = 0
users = nf_get_all_users_on_element(self)
while len(users) == 0:
nf_find_and_press_back(self, user_link)
in_user_page = check_if_in_correct_page(self, user_link)
if not in_user_page:
nf_go_to_user_page(self, username)
nf_go_to_follow_page(self, follow, username)
users = nf_get_all_users_on_element(self)
if len(users) == 0:
delay_random = random.randint(200, 300)
self.logger.info(
"Soft block on see {}, "
"sleeping {} minutes and {} seconds".format(
follow, int(delay_random / 60), delay_random % 60))
sleep(300)
for user in users[1:]:
link = user.get_attribute("href")
user_text = user.text
if user_text not in seen_usernames:
seen_usernames.append(user_text)
if user_text not in usernames:
add_follow_times(self, user_text)
else:
# For loop ended means all users in screen have been saved
scrolled_to_bottom = self.browser.execute_script(
JS.SCROLLED_TO_BOTTOM)
if scrolled_to_bottom:
# already saved all possibles users
break
# will scroll the screen a bit and grab usernames again
for i in range(3):
self.browser.execute_script(JS.SCROLL_SCREEN)
self.quota_supervisor.add_server_call()
sc_rolled += 1
sleep(scroll_nap)
if scrolled_to_bottom:
# already saved all possibles users
break
except Exception:
raise
usernames = usernames.union(set(seen_usernames))
self.logger.info("Grabbed {} {} names".format(len(usernames), follow))
return usernames
def nf_validate_user_call(self,
username: str,
post_link: str = None) -> Tuple[bool, str]:
"""Checks if user can be liked according to declared settings
Also stores post data in database if appropriate
:returns: valid, reason
"""
followers_count = None
following_count = None
number_of_posts = None
if username == self.username:
reason = "---> Username '{}' is yours!\t~skipping user\n".format(
self.username)
return False, reason
if username in self.ignore_users:
reason = ("---> '{}' is in the `ignore_users` list\t~skipping "
"user\n".format(username))
return False, reason
blacklist_file = "{}blacklist.csv".format(self.logfolder)
blacklist_file_exists = os.path.isfile(blacklist_file)
if blacklist_file_exists:
with open("{}blacklist.csv".format(self.logfolder), "rt") as f:
reader = csv.reader(f, delimiter=",")
for row in reader:
for field in row:
if field == username:
return (
False,
"---> {} is in blacklist ~skipping "
"user\n".format(username),
)
potency_ratio = self.potency_ratio
delimit_by_numbers = self.delimit_by_numbers
max_followers = self.max_followers
max_following = self.max_following
min_followers = self.min_followers
min_following = self.min_following
min_posts = self.min_posts
max_posts = self.max_posts
skip_private = self.skip_private
skip_private_percentage = self.skip_private_percentage
skip_no_profile_pic = self.skip_no_profile_pic
skip_no_profile_pic_percentage = self.skip_no_profile_pic_percentage
skip_business = self.skip_business
skip_non_business = self.skip_non_business
skip_business_percentage = self.skip_business_percentage
skip_business_categories = self.skip_business_categories
dont_skip_business_categories = self.dont_skip_business_categories
skip_bio_keyword = self.skip_bio_keyword
if not any([
potency_ratio, delimit_by_numbers, max_followers, max_following,
min_followers, min_following, min_posts, max_posts, skip_private,
skip_private_percentage, skip_no_profile_pic,
skip_no_profile_pic_percentage, skip_business, skip_non_business,
skip_business_percentage, skip_business_categories,
skip_bio_keyword
]):
# Nothing to check, skip going to user page and then back for nothing
return True, "Valid user"
try:
if post_link:
nf_go_from_post_to_profile(self, username)
else:
post_link = get_current_url(self.browser)
self.logger.info("about to start checking user page")
# Checks the potential of target user by relationship status in order
# to delimit actions within the desired boundary
if (potency_ratio
or delimit_by_numbers and (max_followers or max_following
or min_followers or min_following)):
relationship_ratio = None
reverse_relationship = False
# get followers & following counts
self.logger.info("About to get relationship counts")
followers_count, following_count = get_relationship_counts(
self.browser, username, self.logger)
if potency_ratio and potency_ratio < 0:
potency_ratio *= -1
reverse_relationship = True
# division by zero is bad
followers_count = 1 if followers_count == 0 else followers_count
following_count = 1 if following_count == 0 else following_count
if followers_count and following_count:
relationship_ratio = (float(followers_count) /
float(following_count)
if not reverse_relationship else
float(following_count) /
float(followers_count))
self.logger.info(
"User: '******' |> followers: {} |> following: {} |> relationship "
"ratio: {}".format(
username,
followers_count if followers_count else "unknown",
following_count if following_count else "unknown",
truncate_float(relationship_ratio, 2)
if relationship_ratio else "unknown",
))
if followers_count or following_count:
if potency_ratio and not delimit_by_numbers:
if relationship_ratio and relationship_ratio < potency_ratio:
reason = (
"'{}' is not a {} with the relationship ratio of {} "
"~skipping user\n".format(
username,
"potential user" if not reverse_relationship
else "massive follower",
truncate_float(relationship_ratio, 2),
))
return False, reason
elif self.delimit_by_numbers:
if followers_count:
if max_followers:
if followers_count > max_followers:
reason = (
"User '{}'s followers count exceeds maximum "
"limit ~skipping user\n".format(username))
return False, reason
if min_followers:
if followers_count < min_followers:
reason = (
"User '{}'s followers count is less than "
"minimum limit ~skipping user\n".format(
username))
return False, reason
if following_count:
if max_following:
if following_count > max_following:
reason = (
"User '{}'s following count exceeds maximum "
"limit ~skipping user\n".format(username))
return False, reason
if min_following:
if following_count < min_following:
reason = (
"User '{}'s following count is less than "
"minimum limit ~skipping user\n".format(
username))
return False, reason
if potency_ratio:
if relationship_ratio and relationship_ratio < potency_ratio:
reason = (
"'{}' is not a {} with the relationship ratio of "
"{} ~skipping user\n".format(
username,
"potential user"
if not reverse_relationship else "massive "
"follower",
truncate_float(relationship_ratio, 2),
))
return False, reason
if min_posts or max_posts:
# if you are interested in relationship number of posts boundaries
try:
number_of_posts = getUserData(
"graphql.user.edge_owner_to_timeline_media.count",
self.browser)
except WebDriverException:
self.logger.error("~cannot get number of posts for username")
reason = "---> Sorry, couldn't check for number of posts of " "username\n"
return False, reason
if max_posts:
if number_of_posts > max_posts:
reason = (
"Number of posts ({}) of '{}' exceeds the maximum limit "
"given {}\n".format(number_of_posts, username,
max_posts))
return False, reason
if min_posts:
if number_of_posts < min_posts:
reason = (
"Number of posts ({}) of '{}' is less than the minimum "
"limit given {}\n".format(number_of_posts, username,
min_posts))
return False, reason
# Skip users
# skip private
if skip_private:
try:
self.browser.find_element_by_xpath(
"//*[contains(text(), 'This Account is Private')]")
is_private = True
except NoSuchElementException:
is_private = False
if is_private and (random.randint(0, 100) <=
skip_private_percentage):
return False, "{} is private account, by default skip\n".format(
username)
# skip no profile pic
if skip_no_profile_pic:
try:
profile_pic = getUserData("graphql.user.profile_pic_url",
self.browser)
except WebDriverException:
self.logger.error("~cannot get the post profile pic url")
return False, "---> Sorry, couldn't get if user profile pic url\n"
if (profile_pic in default_profile_pic_instagram
or str(profile_pic).find(
"11906329_960233084022564_1448528159_a.jpg") > 0) and (
random.randint(
0, 100) <= skip_no_profile_pic_percentage):
return False, "{} has default instagram profile picture\n".format(
username)
# skip business
if skip_business or skip_non_business:
# if is business account skip under conditions
try:
is_business_account = getUserData(
"graphql.user.is_business_account", self.browser)
except WebDriverException:
self.logger.error(
"~cannot get if user has business account active")
return (
False,
"---> Sorry, couldn't get if user has business "
"account active\n",
)
if skip_non_business and not is_business_account:
return (
False,
"---> Skipping non business because skip_non_business set to True",
)
if is_business_account:
try:
category = getUserData(
"graphql.user.business_category_name", self.browser)
except WebDriverException:
self.logger.error("~cannot get category name for user")
return False, "---> Sorry, couldn't get category name for " "user\n"
if len(skip_business_categories) == 0:
# skip if not in dont_include
if category not in dont_skip_business_categories:
if len(dont_skip_business_categories) == 0 and (
random.randint(
0, 100) <= skip_business_percentage):
return False, "'{}' has a business account\n".format(
username)
else:
return (
False,
("'{}' has a business account in the "
"undesired category of '{}'\n".format(
username, category)),
)
else:
if category in skip_business_categories:
return (
False,
("'{}' has a business account in the "
"undesired category of '{}'\n".format(
username, category)),
)
if len(skip_bio_keyword) != 0:
# if contain stop words then skip
try:
profile_bio = getUserData("graphql.user.biography",
self.browser)
except WebDriverException:
self.logger.error("~cannot get user bio")
return False, "---> Sorry, couldn't get get user bio " "account active\n"
for bio_keyword in skip_bio_keyword:
if bio_keyword.lower() in profile_bio.lower():
return (
False,
"{} has a bio keyword of {}, by default skip\n".format(
username, bio_keyword),
)
# if everything is ok
return True, "Valid user"
except NoSuchElementException:
return False, "Unable to locate element"
finally:
if self.store_in_database:
try:
user = db_get_or_create_user(self, username)
self.db.session.add(user)
user.date_checked = datetime.now()
if followers_count:
user.followers_count = followers_count
if following_count:
user.following_count = following_count
if number_of_posts:
user.posts_count = number_of_posts
except SQLAlchemyError:
self.db.session.rollback()
finally:
self.db.session.commit()
if post_link:
nf_find_and_press_back(self, post_link)
