New York, New York.

(ridiculously) small and powerful micro web framework.

Install the gem:

gem install nyny

Create the myapp.rb file:

# myapp.rb

require 'nyny'
class App < NYNY::App
  get '/' do
    'Hello world!'
  end
end

App.run!

Run the file:

ruby myapp.rb

Open the browser at http://localhost:9292

• TOP
• Motivation
• Philosophy
• Why use NYNY
• Usage
  • Environment
  • Configuration
  • Running
  • Defining routes
  • Request scope
  • Maps
  • Namespaces
  • Templates
  • Filters
  • Middleware
  • Helpers
  • Extensions
• FAQ
• Contributing

Philosophy

NYNY is unassuming, it has all the core stuff to get running, but nothing else. Your app is the framework. However, it's trivial to extend NYNY via its extension interface.

Why use NYNY

• It's very small (~300 LOC), which is just a little overhead on top of Rack.
• You want to dig into the source code and change to your needs (NYNY's source code is more welcoming)
• Each NYNY app is a Rack middleware, so it can be used inside of Sinatra, Rails, or any other Rack-based app.
• It uses Journey for routing (Rails' router), which makes its routing logic a lot more powerful and reliable that in most micro web frameworks.

Usage

A NYNY app must always be in a class which inherits from NYNY::App:

class App < NYNY::App
  get '/' do
    'Hello, World'
  end
end

Environment

To get the directory in which your app is running use NYNY.root

#/some/folder/server.rb
require 'nyny'
puts NYNY.root #=> /some/folder/
puts NYNY.root.join("foo") #=> /some/folder/foo

To get NYNY's environment, use NYNY.env

#env.rb
require 'nyny'
puts NYNY.env
puts NYNY.env.production?

$ ruby env.rb
development
false

$ ruby env.rb RACK_ENV=production
production
true

Configuration

You can configure your app by attaching arbitrary properties to config object:

class App < NYNY::App
  config.foo = 'bar'
end

App.config.foo #=> 'bar'

Or, you can use the configure block:

class App < NYNY::App
  configure do
    config.always = true
  end

  configure :production do
    config.prod = true
  end

  configure :test, :development do
    config.unsafe = true
  end
end

Also, NYNY provides a simple api for hooking into the app's initialization:

class App < NYNY::App
  before_initialize do |app|
    #this will be executed just before the Rack app is compiled
    #'app' is the a App instance
  end

  after_initialize do |app, rack_app|
    #this will be executed after the Rack app is compiled
    #'app' is the a App instance
    #'rack_app' is the main block which will be called on any request
  end
end

Running

There are two ways to run a NYNY app directly [?]:

• by requiring it in a config.ru file, and then passing it as argument to the Rack's run function:

# config.ru

require 'app'
run App.new

• by using the run! method directly on the app class:

# app.rb

# ...app class definition...

App.run!

run! takes the port number as optional argument (the default port is 9292). Also the run! method will include 2 default middlewares to make the development easier: Rack::CommonLogger and BetterErrors::Middleware (only in dev). This will show all requests in the log, and will provide useful details in the case a error occurs during a request.

Defining routes

NYNY uses Journey for routing, that means that NYNY has all the awesomeness the Rails' router has. NYNY supports the following verbs for defining a route: delete, get, head, options, patch, post, put and trace.

class App < NYNY::App
  post '/' do
    'You Posted, dude!'
  end
end

You can use any construct or convention supported in Rails for the path string.

Each route definition call optionally accepts constraints:

class App < NYNY::App
  get '/', :constraints => {:content_type => 'text/html'} do
    'html'
  end
end

What are constraints? Constraints are assertions on the request object. That means that the route from above will match only if request.content_type is 'text/html'.

To group multiple routes for a single constraint, use the constraints block:

class App < NYNY::App
  constraints :content_type => 'text/html' do
    get '/' do
      'html'
    end

    get '/foo' do
      'foo html'
    end
  end
end

Besides the constraints, you can specify defaults:

class App < NYNY::App
  get '/', :defaults => {:format => 'html'} do
    'html'
  end
end

Each block that is passed to a route definition is evaluated in the context of a request scope. See below what methods are available there.

Request scope

As was said above, when you pass a block to a route definition, that block is evaluated in the context of a RequestScope. This means that several methods/objects available inside that block:

• request - A Rack::Request object which encapsulates the request to that route. (see Rack::Request documentation for more info)
• response - A Rack::Response object which encapsulates the response.
• params - a hash which contains both POST body params and GET querystring params.
• headers - a hash with the response headers (ex: headers['Content-Type'] = 'text/html')
• status - allows you to set the status of the response (ex: status 403)
• redirect_to - sets the response to redirect (ex: redirect_to 'http://google.com')
• cookies - a hash which allows you to access/modify/remove cookies (ex: cookies[:foo] = 'bar' or cookies.delete[:foo])
• session - a hash which allows you to access/modify/remove session variables (ex: session[:foo] = 'bar')
• halt - allows you to instantly return a response, interrupting current handler execution (see halt)

Maps

NYNY uses a Rack::Builder underneath to generate a proper Rack app when it is being run. Because of that, NYNY exposes the Builder's map method to allow fine grained low-level url mapping. The most common use case for that is to integrate sprockets into a NYNY app:

require 'sprockets'
require 'nyny'

class AppWithAssets < NYNY::App
  map '/assets' do
    environment = Sprockets::Environment.new
    environment.append_path 'app/assets/javascripts'
    environment.append_path 'app/assets/stylesheets'
    run environment
  end

  get '/' do
    'hello'
  end
end

Please note that none of the NYNY's routing helpers are being exposed in a map block, the block gets executed in the context of a Rack::Builder (stuff you're used to see in a config.ru file)

Namespaces

You can define namespaces for routes in NYNY. Each namespace is an isolated app, which means that you can use the same api that you use in your top app there:

class Ping < NYNY::App
  get '/' do #this will be accessbile at '/ping'
    'pong'
  end
end

class App < NYNY::App
  get '/' do
    'Hello'
  end

  # You can pass a Rack app to the namespace
  namespace '/ping', Ping

  # Or you can pass a block, which will create the app for you
  namespace '/nested' do
    use SomeMiddleware
    helpers SomeHelpers

    get '/' do # this will be accessible at '/nested'
      'Hello from namespace!'
    end
  end
end

Templates

NYNY can render templates, all you need is to call the render function:

class App < NYNY::App
  get '/' do
    render 'index.erb'
  end
end

There are 2 ways to pass data to the template:

Via a instance variable:

class App < NYNY::App
  get '/' do
    @foo = 'bar' #access it as @foo in template
    render 'index.erb'
  end
end

Or via a local variable:

class App < NYNY::App
  get '/' do
    render 'index.erb', :foo => 'bar' #access it as foo in template
  end
end

To render a template with a layout, you need to render both files. It's best to create a helper for that:

class App < NYNY::App
  helpers do
    def template *args
      render 'layout.erb' do
        render *args
      end
    end
  end

  get '/' do
    template 'index.erb'
  end
end

NYNY uses Tilt for templating, so the list of supported engines is pretty complete.

Filters

Unlike Sinatra, NYNY supports only "generic" before and after filters. This means that you can't declare a filter to execute depending on a URL pattern. However, you can obtain the same effect by calling next in a before block if the request.path matches a pattern.

class App < NYNY::App
  before do
    next unless /html/ =~ request.path
    headers['Content-Type'] = 'text/html'
  end

  after do
    puts response.inspect
  end

  get '/' do
    'hello'
  end
end

Middleware

A NYNY app is a Rack middleware, which means that it can be used inside Sinatra, Rails, or any other Rack-based app:

class MyApp < Sinatra::Base
  use MyNYNYApp
end

NYNY also supports middleware itself, and that means you can use Rack middleware (or a Sinatra app) inside a NYNY app:

class App < NYNY::App
  # this will serve all the files in the "public" folder
  use Rack::Static, :url => ['public']
  use SinatraApp
end

I recommend looking at the list of Rack middlewares

Helpers

NYNY supports helpers as Sinatra does:

class App < NYNY::App
  helpers MyHelperModule
  helpers do
    def using_a_block_to_define_helper
      true
    end
  end
end

Using a helper implies that the helper module is included in the RequestScope, and that all the methods in that module will be available inside a route definition block.

Extensions

Since version 2.0.0, NYNY added support for extensions. This makes possible to include helpers, middlewares and custom app class methods inside a single module:

module MyKewlExtension
  class Middleware
    def initialize app
      @app = app
    end

    def call env
      env['KEWL'] = true
      @app.call(env) if @app
    end
  end

  module Helpers
    def the_ultimate_answer
      42
    end
  end

  def get_or_post route, &block
    get route, &block
    post route, &block
  end

  def self.registered app
    app.use Middleware
    app.helpers Helpers

    app.get_or_post '/' do
      "After many years of hard computation, the answer is #{the_ultimate_answer}"
    end
  end
end

class App < NYNY::App
  register MyKewlExtension
end

App.run!

By default, the App class will extend the provided extension module. Optionally, an extension can add a registered method, which will be invoked once the extension is registered. That method will be called with the app class as a parameter.

F. A. Q.

TBD.

Contributing

1. Fork it
2. Create your feature branch (git checkout -b my-new-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin my-new-feature)
5. Create new Pull Request