Process chatgpt conversations.json into the length of time of conversations

to_time.py

# GOAL:
# Determine how long (in seconds) you've spent talking to ChatGPT. 
# This goes through each message sent in order. If there are two messages closer than 5 minutes apart,
# it adds their time delta to the total. If the gap is over 5 minutes, it assumes this is idle time
# and ignores the time delta entirely.

# Outputs a CSV file of total time per conversation, and prints the grand total of all conversations.

# All output is in seconds. 

import json
import csv

# Load JSON data
with open('conversations.json', 'r') as file:
    data = json.load(file)

# Initialize a variable to hold the sum of all conversation times
total_time_of_all_conversations = 0

# Prepare CSV output
with open('output_times.csv', 'w', newline='') as csvfile:
    fieldnames = ['first_message_time', 'last_message_time', 'total_time_seconds']
    writer = csv.DictWriter(csvfile, fieldnames=fieldnames)
    
    writer.writeheader()
    
    for item in data:
        # Extract all create_time values, ignoring None
        times = [message['message']['create_time'] for key, message in item['mapping'].items() if message['message'] and message['message']['create_time'] is not None]
        
        if not times:
            continue  # Skip to the next conversation if no messages
        
        # Sort times to find the first and last message times
        sorted_times = sorted(times)
        first_time = sorted_times[0]
        last_time = sorted_times[-1]
        
        # Initialize total active time for the current conversation
        total_active_time = 0
        
        # Iterate through sorted times and calculate active time
        for i in range(len(sorted_times) - 1):
            current_time = sorted_times[i]
            next_time = sorted_times[i + 1]
            time_diff = next_time - current_time
            
            # If two messages are <= 5 minutes (300 seconds) apart, add their time difference
            if time_diff <= 300:
                total_active_time += time_diff
        
        # For conversations with only one message, consider active time as 1 second
        if len(sorted_times) == 1:
            total_active_time = 1
        
        # Accumulate total time of all conversations
        total_time_of_all_conversations += total_active_time
        
        # Write to CSV
        writer.writerow({
            'first_message_time': first_time,
            'last_message_time': last_time,
            'total_time_seconds': total_active_time
        })

# Print the total time of all conversations in seconds
print(f'Total time of all conversations: {total_time_of_all_conversations} seconds')