erjemin/instaloader
erjemin/instaloader
1
0
Fork 0
Code Issues Wiki Activity

Fix get_followees() and implement get_followers()

Browse Source
This commit is contained in:
André Koch-Kramer 2017-07-20 17:57:12 +02:00
parent dd513f7190
commit 1fdce16f46
2 changed files with 82 additions and 49 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
README.rst
View File

@ -182,8 +182,7 @@ Usage as Python module
You may also use parts of Instaloader as library to do other interesting You may also use parts of Instaloader as library to do other interesting
things. things.
For example, to get a list of all followees of a profile as well as For example, to get a list of all followees and a list of all followers of a profile, do
their follower count, do
.. code:: python .. code:: python
@ -197,8 +196,15 @@ their follower count, do
# Retrieve followees # Retrieve followees
followees = loader.get_followees(PROFILE) followees = loader.get_followees(PROFILE)
print(PROFILE + " follows these profiles:")
for f in followees: for f in followees:
print("%i\t%s\t%s" % (f['follower_count'], f['username'], f['full_name'])) print("\t%s\t%s" % (f['username'], f['full_name']))
# Retrieve followers
followers = loader.get_followers(PROFILE)
print("Followers of " + PROFILE + ":")
for f in followers:
print("\t%s\t%s" % (f['username'], f['full_name']))
Then, you may download all pictures of all followees with Then, you may download all pictures of all followees with

119
instaloader.py
View File

@ -217,60 +217,87 @@ class Instaloader:
raise ProfileNotExistsException("Profile {0} does not exist.".format(profile)) raise ProfileNotExistsException("Profile {0} does not exist.".format(profile))
return int(data['entry_data']['ProfilePage'][0]['user']['id']) return int(data['entry_data']['ProfilePage'][0]['user']['id'])
def get_followees(self, profile: str) -> List[Dict[str, Any]]: def get_followers(self, profile: str) -> List[Dict[str, Any]]:
""" """
Retrieve list of followees of given profile Retrieve list of followers of given profile.
To use this, one needs to be logged in and private profiles has to be followed,
otherwise this returns an empty list.
:param profile: Name of profile to lookup followees :param profile: Name of profile to lookup followers.
:return: List of followees (list of dictionaries), as returned by instagram server :return: List of followers (list of dictionaries).
""" """
tmpsession = copy_session(self.session) tmpsession = copy_session(self.session)
data = self.get_json(profile, session=tmpsession) header = self.default_http_header(empty_session_only=True)
profile_id = data['entry_data']['ProfilePage'][0]['user']['id'] del header['Connection']
query = ["q=ig_user(" + profile_id + ")+%7B%0A" del header['Content-Length']
"++follows.", header['authority'] = 'www.instagram.com'
str(data['entry_data']['ProfilePage'][0]['user']['follows']['count']) + header['scheme'] = 'https'
")+%7B%0A" header['accept'] = '*/*'
"++++count%2C%0A" header['referer'] = 'https://www.instagram.com/' + profile + '/'
"++++page_info+%7B%0A" tmpsession.headers = header
"++++++end_cursor%2C%0A" profile_id = self.get_id_by_username(profile)
"++++++has_next_page%0A" resp = tmpsession.get('https://www.instagram.com/graphql/query/',
"++++%7D%2C%0A" params={'query_id': 17851374694183129,
"++++nodes+%7B%0A" 'variables': '{"id":"' + str(profile_id) + '","first":500}'})
"++++++id%2C%0A"
"++++++full_name%2C%0A"
"++++++username%2C%0A"
"++++++followed_by+%7B%0A"
"++++++++count%0A"
"++++++%7D%0A"
"++++%7D%0A"
"++%7D%0A"
"%7D%0A"
"&ref=relationships%3A%3Afollow_list"]
tmpsession.headers.update(self.default_http_header())
tmpsession.headers.update({'Referer': 'https://www.instagram.com/' + profile + '/following/'})
tmpsession.headers.update({'Content-Type': 'application/x-www-form-urlencoded'})
resp = tmpsession.post('https://www.instagram.com/query/', data=query[0] + "first(" + query[1])
if resp.status_code == 200: if resp.status_code == 200:
data = json.loads(resp.text) data = resp.json()
followees = [] followers = []
while True: while True:
for followee in data['follows']['nodes']: edge_followed_by = data['data']['user']['edge_followed_by']
followee['follower_count'] = followee.pop('followed_by')['count'] followers.extend([follower['node'] for follower in edge_followed_by['edges']])
followees = followees + [followee] page_info = edge_followed_by['page_info']
if data['follows']['page_info']['has_next_page']: if page_info['has_next_page']:
resp = tmpsession.post('https://www.instagram.com/query/', resp = tmpsession.get('https://www.instagram.com/graphql/query/',
data="{0}after({1}%2C+{2}".format(query[0], params={'query_id': 17851374694183129,
data['follows']['page_info']['end_cursor'], 'variables': '{"id":"' + str(profile_id) + '","first":500},"after":"' +
query[1])) page_info['end_cursor'] + '"}'})
data = json.loads(resp.text) data = resp.json()
else: else:
break break
return followees else:
if self.test_login(tmpsession): raise ConnectionException("ConnectionError({0}): unable to gather followers.".format(resp.status_code))
raise ConnectionException("ConnectionError(" + str(resp.status_code) + "): " return followers
"unable to gather followees.")
raise LoginRequiredException("Login required to gather followees.") def get_followees(self, profile: str) -> List[Dict[str, Any]]:
"""
Retrieve list of followees (followings) of given profile.
To use this, one needs to be logged in and private profiles has to be followed,
otherwise this returns an empty list.
:param profile: Name of profile to lookup followers.
:return: List of followees (list of dictionaries).
"""
tmpsession = copy_session(self.session)
header = self.default_http_header(empty_session_only=True)
del header['Connection']
del header['Content-Length']
header['authority'] = 'www.instagram.com'
header['scheme'] = 'https'
header['accept'] = '*/*'
header['referer'] = 'https://www.instagram.com/' + profile + '/'
tmpsession.headers = header
profile_id = self.get_id_by_username(profile)
resp = tmpsession.get('https://www.instagram.com/graphql/query/',
params={'query_id': 17874545323001329,
'variables': '{"id":"' + str(profile_id) + '","first":500}'})
if resp.status_code == 200:
data = resp.json()
followees = []
while True:
edge_follow = data['data']['user']['edge_follow']
followees.extend([followee['node'] for followee in edge_follow['edges']])
page_info = edge_follow['page_info']
if page_info['has_next_page']:
resp = tmpsession.get('https://www.instagram.com/graphql/query/',
params={'query_id': 17874545323001329,
'variables': '{"id":"' + str(profile_id) + '","first":500},"after":"' +
page_info['end_cursor'] + '"}'})
data = resp.json()
else:
break
else:
raise ConnectionException("ConnectionError({0}): unable to gather followees.".format(resp.status_code))
return followees
def download_pic(self, name: str, url: str, date_epoch: float, outputlabel: Optional[str] = None, def download_pic(self, name: str, url: str, date_epoch: float, outputlabel: Optional[str] = None,
filename_suffix: Optional[str] = None) -> bool: filename_suffix: Optional[str] = None) -> bool: