monster_mash

• Typhoeus is an ancient monster.
• A monster mash is a dance party.
• This library inspired by John Nunemaker's awesomely useful HTTParty.
• By law, all Ruby libraries have to have dumbass names.

This library wraps Typhoeus and Typhoeus::Hydra and exposes an easy-to-use DSL for quickly building libraries to interact with HTTP resources. Every method you write will automatically export serial (blocking) and parallel (non-blocking) methods, so you can easily parallelize your HTTP code when possible.

monster_mash has a Sinatra-like syntax, and lets you build client API methods using the 4 HTTP verbs:

• get(method_name, &definition_block)
• post(method_name, &definition_block)
• put(method_name, &definition_block)
• delete(method_name, &definition_block)

Within each definition_block, you can set various Typhoeus options.

• uri (required) - the URI to hit
• handler (required) - a block to handle an HTTP response
• params - hash of URI params
• body - post body
• headers - hash of HTTP headers to send
• timeout - how long to timeout
• cache_timeout - how long to keep HTTP calls cached
• user_agent - a User-Agent string to send
• max_redirects - max number of redirects to follow
• disable_ssl_peer_verification - whether to disable SSL verification

class GoogleJson < MonsterMash::Base
  VERSION = '1.0'

  # Creates a method called +search+ that takes
  # a single +query+ parameter.
  get(:search) do |query|
    uri "http://ajax.googleapis.com/ajax/services/search/web"
    params 'v' => VERSION,
           'q' => query,
           'rsz' => 'large'
    handler do |response|
      json = JSON.parse(response.body)

      # returns results
      json['responseData']['results']
    end
  end
end

To make serial (blocking) calls using this code, you would then call the class method:

# blocks
results = GoogleJson.search("my search query")
results.each do |result|
  puts result['unescapedUrl']
  # do other stuff with the response
end

The search(query) method returns whatever your handler block returns.

To make parallel (non-blocking) calls, you need an instance of Typhoeus::Hydra:

hydra = Typhoeus::Hydra.new
google = GoogleJson.new(hydra)
10.times do
  google.search("my query") do |results, error|
    if error
      # handle error
    else
      results.each do |result|
        puts result['unescapedUrl']
      end
    end
  end
end

# blocks until all 10 queries complete.
hydra.run

monster_mash will correctly delegate method calls from your handler block to your API class. Example:

class GoogleJson < MonsterMash::Base
  VERSION = '1.0'

  # Creates a method called +search+ that takes
  # a single +query+ parameter.
  get(:search) do |query|
    uri "http://ajax.googleapis.com/ajax/services/search/web"
    params 'v' => VERSION,
           'q' => query,
           'rsz' => 'large'
    handler do |response|
      json = JSON.parse(response.body)

      # Calls the correct method on GoogleJson.
      parse_results(json)
    end
  end

  def self.parse_results(json)
    json['responseData']['results']
  end
end

If you have Typhoeus settings you want to happen for every request, you can set them in a defaults block:

class GoogleJson < MonsterMash::Base
  defaults do
    user_agent "GoogleJson Ruby Library"
    disable_ssl_peer_verification true
  end

  # ...
end

If all of your requests share a common base URI, you can set that in your defaults block:

class GoogleJson < MonsterMash::Base
  defaults do
    base_uri "http://google.com"
  end

  get(:search) do
    uri "/search" # expands to http://google.com/search
  end

  post(:authenticate) do
    uri "https://auth.google.com" # ignores the base_uri
  end
end

As well, if you set params or headers in the defaults block, any params or headers added later will be merged into the hash.

class GoogleJson < MonsterMash::Base
  defaults do
    params 'api_key' => 'fdas'
  end

  # The full params hash will look like:
  #   :q => +query+,
  #   :v => '1.0',
  #   :api_key => 'fdas'
  get(:search) do |query|
    params 'q' => query,
           'v' => '1.0'
    uri "..."
    handler do |response|
      # ...
    end
  end
end

• All serial (blocking) methods will simply raise an error if anything wrong happens. You just need to rescue said error.
• When interacting with Hydra requests, the block you pass to it will receive an error to it if any error was caught during the handler's run. You need to check for the error in your block and handle it there.

• http://github.com/dbalatero/alchemy_api

• Fork the project.
• Make your feature addition or bug fix.
• Add tests for it. This is important so I don't break it in a future version unintentionally.
• Commit, do not mess with rakefile, version, or history. (if you want to have your own version, that is fine but bump version in a commit by itself I can ignore when I pull)
• Send me a pull request. Bonus points for topic branches.

Copyright (c) 2010 David Balatero. See LICENSE for details.