profile_performance.py

import time
import pandas as pd
import preprocessor
import random

def generate_large_chat(lines=10000):
    """Generates a synthetic WhatsApp chat log."""
    senders = ["User1", "User2", "User3"]
    messages = [
        "Hello there, how are you?",
        "I am doing great, thanks for asking! Project update?",
        "This is a test message to simulate a long chat about artificial intelligence.",
        "Meeting is at 10 AM tomorrow to discuss the roadmap.",
        "Check out this link: https://example.com",
        "Haha that is funny 😂",
        "Je parle un peu français aussi. C'est la vie.",
        "Non, je ne crois pas. Il fait beau aujourd'hui.",
        "Ok, see you later. Don't forget the deadline.",
        "Python is a great programming language for data science.",
        "Streamlit makes building apps very easy and fast."
    ]
    
    chat_data = []
    for _ in range(lines):
        date = f"{random.randint(1, 12)}/{random.randint(1, 28)}/23"
        hour = random.randint(1, 12)
        minute = random.randint(10, 59)
        ampm = random.choice(["AM", "PM"])
        time_str = f"{hour}:{minute} {ampm}"
        sender = random.choice(senders)
        message = random.choice(messages)
        chat_data.append(f"{date}, {time_str} - {sender}: {message}")
    
    return "\n".join(chat_data)

def profile_preprocessing():
    print("Generating synthetic data (10,000 lines)...")
    raw_data = generate_large_chat(10000)
    print(f"Data size: {len(raw_data) / 1024 / 1024:.2f} MB")
    
    print("\nStarting profiling...")
    start_total = time.time()
    
    # We can't easily profile inside the function without modifying it, 
    # so we will measure the total time and infer from code analysis 
    # or modify preprocessor.py temporarily to print timings.
    # For now, let's just run it and see the total time.
    
    try:
        start_time = time.time()
        
        # Step 1: Parse
        df = preprocessor.parse_data(raw_data)
        print(f"Parsing took: {time.time() - start_time:.2f}s")
        
        # Step 2: Analyze
        step_start = time.time()
        df, topics = preprocessor.analyze_sentiment_and_topics(df)
        print(f"Analysis took: {time.time() - step_start:.2f}s")
        end_total = time.time()
        print(f"\nTotal Preprocessing Time: {end_total - start_total:.2f} seconds")
        print(f"Messages processed: {len(df)}")
        
    except Exception as e:
        print(f"Error: {e}")
        import traceback
        traceback.print_exc()

if __name__ == "__main__":
    profile_preprocessing()