HTTP Capture

Allows access to the responses from a wide range of ruby HTTP libraries. This is super helpful for testing (especially with Cucumber).

Currently supports:

Net::HTTP (and anything that uses it, specifically including:)
- Mechanize
- HTTParty
HTTPClient
Rack::Test

They can all be used by requiring <library folder space>/capture from your code, eg:

require 'net/http/capture'
require 'mechanize/capture'
require 'httparty/capture'
require 'httpclient/capture'
require 'rack/test/capture'

Install

Install the usual way:

$ gem install http_capture

Or require it from within your gemspec and use bundle.

Usage

Because it's beneficial to use a number of different HTTP libraries during our testing I've written a small script to capture the HTTP response information from each of the supported libraries so they can be examined after-the-fact. For example:

require 'httpclient'
# Include the relevant capture library
require 'httpclient/capture'

# Perform a request, as usual
clnt = HTTPClient.new
clnt.get_content("http://google.com") # => "<html>..."

# Now you can retrieve all the captured responses:
HttpCapture::RESPONSES.last.status # => 200
HttpCapture::RESPONSES.last['Content-Length'] # => 12345
HttpCapture::RESPONSES.last.body # => "<html>..."
HttpCapture::RESPONSES.last.response # => <HTTPClient::Response>

The objects in the HttpCapture::RESPONSES array contain the exact objects, wrapped by helper functions which expose three methods: code for the status code, [](key) for accessing headers and body for reading the body of the document.