Enhance User Session Management with TTL Functionality (#18)

* Update main.py

* Update README.md

* Update gcp-deploy.yml

.github/workflows/gcp-deploy.yml

@@ -23,4 +23,4 @@ jobs:
         with:
           project_id: ${{ secrets.GCP_PROJECT_ID }}
 
-      - run: gcloud functions deploy ${{ secrets.GCP_FUNCTION }} --runtime python311 --trigger-http --allow-unauthenticated --entry-point process_event --region ${{ secrets.GCP_REGION }} --set-env-vars OPENAI_API_KEY=${{ secrets.OPENAI_API_KEY }},MODEL_NAME=${{ secrets.MODEL_NAME }},SYSTEM_PROMPT="${{ secrets.SYSTEM_PROMPT }}",MAX_TURNS="${{ secrets.MAX_TURNS }}"
+      - run: gcloud functions deploy ${{ secrets.GCP_FUNCTION }} --runtime python311 --trigger-http --allow-unauthenticated --entry-point process_event --region ${{ secrets.GCP_REGION }} --set-env-vars OPENAI_API_KEY=${{ secrets.OPENAI_API_KEY }},MODEL_NAME=${{ secrets.MODEL_NAME }},SYSTEM_PROMPT="${{ secrets.SYSTEM_PROMPT }}",MAX_TURNS="${{ secrets.MAX_TURNS }}",TTL="${{ secrets.TTL }}"

README.md

@@ -8,6 +8,7 @@ Chat²GPT is a ChatGPT chat bot for Google Chat. 💬🤖 It enhances communicat
 
 ## 📖 Table of Contents
 - [🛠️ Setup](#%EF%B8%8F-setup)
+- [🧑‍💻 Usage](#-usage)
 - [🌐 Community](#-community)
   - [Contributing 👥🤝](#contributing-)
   - [Reporting Bugs 🐛📝](#reporting-bugs-)
@@ -58,6 +59,8 @@ In your GitHub repository:
   - `OPENAI_API_KEY`: Your OpenAI API key.
   - `MODEL_NAME`: The name of the OpenAI model you're using. For this project, we recommend "gpt-3.5-turbo".
   - `SYSTEM_PROMPT`: The system prompt to use for the OpenAI API.
+  - `MAX_TURNS`: This sets the maximum number of exchanges the bot remembers in a user session before resetting. Default: 10 exchanges.
+  - `TTL`: This sets the duration (in seconds) a user session stays active from the last received message before it resets. Default: 600 seconds (10 minutes).
 
 **5. GitHub Actions 🚀**
 
@@ -73,7 +76,9 @@ The GitHub Actions workflow is configured to automatically deploy the bot to Goo
 
 Now, your bot can be added to any room within your Google Workspace.
 
-Your bot is now ready! It can interact in any chat room when it's explicitly mentioned (@botname) or directly messaged, depending on the functionality you've programmed it with.
+## 🧑‍💻 Usage
+
+Your bot is all set and ready for action! It's capable of interacting in any chat room, responding when directly mentioned (@botname), or when it receives a direct message, based on the functionality you've programmed. Our bot is designed to remember several rounds of a conversation per user session, providing a coherent and continuous interaction. This means you can ask a question, receive a response, and continue the conversation by referencing the initial query or its response. However, for performance optimization, the length of the conversation is limited by a configurable setting; we recommend setting it to 5-10 turns. Importantly, each session also tracks the time since the last received message, automatically resetting if it exceeds the specified 'Time to Live' (TTL) value. This ensures a seamless and efficient conversation experience with the bot.
 
 ## 🌐 Community
 

main.py

@@ -1,4 +1,5 @@
 import os
+import datetime
 from flask import jsonify
 from simpleaichat import AIChat
 
@@ -24,9 +25,16 @@
 except Exception as e:
     print(f"Error getting max turns: {str(e)}")
 
+# Try to get the TTL from an environment variable
+try:
+    TTL = int(os.getenv('TTL', 600))  # Default to 600 seconds (10 minutes)
+except Exception as e:
+    print(f"Error getting TTL: {str(e)}")
+
 # Define globals
 user_sessions = {}  # A dictionary to track the AIChat instances for each user
 turn_counts = {}  # A dictionary to track the turn count for each user
+last_received_times = {}  # A dictionary to track the last received time for each user
 
 def process_event(request):
     try:
@@ -57,24 +65,29 @@ def process_event(request):
 
 def handle_message(user_id, user_message):
     try:
+        current_time = datetime.datetime.now()
+
         # Get the AIChat instance for the user, or create a new one
         ai_chat = user_sessions.get(user_id)
         turn_count = turn_counts.get(user_id, 0)
-        if ai_chat is None or turn_count >= MAX_TURNS:
+        last_received_time = last_received_times.get(user_id)
+
+        if ai_chat is None or turn_count >= MAX_TURNS or (last_received_time is not None and (current_time - last_received_time).total_seconds() > TTL):
             ai_chat = AIChat(api_key=openai_api_key, system=system_prompt)
             user_sessions[user_id] = ai_chat
             turn_count = 0
 
         # Generate the response
         response = ai_chat(user_message)
 
-        # Update the turn count
+        # Update the turn count and the last received time
         turn_count += 1
         turn_counts[user_id] = turn_count
+        last_received_times[user_id] = current_time
 
         bot_message = response
 
     except Exception as e:
         print(f"Error calling OpenAI API: {str(e)}")
         bot_message = "Sorry, I'm currently unable to generate a response."
-    return jsonify({'text': bot_message})
+    return jsonify({'text': bot_message})