Data.py

###############################################
import tweepy
import pandas as pd
import time
import os

# 🔐 Add your token here
BEARER_TOKEN = 'AAAAAAAAAAAAAAAAAAAAAKuZ2gEAAAAAjXozzWflF%2FYO9FsZ1CuRuNGAeSI%3DS8fg9ON3DifgUCIqIDZ9KMvbWp3gcT7BXrL5i3fxi1DHDu4e5T'

client = tweepy.Client(bearer_token=BEARER_TOKEN)

queries = [
    "#AI"
]

all_data = []

for tag in queries:
    print(f"📡 Collecting tweets for: {tag}")
    query = f"{tag} lang:en -is:retweet"

    success = False
    while not success:
        try:
            response = client.search_recent_tweets(query=query, max_results=900)

            if response.data:
                for tweet in response.data:
                    text = tweet.text
                    hashtags = [word for word in text.split() if word.startswith("#")]
                    if hashtags:
                        all_data.append({
                            "text": text,
                            "hashtags": " ".join(hashtags)
                        })

            success = True
            time.sleep(3)  # gentle delay between calls

        except tweepy.TooManyRequests:
            print("⏳ Rate limit hit. Sleeping for 15 minutes...")
            time.sleep(15 * 60)
        except ConnectionResetError as e:
            print(f"❌ Connection error: {e}. Retrying in 1 minute...")
            time.sleep(60)
        except Exception as e:
            print(f"❌ Unknown error: {e}. Skipping to next hashtag.")
            success = True  # skip to next hashtag to avoid infinite loop

# Save CSV
if all_data:
    df = pd.DataFrame(all_data)
    df.to_csv("all_tweets.csv", index=False)
    print(f"✅ Saved {len(df)} tweets to 'all_tweets.csv'")
else:
    print("⚠️ No tweets collected. Try again later.")