BraintreeHttp is a generic HTTP Client.

In it's simplest form, an HttpClient exposes an #execute method which takes an HTTP request, executes it against the domain described in an Environment, and returns an HTTP response.

An Environment describes a domain that hosts a REST API, against which an HttpClient will make requests. Environment is a simple class that contains one property, base_url.

env = Environment.new('https://example.com')

HTTP requests contain all the information needed to make an HTTP request against the REST API. Specifically, one request describes a path, a verb, any path/query/form parameters, headers, attached files for upload, and body data. In Ruby, an HTTP request is simply an OpenStruct literal with path, verb, and optionally, request_body, and headers populated.

HTTP responses contain information returned by a server in response to a request as described above. They are simple objects which contain a status_code, headers, and a result, which reprepsents any data returned by the server.

req = OpenStruct.new({
  :path => 'path/to/resource',
  :verb => 'GET',
  :headers => {
    'X-Custom-Header' => 'custom_value'
  }
})

resp = client.execute(req)
status_code = resp.status_code;
headers = resp.headers;
response_data = resp.result;

Injectors are blocks that can be used for executing arbitrary pre-flight logic, such as modifying a request or logging data. Injectors are attached to an HttpClient using the #add_injector method. They must take one argument (a request).

The HttpClient executes its injectors in a first-in, first-out order, before each request.

let client = new HttpClient(env);
client.add_injector do |request|
  puts req
end

client.add_injector do |request|
  req.headers['Request-Id'] = 'abcd'
end

...

HttpClient#execute may throw an IOError if something went wrong during the course of execution. If the server returned a non-200 response, this execption will be an instance of HttpError that will contain a status code and headers you can use for debugging.

begin
  resp = http_client.execute(req)
rescue => e
  if e.is_a? HttpError
    # Inspect this exception for details
    status_code = e.status_code;
  end
end

(De)Serialization of request and response data is done by instances of Encoder. BraintreeHttp currently supports json encoding out of the box.

BraintreeHttp-Ruby is open source and available under the MIT license. See the LICENSE file for more information.

Pull requests and issues are welcome. Please see CONTRIBUTING.md for more details.