Custom HTTP Clients for the Twilio Ruby Helper Library

If you are working with the Twilio Ruby Helper Library, and you need to be able to modify the HTTP requests that the library makes to the Twilio servers, you’re in the right place. The most common need to alter the HTTP request is to connect and authenticate with an enterprise’s proxy server. We’ll provide sample code that you can drop right into your app to handle this use case.

Connect and authenticate with a proxy server

To connect and provide credentials to a proxy server that may be between your app and Twilio, you need a way to modify the HTTP requests that the Twilio helper library makes on your behalf to invoke the Twilio REST API.

The Twilio Ruby helper library uses the Faraday gem under the hood to make the HTTP requests. You can provide your own httpClient for making API requests.

So the question becomes how do we apply this to a typical Twilio REST API example?

@client = Twilio::REST::Client.new(account_sid, auth_token)

message = @client.messages
  .create(
     to: '+15558675310',
     body: 'Hey there!',
     from: '+15017122661'
   )

Where does a http_client get created and used? Out of the box, the helper library is creating a default Twilio::Http::Client for you, using the Twilio credentials you pass to the initialize method. However, there’s nothing stopping you from creating your own and using that.

Once you have your own Twilio::Http::Client, you can pass it to any Twilio REST API resource action you want. Here’s an example of sending an SMS message with a custom client:

Create your custom TwilioRestClient

When you take a closer look at the constructor for Twilio::Http::Client, you see that this class provides it to the Twilio helper library to make the necessary HTTP requests.

Call Twilio through the proxy server

Now that we understand how all the components fit together, we can create our own http_client that can connect through a proxy server. To make this reusable, here’s a class that you can use to create this http_client whenever you need one.

In this example, we are using some environment variables loaded at the program startup to retrieve various configuration settings:

• Your Twilio Account Sid and Auth Token (found here, in the Twilio console)

• A proxy address in the form of http://127.0.0.1:8888

These settings are located in a file like .env like so:

ACCOUNT_SID=ACxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
AUTH_TOKEN= your_auth_token

HTTPS_PROXY=https://127.0.0.1:8888
HTTP_PROXY=http://127.0.0.1:8888

Here’s the full console program that sends a text message and shows how it all can work together. It loads the .env file.

What else can this technique be used for?

Now that you know how to inject your own http_client into the Twilio API request pipeline, you could use this technique to add custom HTTP headers and authorization to the requests (perhaps as required by an upstream proxy server).

You could also implement your own http_client to mock the Twilio API responses so your unit and integration tests can run quickly without the need to make a connection to Twilio. In fact, there’s already an example online showing how to do exactly that on C#.

We can’t wait to see what you build!