langchaingo/examples/google-cloudsql-chat-message-history-example

Google CloudSQL Chat Message History Example

This example demonstrates how to use CloudSQL for Postgres as a backend for the ChatMessageHistory for LangChain in Go.

What This Example Does

1. Creates a CloudSQL Chat Message History:

   • Initializes the cloudsql.PostgresEngine object to establish a connection to the CloudSQL database.
   • Creates a new table to store the chat message history if it doesn't exist.
   • Initializes a cloudsql.ChatMessageHistory object, which provides methods to store, retrieve, and clear message contents with a specific session ID.

2. Add Single Messages:

   • Creates individual AI and Human messages and stores them in the chat message history.
   • Prints the stored messages to the console.

3. Add Multiple Messages:

   • Creates multiple messages of different types (AI and Human) and stores them all at once in the chat message history.
   • Prints the stored messages to the console.

4. Overwrite Messages:

   • Clears the existing messages and then stores a new set of messages.
   • Prints the updated list of stored messages.

5. Clear All Messages:

   • Clears all messages stored for the current session.

How to Run the Example

1. Set the following environment variables. Your CloudSQL values can be found in the Google Cloud Console:

   export PROJECT_ID=<your project Id>
   export POSTGRES_USERNAME=<your user>
   export POSTGRES_PASSWORD=<your password>
   export POSTGRES_REGION=<your region>
   export POSTGRES_INSTANCE=<your instance>
   export POSTGRES_DATABASE=<your database>
   export POSTGRES_TABLE=<your tablename>
   export POSTGRES_SESSION_ID=<your sessionID>
   

2. Run the Go example:

   go run google_cloudsql_chat_message_history_example.go
   

Key Features

• CloudSQL Integration: Connects to a CloudSQL instance for storing and managing chat messages.
• Message Storage: Provides methods for storing chat messages with different operations like add, overwrite, and clear.
• Session-based Message Management: Messages are stored and retrieved using a unique session ID, making it easy to manage separate chat sessions.
• Clear and Overwrite Capabilities: Allows overwriting and clearing messages to maintain the chat history as needed.