HTTPI

HTTPI provides a common interface for Ruby HTTP libraries.

Wishlist | Bugs | Docs

Installation

The gem is available through Rubygems and can be installed via:

$ gem install httpi

Some examples

Executing a POST request with the most basic request object:

request = HTTPI::Request.new "http://example.com"
HTTPI.get request

Here's a POST request with a request object:

request = HTTPI::Request.new
request.url = "http://post.example.com"
request.body = "send me"

HTTPI.post request

And a GET request using HTTP basic auth and the Curb adapter:

request = HTTPI::Request.new
request.url = "http://auth.example.com"
request.auth.basic "username", "password"

HTTPI.get request, :curb

HTTPI also comes shortcuts. This executes a PUT request:

HTTPI.put "http://example.com", "<some>xml</some>"

And this executes a DELETE request:

HTTPI.delete "http://example.com"

HTTPI

The HTTPI module uses one of the available adapters to execute HTTP requests.

GET

HTTPI.get(request, adapter = nil)
HTTPI.get(url, adapter = nil)

POST

HTTPI.post(request, adapter = nil)
HTTPI.post(url, body, adapter = nil)

HTTPI.head(request, adapter = nil)
HTTPI.head(url, adapter = nil)

PUT

HTTPI.put(request, adapter = nil)
HTTPI.put(url, body, adapter = nil)

DELETE

HTTPI.delete(request, adapter = nil)
HTTPI.delete(url, adapter = nil)

Notice

  • You can specify the adapter to use per request
  • And request methods always return an HTTPI::Response

More control

If you need more control over the request, you can access the HTTP client instance represented by your adapter in a block:

HTTPI.post request do |http|
  http.use_ssl = true  # Curb example
end

HTTPI::Adapter

HTTPI uses adapters to support multiple HTTP libraries. It currently contains adapters for:

You can manually specify the adapter to use via:

HTTPI::Adapter.use = :curb  # or one of [:httpclient, :net_http]

If you don't specify which adapter to use, HTTPI try to load HTTPClient, then Curb and finally NetHTTP.

Notice: HTTPI does not force you to install any of these libraries. If you'd like to use on of the more advanced libraries (HTTPClient or Curb), you have to make sure they're in your LOAD_PATH. HTTPI will then load the library when executing HTTP requests.

HTTPI::Request

The HTTPI::Request serves as a common denominator of options that HTTPI adapters need to support.
It represents an HTTP request and lets you customize various settings through the following methods:

#url           # the URL to access
#proxy         # the proxy server to use
#ssl           # whether to use SSL
#headers       # a Hash of HTTP headers
#body          # the HTTP request body
#open_timeout  # the open timeout (sec)
#read_timeout  # the read timeout (sec)

Usage example

request = HTTPI::Request.new
request.url = "http://example.com"
request.read_timeout = 30

HTTPI::Auth

HTTPI::Auth supports HTTP basic and digest authentication.

#basic(username, password)   # HTTP basic auth credentials
#digest(username, password)  # HTTP digest auth credentials
#ntlm(username, password)    # NTLM auth credentials

Usage example

request = HTTPI::Request.new
request.auth.basic "username", "password"

HTTPI::Auth::SSL

HTTPI::Auth::SSL manages SSL client authentication.

#cert_key_file  # the private key file to use
#cert_file      # the certificate file to use
#ca_cert_file   # the ca certificate file to use
#verify_mode    # one of [:none, :peer, :fail_if_no_peer_cert, :client_once]

Usage example

request = HTTPI::Request.new
request.auth.ssl.cert_key_file = "client_key.pem"
request.auth.ssl.cert_key_password = "C3rtP@ssw0rd"
request.auth.ssl.cert_file = "client_cert.pem"
request.auth.ssl.verify_mode = :none

HTTPI::Response

As mentioned before, every request method return an HTTPI::Response. It contains the response code, headers and body.

response = HTTPI.get request

response.code     # => 200
response.headers  # => { "Content-Encoding" => "gzip" }
response.body     # => "<!DOCTYPE HTML PUBLIC ...>"

The response.body handles gzipped and DIME encoded responses.

TODO

  • Return the original HTTPI::Request for debugging purposes
  • Return the time it took to execute the request

Logging

HTTPI by default logs each HTTP request to STDOUT using a log level of :debug.

HTTPI.log       = false     # disabling logging
HTTPI.logger    = MyLogger  # changing the logger
HTTPI.log_level = :info     # changing the log level

Participate

Any help and feedback appreciated. So please get in touch!