16

I'm looking for a possibility to get a followers and following list in JSON format via web request (in the same way as on the Instagram web site). For example, I can login via requests, and get user info:

def get_user_info(self, user_name):
    url = "https://www.instagram.com/" + user_name + "/?__a=1"
    try:
        r = requests.get(url)
    except requests.exceptions.ConnectionError:
        print 'Seems like dns lookup failed..'
        time.sleep(60)
        return None
    if r.status_code != 200:
        print 'User: ' + user_name + ' status code: ' + str(r.status_code)
        print r
        return None
    info = json.loads(r.text)
    return info['user']

I tried to see what request chrome send to server, but was unsuccessful. The question is: how to prepare a similar get or post request to retrieve followers list without the Instagram API?

Jongware
  • 22,200
  • 8
  • 54
  • 100
Ilya Kryukov
  • 461
  • 1
  • 5
  • 11

2 Answers2

14

GraphQL queries with query_hash = "58712303d941c6855d4e888c5f0cd22f" (followings) and "37479f2b8209594dde7facb0d904896a" (followers) return this information. With being logged in, do a GET query to instagram.com/graphql/query with parameters query_hash and variables, where variables is a JSON-formatted set of variables id (user id, as in the return dictionary of your get_user_info() function), first (a page length, it seems the current maximum is 50) and in subsequent requests after set to the end_cursor in the previous response dictionary.

Alternatively, the Instaloader library provides a convenient way to login and then programmatically access a profile's followers and followings list.

import instaloader

# Get instance
L = instaloader.Instaloader()

# Login or load session
L.login(USER, PASSWORD)        # (login)
L.interactive_login(USER)      # (ask password on terminal)
L.load_session_from_file(USER) # (load session created w/
                               #  `instaloader -l USERNAME`)

# Obtain profile metadata
profile = instaloader.Profile.from_username(L.context, PROFILE)

# Print list of followees
for followee in profile.get_followees():
    print(followee.username)

# (likewise with profile.get_followers())

Besides username, the attributes full_name, userid, followed_by_viewer and many more are defined in the Profile instance that is returned for each followee.

aandergr
  • 627
  • 7
  • 13
  • Seems like a lot of similar tool is using `GraphQL` with `query_hash` for followers. I am trying to approach the same thing use Java. Do you think it is applicable? Do you have any code to share? Thanks a lot. – chrisTina Mar 29 '19 at 20:09
  • @aandergr is that ok that every 1k scraped followers i get blocked to ~10 minutes cuz of: `Too many queries in the last time. Need to wait 630 seconds, until ...` Can I somehow pass through that? – Chiefir Oct 08 '19 at 08:55
-4

Easy(just replace _a with __a)

'https://www.instagram.com/'+user_name+'/followers/?_a=1'
'https://www.instagram.com/'+user_name+'/following/?_a=1'
Вадим Виноградов
  • 21
  • 1
  • 2