This is super simple, quick & dirty Python script to retrieve all your (recent) tweets, and store them as combined JSON file on your disk.
It's really basic (no lookup/storage of media links, no saving threads, ...).
Get your Twitter App credentials by registering as a developer.
export APP_TOKEN="your bearer token"
Then run with
tw_get_tweets.py -i <your user id>
Note: The user id is a number, not your @ handle.
| #!/usr/bin/env python3 | |
| import argparse | |
| import requests | |
| import colorama | |
| import os, sys, time | |
| import json | |
| colorama.init() | |
| def add_authentication(inout_request): | |
| inout_request.headers["Authorization"] = f'Bearer {os.environ.get("APP_TOKEN")}' | |
| inout_request.headers["User-Agent"] = "tw_get_tweets" | |
| return inout_request; | |
| def get_tweets(id, start): | |
| params = {} | |
| params["expansions"] = ",".join({"attachments.poll_ids", "attachments.media_keys", "author_id", "edit_history_tweet_ids", "entities.mentions.username", "geo.place_id", "in_reply_to_user_id", "referenced_tweets.id", "referenced_tweets.id.author_id"}) | |
| params["tweet.fields"] = ",".join({"attachments", "author_id", "context_annotations", "conversation_id", "created_at", "edit_controls", "entities", "geo", "id", "in_reply_to_user_id", "lang", "public_metrics", "possibly_sensitive", "referenced_tweets", "reply_settings", "source", "text", "withheld"}); | |
| params["user.fields"] = ",".join({"created_at", "description", "entities", "id", "location", "name", "pinned_tweet_id", "profile_image_url", "protected", "public_metrics", "url", "username", "verified", "withheld"}) | |
| params["media.fields"] = ",".join({"duration_ms", "height", "media_key", "preview_image_url", "type", "url", "width", "public_metrics", "alt_text", "variants"}) | |
| params["place.fields"] = ",".join({"contained_within", "country", "country_code", "full_name", "geo", "id", "name", "place_type"}) | |
| params["poll.fields"] = ",".join({"duration_minutes", "end_datetime", "id", "options", "voting_status"}) | |
| params["max_results"] = 100 | |
| if start != "": | |
| params["pagination_token"] = start | |
| tweets_url = f"https://api.twitter.com/2/users/{id}/tweets" | |
| response = requests.request("GET", tweets_url, auth=add_authentication, params=params) | |
| return response | |
| def main(args): | |
| assert(args.id) | |
| count = 0 | |
| start_token = '' | |
| while True: | |
| response = get_tweets(args.id, start_token) | |
| print(count, response.status_code) | |
| if response.status_code != 200: | |
| print("Error!") | |
| print(response.json()) | |
| break | |
| json_response = response.json() | |
| with open(f"tweets_{count}.json", 'w') as handler: | |
| handler.write(json.dumps(json_response, indent=2)) | |
| if "meta" in json_response and "next_token" in json_response["meta"]: | |
| start_token = json_response["meta"]["next_token"] | |
| # 15 minutes -> 900 seconds | |
| # 300 requests / 15 minute => 1 request / 3 second | |
| ## sleep for 3s to avoid breaking stuff | |
| time.sleep(3) | |
| count = count + 1 | |
| if __name__ == "__main__": | |
| parser = argparse.ArgumentParser() | |
| parser.add_argument('-i', '--id', help='user id', type=str, required=True) | |
| args = parser.parse_args() | |
| print(args) | |
| main(args) |