Instagram-Profilecrawl

Quickly crawl the information (e.g. followers, tags etc...) of an instagram profile. No login required!

Automation Script for crawling information from ones instagram profile.
Like e.g. the number of posts, followers, and the tags of the the posts

Getting started

Just do:

git clone https://github.com/timgrossmann/instagram-profilecrawl.git

It uses selenium and requests to get all the information so install them with:

pip install -r requirements.txt

Install the proper chromedriver for your operating system. Once you (download it)[https://sites.google.com/a/chromium.org/chromedriver/downloads] just drag and drop it into instagram-profilecrawl/assets directory.

Use it!

Now you can start using it following this example:

python3.5 crawl_profile.py username1 username2 ... usernameX

Settings: To limit the amount of posts to be analyzed, change variable limit_amount in crawl_profile.py. Default value is 12000.\

Optional login

If you want to access more features (such as private accounts which you followed with yours will be accessible) you must enter your username and password in setting.py. Remember, it's optional.

Here are the steps to do so:

1. Open Settings.py
2. Search for login_username & login_username
3. Put your information inside the quotation marks

Second option: just the settings to your script

Settings.login_username = 'my_insta_account'
Settings.login_password = 'my_passwort_xxx'

Run on Raspberry Pi

To run the crawler on Raspberry Pi with Firefox, follow these steps:

1. Install Firefox: sudo apt-get install firefox-esr
2. Get the geckodriver as described here
3. Install pyvirtualdisplay: sudo pip3 install pyvirtualdisplay
4. Run the script for RPi: python3 crawl_profile_pi.py username1 username2 ...

Collecting stats:

If you are interested in collecting and logging stats from a crawled profile, use the log_stats.py script after runnig crawl_profile.py (or crawl_profile_pi.py). For example, on Raspberry Pi run:

1. Run python3 crawl_profile_pi.py username
2. Run python3 log_stats.py -u username for specific user or python3 log_stats.py for all user

This appends the collected profile info to stats.csv. Can be useful for monitoring the growth of an Instagram account over time. The logged stats are: Time, username, total number of followers, following, posts, likes, and comments. The two commands can simply be triggered using crontab (make sure to trigger log_stats.py several minutes after crawl_profile_pi.py).

Settings:

Path to the save the profile jsons:

Settings.profile_location = os.path.join(BASE_DIR, 'profiles')

Should the profile json file should get a timestamp

Settings.profile_file_with_timestamp = True

Path to the save the commentors:

Settings.profile_commentors_location = os.path.join(BASE_DIR, 'profiles')

Should the commentors file should get a timestamp

Settings.profile_commentors_file_with_timestamp = True

Scrap & save the posts json

Settings.scrap_posts_infos = True

How many (max) post should be scraped

Settings.limit_amount = 12000

Should the comments also be saved in json files

Settings.output_comments = False

Should the mentions in the post image saved in json files

Settings.mentions = True

Should the users who liked the post saved in json files Attention: be aware it would take a lot of time. script just can load 12 like at once. before making a break and load again

Settings.scrap_posts_likers = True

Time between post scrolling (increase if you got errors)

Settings.sleep_time_between_post_scroll = 1.5

Time between comment scrolling (increase if you got errors)

Settings.sleep_time_between_comment_loading = 1.5

Output debug messages to Console

Settings.log_output_toconsole = True

Path to the logfile

Settings.log_location = os.path.join(BASE_DIR, 'logs')

Output debug messages to File

Settings.log_output_tofile = True

New logfile for every run

Settings.log_file_per_run = False

The information will be saved in a JSON-File in ./profiles/{username}.json

Example of a files data

{
  "alias": "Tim Gro\u00dfmann",
  "username": "grossertim",
  "num_of_posts": 127,
  "posts": [
    {
      "caption": "It was a good day",
      "location": {
        "location_url": "https://www.instagram.com/explore/locations/345421482541133/caffe-fernet/",
        "location_name": "Caffe Fernet",
        "location_id": "345421482541133",
        "latitude": 1.2839,
        "longitude": 103.85333
      },
      "img": "https://scontent.cdninstagram.com/t51.2885-15/e15/p640x640/16585292_1355568261161749_3055111083476910080_n.jpg?ig_cache_key=MTQ0ODY3MjA3MTQyMDA3Njg4MA%3D%3D.2",
      "date": "2018-04-26T15:07:32.000Z",
      "tags": ["#fun", "#good", "#goodday", "#goodlife", "#happy", "#goodtime", "#funny", ...],
      "likes": 284,
      "comments": {
        "count": 0,
        "list": [],
       },
     },
     {
      "caption": "Wild Rocket Salad with Japanese Sesame Sauce",
      "location": {
        "location_url": "https://www.instagram.com/explore/locations/318744905241462/junior-kuppanna-restaurant-singapore/",
        "location_name": "Junior Kuppanna Restaurant, Singapore",
        "location_id": "318744905241462",
        "latitude": 1.31011,
        "longitude": 103.85672
      },
      "img": "https://scontent.cdninstagram.com/t51.2885-15/e35/16122741_405776919775271_8171424637851271168_n.jpg?ig_cache_key=MTQ0Nzk0Nzg2NDI2ODc5MTYzNw%3D%3D.2",
      "date": "2018-04-26T15:07:32.000Z",
      "tags": ["#vegan", "#veganfood", "#vegansofig", "#veganfoodporn", "#vegansofig", ...],
      "likes": 206,
      "comments": {
        "count": 1,
        "list": [
          {
            "user": "pastaglueck",
            "comment": "nice veganfood"
           },
         ],
       },
     },
     .
     .
     .
     ],
  "prof_img": "https://scontent.cdninstagram.com/t51.2885-19/s320x320/14564896_1313394225351599_6953533639699202048_a.jpg",
  "followers": 1950,
  "following": 310
}

The script also collects usernames of users who commented on the posts and saves it in ./profiles/{username}_commenters.txt file, sorted by comment frequency.

With the help of Wordcloud you could do something like that with your used tags

---

Have Fun & Feel Free to report any issues