azure-python-signalr

A selfhosted client which uses Azure Function endpoints to get requests form the client and then uses a SignalR service to send replies back.

Setup

The following setup is done using the terminal in VS Code.

The solution uses a SignalR service in Azure, so set that up in the portal. It can run on the free tier. The important setting is that Service Mode is set to Serverless. This is required in order to have the Azure Function use it. Once created add the connectionString for the SignalR service to the local.settings.json file by using this command in the terminal func settings add AzureSignalRConnectionString "<signalr-connection-string>"

The solution should run in a vitual environment, VS Code set that up once it detects the python function app.

In order to make the Azure Function run locally we use Azurite. Install that and then from the command line (Ctrl+Shift+P) run Azurtie: Start.

Once the setup is complete the solution is ready to run.

Function endpoints.

All endpoints expect body user input in json format (unless none stated). All endpoints return to 'target': 'newMessage' (newMessage). The app has the following endpoints;

1. index - Launches the selfhosted website which displays messages from SignalR using the mock_query method.
2. mock_query - ["POST"] (Broadcast) Simulates program execution, waits two programmatically set times. Wait time is in seconds. Before each wait a status message is returned to all clients.

{
    "wait1": "<wait in seconds>",
    "wait2": "<wait in seconds>"
}

3. send_to_user - (Single User) Sends a message to a single user based the id created in the SignalR setup. Returns a simple string "Hello userId {userId}, this is for you only."

{
    "userId": "<userId>"
}

4. send_status - "Private" endpoint used for sending out status messages with SignalR.
5. negotitate - ["POST"] Initializes the connection with SignalR and the client.