How to crawl Instagram data using its public API and Python?
In 2020, the official Instagram API allow you to access only your own posts and not even public comments and posts on Instagram because of the rising privacy concerns from the users and frequent accusations of data-breach at many big companies including Facebook. This has made it difficult for programmers to crawl Instagram data.
So, how to crawl Instagram data?
There’s still a workaround. It does provide an API which is publicly accessible.
Let’s try to hit this URL.
Eureka, it’s a JSON response:
URL & JSON response
Here, travel is the hashtag, as we can also see in the JSON response. And JSON response consist of all the posts containing hashtag travel. Now JSON response is easy to understand. Edges is the list that contains posts’ data. So, now all we need is to parse this JSON to get the data.
Programmatically parsing response using Python
Libraries required: requests
Here’s a quick Python code to get the captions from the posts, you can modify it for your own use:
import requests class Parser: HASH_KEY = "graphql" HASHTAG_KEY = "hashtag" MEDIA_KEY = "edge_hashtag_to_media" LIST_KEY = "edges" NODE_KEY = "node" CAPTION_LIST_KEY = "edge_media_to_caption" TEXT_KEY = "text" def __init__(self, tag): self.tag = tag def get_url(self): url = "https://www.instagram.com/explore/tags/" + self.tag + "/?__a=1" return url def get_request_response(self): r = requests.get(url=self.get_url(), params="") data = r.json() return data def get_captions(self): captions =  data = self.get_request_response() nodes_list = data[Parser.HASH_KEY][Parser.HASHTAG_KEY][Parser.MEDIA_KEY][Parser.LIST_KEY] for obj in nodes_list: caption_list = obj[Parser.NODE_KEY][Parser.CAPTION_LIST_KEY][Parser.LIST_KEY] if len(caption_list) > 0: caption = caption_list[Parser.NODE_KEY][Parser.TEXT_KEY] captions.append(caption) print(caption) def main(): parser = Parser("travel") parser.get_captions() if __name__ == "__main__": main()
Later, we would be posting more programming tutorials to get you started with Python. We would advance from beginner level to intermediate and then to Machine Learning. If you like our posts, please like, comment and share. Also, don’t forget to subscribe to our awesome newsletter.