Create chatgptarchive.py

ChatGPTArchive/chatgptarchive.py

@@ -0,0 +1,107 @@
+# chatgptarchive.py
+# Description: Parse and save conversations from a JSON file to individual files
+import json
+import os
+import re
+from datetime import datetime
+
+
+def parse_conversations(conversations):
+    parsed_data = []
+
+    for conversation in conversations:
+        conversation_id = conversation.get('id')
+        title = conversation.get('title')
+
+        # Handle None or empty titles
+        if not title:
+            title = 'conversation'
+
+        title = title.replace(" ", "_")
+        messages = conversation.get('mapping', {}).values()
+
+        conversation_data = {
+            'conversation_id': conversation_id,
+            'title': title,
+            'messages': []
+        }
+
+        for message_entry in messages:
+            if not message_entry or not message_entry.get('message'):
+                continue
+
+            message = message_entry['message']
+            author = message.get('author', {}).get('role') if message.get('author') else None
+            content = message.get('content', {}).get('parts', []) if message.get('content') else []
+            create_time = message.get('create_time')
+            update_time = message.get('update_time')
+            parent_id = message_entry.get('parent')
+            children_ids = message_entry.get('children', [])
+
+            conversation_data['messages'].append({
+                'message_id': message_entry.get('id'),
+                'author': author,
+                'content': content,
+                'create_time': create_time,
+                'update_time': update_time,
+                'parent_id': parent_id,
+                'children_ids': children_ids
+            })
+
+        parsed_data.append(conversation_data)
+
+    return parsed_data
+
+
+def sanitize_filename(filename):
+    """Sanitize the filename by removing invalid characters and ensuring it is not empty."""
+    sanitized = re.sub(r'[^a-zA-Z0-9_-]', '', filename)
+    return sanitized if sanitized else 'conversation'
+
+
+def save_conversations_to_files(parsed_conversations, output_dir):
+    if not os.path.exists(output_dir):
+        os.makedirs(output_dir)
+
+    filename_counts = {}
+
+    for conversation in parsed_conversations:
+        title = conversation['title']
+        base_filename = sanitize_filename(title[:10])
+
+        if base_filename in filename_counts:
+            filename_counts[base_filename] += 1
+            filename = f"{base_filename}_{filename_counts[base_filename]}.json"
+        else:
+            filename_counts[base_filename] = 1
+            filename = f"{base_filename}.json"
+
+        filepath = os.path.join(output_dir, filename)
+
+        with open(filepath, 'w') as file:
+            json.dump(conversation, file, indent=4)
+
+
+def main():
+    # Prompt the user for the path to the conversations.json file
+    conversations_json_path = input("Enter the path to the conversations.json file: ").strip()
+
+    # Load the JSON data
+    with open(conversations_json_path, 'r') as file:
+        conversations_data = json.load(file)
+
+    # Parse the conversations data
+    parsed_conversations = parse_conversations(conversations_data)
+
+    # Define the output directory with a timestamp
+    timestamp = datetime.now().strftime("%d%m%y")
+    output_dir = os.path.join(os.path.dirname(conversations_json_path), f'gptlogs-{timestamp}')
+
+    # Save each parsed conversation to its own file
+    save_conversations_to_files(parsed_conversations, output_dir)
+
+    print(f"Parsed conversations have been saved to {output_dir}")
+
+
+if __name__ == "__main__":
+    main()