Action Controllers are the core of a web request in Rails. They are made up of one or more actions that are executed on request and then either render a template or redirect to another action. An action is defined as a public method on the controller, which will automatically be made accessible to the web-server through Rails Routes.

A sample controller could look like this:

class GuestBookController < ActionController::Base
  def index
    @entries = Entry.find(:all)
  def sign
    redirect_to :action => "index"

Actions, by default, render a template in the app/views directory corresponding to the name of the controller and action after executing code in the action. For example, the indexaction of the GuestBookController would render the template app/views/guestbook/index.erbby default after populating the @entriesinstance variable.

Unlike index, the sign action will not render a template. After performing its main purpose (creating a new entry in the guest book), it initiates a redirect instead. This redirect works by returning an external “302 Moved” HTTP response that takes the user to the index action.

The index and sign represent the two basic action archetypes used in Action Controllers. Get-and-show and do-and-redirect. Most actions are variations of these themes.


Requests are processed by the Action Controller framework by extracting the value of the “action” key in the request parameters. This value should hold the name of the action to be performed. Once the action has been identified, the remaining request parameters, the session (if one is available), and the full request with all the HTTP headers are made available to the action through instance variables. Then the action is performed.

The full request object is available with the request accessor and is primarily used to query for HTTP headers. These queries are made by accessing the environment hash, like this:

def server_ip
  location = request.env["SERVER_ADDR"]
  render :text => "This server hosted at #{location}"


All request parameters, whether they come from a GET or POST request, or from the URL, are available through the params method which returns a hash. For example, an action that was performed through /weblog/list?category=All&limit=5will include { "category" => "All", "limit" => 5 } in params.

It's also possible to construct multi-dimensional parameter hashes by specifying keys using brackets, such as:

<input type="text" name="post[name]" value="david">
<input type="text" name="post[address]" value="hyacintvej">

A request stemming from a form holding these inputs will include { "post" => { "name" => "david", "address" => "hyacintvej" } }. If the address input had been named “post[street]”, the params would have included { "post" => { "address" => { "street" => "hyacintvej" } } }. There's no limit to the depth of the nesting.


Sessions allows you to store objects in between requests. This is useful for objects that are not yet ready to be persisted, such as a Signup object constructed in a multi-paged process, or objects that don't change much and are needed all the time, such as a User object for a system that requires login. The session should not be used, however, as a cache for objects where it's likely they could be changed unknowingly. It's usually too much work to keep it all synchronized – something databases already excel at.

You can place objects in the session by using the session method, which accesses a hash:

session[:person] = Person.authenticate(user_name, password)

And retrieved again through the same hash:

Hello #{session[:person]}

For removing objects from the session, you can either assign a single key to nil:

# removes :person from session
session[:person] = nil

or you can remove the entire session with reset_session.

Sessions are stored by default in a browser cookie that's cryptographically signed, but unencrypted. This prevents the user from tampering with the session but also allows him to see its contents.

Do not put secret information in cookie-based sessions!

Other options for session storage are:

  • ActiveRecord::SessionStore - Sessions are stored in your database, which works better than PStore with multiple app servers and, unlike CookieStore, hides your session contents from the user. To use ActiveRecord::SessionStore, set

    config.action_controller.session_store = :active_record_store

    in your config/environment.rband run rake db:sessions:create .

  • MemCacheStore - Sessions are stored as entries in your memcached cache. Set the session store type in config/environment.rb:

    config.action_controller.session_store = :mem_cache_store

    This assumes that memcached has been installed and configured properly. See the MemCacheStore docs for more information.


Each action results in a response, which holds the headers and document to be sent to the user's browser. The actual response object is generated automatically through the use of renders and redirects and requires no user intervention.


Action Controller sends content to the user by using one of five rendering methods. The most versatile and common is the rendering of a template. Included in the Action Pack is the Action View, which enables rendering of ERb templates. It's automatically configured. The controller passes objects to the view by assigning instance variables:

def show
  @post = Post.find(params[:id])

Which are then automatically available to the view:

Title: <%= @post.title %>

You don't have to rely on the automated rendering. Especially actions that could result in the rendering of different templates will use the manual rendering methods:

def search
  @results = Search.find(params[:query])
  case @results
    when 0 then render :action => "no_results"
    when 1 then render :action => "show"
    when 2..10 then render :action => "show_many"

Read more about writing ERb and Builder templates in classes/ActionView/Base.html.


Redirects are used to move from one action to another. For example, after a createaction, which stores a blog entry to a database, we might like to show the user the new entry. Because we're following good DRY principles (Don't Repeat Yourself), we're going to reuse (and redirect to) a showaction that we'll assume has already been created. The code might look like this:

def create
  @entry =[:entry])
    # The entry was saved correctly, redirect to show
    redirect_to :action => 'show', :id =>
    # things didn't go so well, do something else

In this case, after saving our new entry to the database, the user is redirected to the showmethod which is then executed.

Calling multiple redirects or renders

An action may contain only a single render or a single redirect. Attempting to try to do either again will result in a DoubleRenderError:

def do_something
  redirect_to :action => "elsewhere"
  render :action => "overthere" # raises DoubleRenderError

If you need to redirect on the condition of something, then be sure to add “and return” to halt execution.

def do_something
  redirect_to(:action => "elsewhere") and return if monkeys.nil?
  render :action => "overthere" # won't be called if monkeys is nil
[RW] action_name

Returns the name of the action this controller is processing.

Class Public methods

Adds a view_path to the end of the ::view_paths array. If the current class has no view paths, copy them from the superclass. This change will be visible for all future requests.

ArticleController.append_view_path(["views/default", "views/custom"])
# File actionpack/lib/action_controller/base.rb, line 455
def append_view_path(path)
  @view_paths = superclass.view_paths.dup if @view_paths.nil?
# File actionpack/lib/action_controller/base.rb, line 382
def call(env)
  # HACK: For global rescue to have access to the original request and response
  request = env["action_controller.rescue.request"] ||=
  response = env["action_controller.rescue.response"] ||=
  process(request, response)

Converts the class name from something like “OneModule::TwoModule::NeatController” to “NeatController”.

# File actionpack/lib/action_controller/base.rb, line 395
def controller_class_name
  @controller_class_name ||= name.demodulize

Converts the class name from something like “OneModule::TwoModule::NeatController” to “neat”.

# File actionpack/lib/action_controller/base.rb, line 400
def controller_name
  @controller_name ||= controller_class_name.sub(/Controller$/, '').underscore

Converts the class name from something like “OneModule::TwoModule::NeatController” to “one_module/two_module/neat”.

# File actionpack/lib/action_controller/base.rb, line 405
def controller_path
  @controller_path ||= name.gsub(/Controller$/, '').underscore
filter_parameter_logging(*filter_words, &block)

Replace sensitive parameter data from the request log. Filters parameters that have any of the arguments as a substring. Looks in all subhashes of the param hash for keys to filter. If a block is given, each key and value of the parameter hash and all subhashes is passed to it, the value or key can be replaced using String#replace or similar method.


=> Does nothing, just slows the logging process down
filter_parameter_logging :password
=> replaces the value to all keys matching /password/ with "[FILTERED]"
filter_parameter_logging :foo, "bar"
=> replaces the value to all keys matching /foo|bar/ with "[FILTERED]"
filter_parameter_logging { |k,v| v.reverse! if k =~ /secret/ }
=> reverses the value to all keys matching /secret/
filter_parameter_logging(:foo, "bar") { |k,v| v.reverse! if k =~ /secret/ }
=> reverses the value to all keys matching /secret/, and
   replaces the value to all keys matching /foo|bar/ with "[FILTERED]"
# File actionpack/lib/action_controller/base.rb, line 483
def filter_parameter_logging(*filter_words, &block)
  parameter_filter ={ |s| s.to_s }.join('|'), true) if filter_words.length > 0
  define_method(:filter_parameters) do |unfiltered_parameters|
    filtered_parameters = {}
    unfiltered_parameters.each do |key, value|
      if key =~ parameter_filter
        filtered_parameters[key] = '[FILTERED]'
      elsif value.is_a?(Hash)
        filtered_parameters[key] = filter_parameters(value)
      elsif value.is_a?(Array)
        filtered_parameters[key] = value.collect do |item|
          case item
          when Hash, Array
      elsif block_given?
        key = key.dup
        value = value.dup if value.duplicable?
        yield key, value
        filtered_parameters[key] = value
        filtered_parameters[key] = value
  protected :filter_parameters

Return an array containing the names of public methods that have been marked hidden from the action processor. By default, all methods defined in ActionController::Base and included modules are hidden. More methods can be hidden using hide_actions.

# File actionpack/lib/action_controller/base.rb, line 412
def hidden_actions
  read_inheritable_attribute(:hidden_actions) || write_inheritable_attribute(:hidden_actions, [])

Hide each of the given methods from being callable as actions.

# File actionpack/lib/action_controller/base.rb, line 417
def hide_action(*names)
  write_inheritable_attribute(:hidden_actions, hidden_actions | { |name| name.to_s })

Adds a view_path to the front of the ::view_paths array. If the current class has no view paths, copy them from the superclass. This change will be visible for all future requests.

ArticleController.prepend_view_path(["views/default", "views/custom"])
# File actionpack/lib/action_controller/base.rb, line 443
def prepend_view_path(path)
  @view_paths = superclass.view_paths.dup if !defined?(@view_paths) || @view_paths.nil?

View load paths determine the bases from which template references can be made. So a call to render(“test/template”) will be looked up in the view load paths array and the closest match will be returned.

# File actionpack/lib/action_controller/base.rb, line 424
def view_paths
  if defined? @view_paths
# File actionpack/lib/action_controller/base.rb, line 432
def view_paths=(value)
  @view_paths = ActionView::Base.process_view_paths(value) if value
Instance Public methods

Adds a view_path to the end of the ::view_paths array. This change affects the current request only.

self.append_view_path(["views/default", "views/custom"])
# File actionpack/lib/action_controller/base.rb, line 677
def append_view_path(path)

Converts the class name from something like “OneModule::TwoModule::NeatController” to “NeatController”.

# File actionpack/lib/action_controller/base.rb, line 632
def controller_class_name

Converts the class name from something like “OneModule::TwoModule::NeatController” to “neat”.

# File actionpack/lib/action_controller/base.rb, line 637
def controller_name

Converts the class name from something like “OneModule::TwoModule::NeatController” to “one_module/two_module/neat”.

# File actionpack/lib/action_controller/base.rb, line 642
def controller_path

Adds a view_path to the front of the ::view_paths array. This change affects the current request only.

self.prepend_view_path(["views/default", "views/custom"])
# File actionpack/lib/action_controller/base.rb, line 667
def prepend_view_path(path)
# File actionpack/lib/action_controller/base.rb, line 539
def send_response
# File actionpack/lib/action_controller/base.rb, line 646
def session_enabled?
  ActiveSupport::Deprecation.warn("Sessions are now lazy loaded. So if you don't access them, consider them disabled.", caller)
url_for(options = {})

Returns a URL that has been rewritten according to the options hash and the defined routes. (For doing a complete redirect, use redirect_to).

url_foris used to:

All keys given to url_forare forwarded to the Route module, save for the following:

  • :anchor- Specifies the anchor name to be appended to the path. For example, url_for :controller => 'posts', :action => 'show', :id => 10, :anchor => 'comments' will produce “/posts/show/10#comments”.

  • :only_path- If true, returns the relative URL (omitting the protocol, host name, and port) ( falseby default).

  • :trailing_slash- If true, adds a trailing slash, as in “/archive/2005/”. Note that this is currently not recommended since it breaks caching.

  • :host- Overrides the default (current) host if provided.

  • :protocol- Overrides the default (current) protocol if provided.

  • :port- Optionally specify the port to connect to.

  • :user- Inline HTTP authentication (only plucked out if :passwordis also present).

  • :password- Inline HTTP authentication (only plucked out if :useris also present).

  • :skip_relative_url_root- If true, the url is not constructed using the relative_url_rootof the request so the path will include the web server relative installation directory.

The URL is generated from the remaining keys in the hash. A URL contains two key parts: the <base> and a query string. Routes composes a query string as the key/value pairs not included in the <base>.

The default Routes setup supports a typical Rails path of “controller/action/id” where action and id are optional, with action defaulting to 'index' when not given. Here are some typical #url_for statements and their corresponding URLs:

url_for :controller => 'posts', :action => 'recent'                # => 'proto://'
url_for :controller => 'posts', :action => 'index'                 # => 'proto://'
url_for :controller => 'posts', :action => 'index', :port=>'8033'  # => 'proto://'
url_for :controller => 'posts', :action => 'show', :id => 10       # => 'proto://'
url_for :controller => 'posts', :user => 'd', :password => '123'   # => 'proto://'

When generating a new URL, missing values may be filled in from the current request's parameters. For example, url_for :action => 'some_action' will retain the current controller, as expected. This behavior extends to other parameters, including :controller, :id, and any other parameters that are placed into a Route's path.   The URL helpers such as url_forhave a limited form of memory: when generating a new URL, they can look for missing values in the current request's parameters. Routes attempts to guess when a value should and should not be taken from the defaults. There are a few simple rules on how this is performed:

  • If the controller name begins with a slash no defaults are used:

    url_for :controller => '/home'

    In particular, a leading slash ensures no namespace is assumed. Thus, while url_for :controller => 'users'may resolve to Admin::UsersControllerif the current controller lives under that module, url_for :controller => '/users' ensures you link to ::UsersControllerno matter what.

  • If the controller changes, the action will default to index unless provided

The final rule is applied while the URL is being generated and is best illustrated by an example. Let us consider the route given by map.connect 'people/:last/:first/:action', :action => 'bio', :controller => 'people'.

Suppose that the current URL is “people/hh/david/contacts”. Let's consider a few different cases of URLs which are generated from this page.

  • url_for :action => 'bio'– During the generation of this URL, default values will be used for the first and

last components, and the action shall change. The generated URL will be, “people/hh/david/bio”.

  • url_for :first => 'davids-little-brother'This generates the URL 'people/hh/davids-little-brother' – note that this URL leaves out the assumed action of 'bio'.

However, you might ask why the action from the current request, 'contacts', isn't carried over into the new URL. The answer has to do with the order in which the parameters appear in the generated path. In a nutshell, since the value that appears in the slot for :firstis not equal to default value for :first we stop using defaults. On its own, this rule can account for much of the typical Rails URL behavior.   Although a convenience, defaults can occasionally get in your way. In some cases a default persists longer than desired. The default may be cleared by adding :name => nilto url_for's options. This is often required when writing form helpers, since the defaults in play may vary greatly depending upon where the helper is used from. The following line will redirect to PostController's default action, regardless of the page it is displayed on:

url_for :controller => 'posts', :action => nil
# File actionpack/lib/action_controller/base.rb, line 619
def url_for(options = {})
  options ||= {}
  case options
    when String
    when Hash

View load paths for controller.

# File actionpack/lib/action_controller/base.rb, line 653
def view_paths
# File actionpack/lib/action_controller/base.rb, line 657
def view_paths=(value)
  @template.view_paths = ActionView::Base.process_view_paths(value)
Instance Protected methods
default_url_options(options = nil)

Overwrite to implement a number of default options that all url_for-based methods will use. The default options should come in the form of a hash, just like the one you would use for #url_for directly. Example:

def default_url_options(options)
  { :project => ? @project.url_name : "unknown" }

As you can infer from the example, this is mostly useful for situations where you want to centralize dynamic decisions about the urls as they stem from the business domain. Please note that any individual #url_for call can always override the defaults set by this method.

# File actionpack/lib/action_controller/base.rb, line 1063
def default_url_options(options = nil)
expires_in(seconds, options = {})

Sets a HTTP 1.1 Cache-Control header. Defaults to issuing a “private” instruction, so that intermediate caches shouldn't cache the response.


expires_in 20.minutes
expires_in 3.hours, :public => true
expires in 3.hours, 'max-stale' => 5.hours, :public => true

This method will overwrite an existing Cache-Control header. See for more possibilities.

# File actionpack/lib/action_controller/base.rb, line 1218
def expires_in(seconds, options = {}) #:doc:
  cache_control = response.headers["Cache-Control"].split(",").map {|k| k.strip }
  cache_control << "max-age=#{seconds}"
  if options[:public]
    cache_control << "public"
    cache_control << "private"
  # This allows for additional headers to be passed through like 'max-stale' => 5.hours
  cache_control += options.symbolize_keys.reject{|k,v| k == :public || k == :private }.map{ |k,v| v == true ? k.to_s : "#{k.to_s}=#{v.to_s}"}
  response.headers["Cache-Control"] = cache_control.join(', ')

Sets a HTTP 1.1 Cache-Control header of “no-cache” so no caching should occur by the browser or intermediate caches (like caching proxy servers).

# File actionpack/lib/action_controller/base.rb, line 1238
def expires_now #:doc:
  response.headers["Cache-Control"] = "no-cache"

Sets the etag, last_modified, or both on the response and renders a “304 Not Modified” response if the request is already fresh.


  • :etag

  • :last_modified

  • :publicBy default the Cache-Control header is private, set this to true if you want your application to be cachable by other devices (proxy caches).


def show
  @article = Article.find(params[:id])
  fresh_when(:etag => @article, :last_modified => @article.created_at.utc, :public => true)

This will render the show template if the request isn't sending a matching etag or If-Modified-Since header and just a “304 Not Modified” response if there's a match.

# File actionpack/lib/action_controller/base.rb, line 1189
def fresh_when(options)
  options.assert_valid_keys(:etag, :last_modified, :public)
  response.etag          = options[:etag]          if options[:etag]
  response.last_modified = options[:last_modified] if options[:last_modified]
  if options[:public]
    cache_control = response.headers["Cache-Control"].split(",").map {|k| k.strip }
    cache_control << "public"
    response.headers["Cache-Control"] = cache_control.join(', ')
  if request.fresh?(response)
    head :not_modified

Return a response that has no content (merely headers). The options argument is interpreted to be a hash of header names and values. This allows you to easily return a response that consists only of significant headers:

head :created, :location => person_path(@person)

It can also be used to return exceptional conditions:

return head(:method_not_allowed) unless
return head(:bad_request) unless valid_request?
# File actionpack/lib/action_controller/base.rb, line 1005
def head(*args)
  if args.length > 2
    raise ArgumentError, "too many arguments to head"
  elsif args.empty?
    raise ArgumentError, "too few arguments to head"
  options = args.extract_options!
  status = interpret_status(args.shift || options.delete(:status) || :ok)
  options.each do |key, value|
    headers[key.to_s.dasherize.split(/-/).map { |v| v.capitalize }.join("-")] = value.to_s
  render :nothing => true, :status => status
redirect_to(options = {}, response_status = {})

Redirects the browser to the target specified in options. This parameter can take one of three forms:

  • Hash- The URL will be generated by calling #url_for with the options.

  • Record- The URL will be generated by calling #url_for with the options, which will reference a named URL for that record.

  • Stringstarting with protocol://(like http://) - Is passed straight through as the target for redirection.

  • Stringnot containing a protocol - The current protocol and host is prepended to the string.

  • :back- Back to the page that issued the request. Useful for forms that are triggered from multiple places. Short-hand for redirect_to(request.env["HTTP_REFERER"])


redirect_to :action => "show", :id => 5
redirect_to post
redirect_to ""
redirect_to "/images/screenshot.jpg"
redirect_to articles_url
redirect_to :back

The redirection happens as a “302 Moved” header unless otherwise specified.


redirect_to post_url(@post), :status => :found
redirect_to :action=>'atom', :status => :moved_permanently
redirect_to post_url(@post), :status => 301
redirect_to :action=>'atom', :status => 302

The status code can either be a standard HTTP Status code as an integer, or a symbol representing the downcased, underscored and symbolized description.

It is also possible to assign a flash message as part of the redirection. There are two special accessors for commonly used the flash names alertand noticeas well as a general purpose flashbucket.


redirect_to post_url(@post), :alert => "Watch it, mister!"
redirect_to post_url(@post), :status=> :found, :notice => "Pay attention to the road"
redirect_to post_url(@post), :status => 301, :flash => { :updated_post_id => }
redirect_to { :action=>'atom' }, :alert => "Something serious happened"

When using redirect_to :back, if there is no referrer, RedirectBackError will be raised. You may specify some fallback behavior for this case by rescuing RedirectBackError.

# File actionpack/lib/action_controller/base.rb, line 1105
def redirect_to(options = {}, response_status = {}) #:doc:
  raise"Cannot redirect to nil!") if options.nil?
  if options.is_a?(Hash) && options[:status]
    status = options.delete(:status)
  elsif response_status[:status]
    status = response_status[:status]
    status = 302
  response.redirected_to = options
  case options
    # The scheme name consist of a letter followed by any combination of
    # letters, digits, and the plus ("+"), period ("."), or hyphen ("-")
    # characters; and is terminated by a colon (":").
    when %r{^\w[\w\d+.-]*:.*}
      redirect_to_full_url(options, status)
    when String
      redirect_to_full_url(request.protocol + request.host_with_port + options, status)
    when :back
      if referer = request.headers["Referer"]
        redirect_to(referer, :status=>status)
        raise RedirectBackError
      redirect_to_full_url(url_for(options), status)
redirect_to_full_url(url, status)
# File actionpack/lib/action_controller/base.rb, line 1137
def redirect_to_full_url(url, status)
  raise DoubleRenderError if performed?"Redirected to #{url}") if logger &&
  response.redirect(url, interpret_status(status))
  @performed_redirect = true
render(options = nil, extra_options = {}, &block)

Renders the content that will be returned to the browser as the response body.

Rendering an action

Action rendering is the most common form and the type used automatically by Action Controller when nothing else is specified. By default, actions are rendered within the current layout (if one exists).

# Renders the template for the action "goal" within the current controller
render :action => "goal"
# Renders the template for the action "short_goal" within the current controller,
# but without the current active layout
render :action => "short_goal", :layout => false
# Renders the template for the action "long_goal" within the current controller,
# but with a custom layout
render :action => "long_goal", :layout => "spectacular"

Rendering partials

Partial rendering in a controller is most commonly used together with Ajax calls that only update one or a few elements on a page without reloading. Rendering of partials from the controller makes it possible to use the same partial template in both the full-page rendering (by calling it from within the template) and when sub-page updates happen (from the controller action responding to Ajax calls). By default, the current layout is not used.

# Renders the same partial with a local variable.
render :partial => "person", :locals => { :name => "david" }
# Renders the partial, making @new_person available through
# the local variable 'person'
render :partial => "person", :object => @new_person
# Renders a collection of the same partial by making each element
# of @winners available through the local variable "person" as it
# builds the complete response.
render :partial => "person", :collection => @winners
# Renders a collection of partials but with a custom local variable name
render :partial => "admin_person", :collection => @winners, :as => :person
# Renders the same collection of partials, but also renders the
# person_divider partial between each person partial.
render :partial => "person", :collection => @winners, :spacer_template => "person_divider"
# Renders a collection of partials located in a view subfolder
# outside of our current controller.  In this example we will be
# rendering app/views/shared/_note.r(html|xml)  Inside the partial
# each element of @new_notes is available as the local var "note".
render :partial => "shared/note", :collection => @new_notes
# Renders the partial with a status code of 500 (internal error).
render :partial => "broken", :status => 500

Note that the partial filename must also be a valid Ruby variable name, so e.g. 2005 and register-user are invalid.

Automatic etagging

Rendering will automatically insert the etag header on 200 OK responses. The etag is calculated using MD5 of the response body. If a request comes in that has a matching etag, the response will be changed to a 304 Not Modified and the response body will be set to an empty string. No etag header will be inserted if it's already set.

Rendering a template

Template rendering works just like action rendering except that it takes a path relative to the template root. The current layout is automatically applied.

# Renders the template located in [TEMPLATE_ROOT]/weblog/show.r(html|xml) (in Rails, app/views/weblog/show.erb)
render :template => "weblog/show"
# Renders the template with a local variable
render :template => "weblog/show", :locals => {:customer =>}

Rendering a file

File rendering works just like action rendering except that it takes a filesystem path. By default, the path is assumed to be absolute, and the current layout is not applied.

# Renders the template located at the absolute filesystem path
render :file => "/path/to/some/template.erb"
render :file => "c:/path/to/some/template.erb"
# Renders a template within the current layout, and with a 404 status code
render :file => "/path/to/some/template.erb", :layout => true, :status => 404
render :file => "c:/path/to/some/template.erb", :layout => true, :status => 404

Rendering text

Rendering of text is usually used for tests or for rendering prepared content, such as a cache. By default, text rendering is not done within the active layout.

# Renders the clear text "hello world" with status code 200
render :text => "hello world!"
# Renders the clear text "Explosion!"  with status code 500
render :text => "Explosion!", :status => 500
# Renders the clear text "Hi there!" within the current active layout (if one exists)
render :text => "Hi there!", :layout => true
# Renders the clear text "Hi there!" within the layout
# placed in "app/views/layouts/special.r(html|xml)"
render :text => "Hi there!", :layout => "special"

Streaming data and/or controlling the page generation

The :textoption can also accept a Proc object, which can be used to:

  1. stream on-the-fly generated data to the browser. Note that you should use the methods provided by ActionController::Steaming instead if you want to stream a buffer or a file.

  2. manually control the page generation. This should generally be avoided, as it violates the separation between code and content, and because almost everything that can be done with this method can also be done more cleanly using one of the other rendering methods, most notably templates.

Two arguments are passed to the proc, a responseobject and an outputobject. The response object is equivalent to the return value of the ActionController::Base#response method, and can be used to control various things in the HTTP response, such as setting the Content-Type header. The output object is an writable IO-like object, so one can call writeand flushon it.

The following example demonstrates how one can stream a large amount of on-the-fly generated data to the browser:

# Streams about 180 MB of generated data to the browser.
render :text => proc { |response, output|
  10_000_000.times do |i|
    output.write("This is line #{i}\n")

Another example:

# Renders "Hello from code!"
render :text => proc { |response, output| output.write("Hello from code!") }

Rendering XML

Rendering XML sets the content type to application/xml.

# Renders '<name>David</name>'
render :xml => {:name => "David"}.to_xml

It's not necessary to call to_xmlon the object you want to render, since renderwill automatically do that for you:

# Also renders '<name>David</name>'
render :xml => {:name => "David"}

Rendering JSON

Rendering JSON sets the content type to application/json and optionally wraps the JSON in a callback. It is expected that the response will be parsed (or eval'd) for use as a data structure.

# Renders '{"name": "David"}'
render :json => {:name => "David"}.to_json

It's not necessary to call to_jsonon the object you want to render, since renderwill automatically do that for you:

# Also renders '{"name": "David"}'
render :json => {:name => "David"}

Sometimes the result isn't handled directly by a script (such as when the request comes from a SCRIPT tag), so the :callbackoption is provided for these cases.

# Renders 'show({"name": "David"})'
render :json => {:name => "David"}.to_json, :callback => 'show'

Rendering an inline template

Rendering of an inline template works as a cross between text and action rendering where the source for the template is supplied inline, like text, but its interpreted with ERb or Builder, like action. By default, ERb is used for rendering and the current layout is not used.

# Renders "hello, hello, hello, again"
render :inline => "<%= 'hello, ' * 3 + 'again' %>"
# Renders "<p>Good seeing you!</p>" using Builder
render :inline => "xml.p { 'Good seeing you!' }", :type => :builder
# Renders "hello david"
render :inline => "<%= 'hello ' + name %>", :locals => { :name => "david" }

Rendering inline JavaScriptGenerator page updates

In addition to rendering JavaScriptGenerator page updates with Ajax in RJS templates (see ActionView::Base for details), you can also pass the :updateparameter to render, along with a block, to render page updates inline.

render :update do |page|
  page.replace_html  'user_list', :partial => 'user', :collection => @users
  page.visual_effect :highlight, 'user_list'

Rendering vanilla JavaScript

In addition to using RJS with render :update, you can also just render vanilla JavaScript with :js.

# Renders "alert('hello')" and sets the mime type to text/javascript
render :js => "alert('hello')"

Rendering with status and location headers

All renders take the :statusand :location options and turn them into headers. They can even be used together:

render :xml => post.to_xml, :status => :created, :location => post_url(post)
# File actionpack/lib/action_controller/base.rb, line 890
def render(options = nil, extra_options = {}, &block) #:doc:
  raise DoubleRenderError, "Can only render or redirect once per action" if performed?
  validate_render_arguments(options, extra_options, block_given?)
  if options.nil?
    options = { :template => default_template, :layout => true }
  elsif options == :update
    options = extra_options.merge({ :update => true })
  elsif options.is_a?(String) || options.is_a?(Symbol)
    case options.to_s.index('/')
    when 0
      extra_options[:file] = options
    when nil
      extra_options[:action] = options
      extra_options[:template] = options
    options = extra_options
  elsif !options.is_a?(Hash)
    extra_options[:partial] = options
    options = extra_options
  layout = pick_layout(options)
  response.layout = layout.path_without_format_and_extension if layout"Rendering template within #{layout.path_without_format_and_extension}") if logger && layout
  if content_type = options[:content_type]
    response.content_type = content_type.to_s
  if location = options[:location]
    response.headers["Location"] = url_for(location)
  if options.has_key?(:text)
    text = layout ? @template.render(options.merge(:text => options[:text], :layout => layout)) : options[:text]
    render_for_text(text, options[:status])
    if file = options[:file]
      render_for_file(file, options[:status], layout, options[:locals] || {})
    elsif template = options[:template]
      render_for_file(template, options[:status], layout, options[:locals] || {})
    elsif inline = options[:inline]
      render_for_text(@template.render(options.merge(:layout => layout)), options[:status])
    elsif action_name = options[:action]
      render_for_file(default_template(action_name.to_s), options[:status], layout)
    elsif xml = options[:xml]
      response.content_type ||= Mime::XML
      render_for_text(xml.respond_to?(:to_xml) ? xml.to_xml : xml, options[:status])
    elsif js = options[:js]
      response.content_type ||= Mime::JS
      render_for_text(js, options[:status])
    elsif options.include?(:json)
      json = options[:json]
      json = ActiveSupport::JSON.encode(json) unless json.is_a?(String)
      json = "#{options[:callback]}(#{json})" unless options[:callback].blank?
      response.content_type ||= Mime::JSON
      render_for_text(json, options[:status])
    elsif options[:partial]
      options[:partial] = default_template_name if options[:partial] == true
      if layout
        render_for_text(@template.render(:text => @template.render(options), :layout => layout), options[:status])
        render_for_text(@template.render(options), options[:status])
    elsif options[:update]
      generator =, &block)
      response.content_type = Mime::JS
      render_for_text(generator.to_s, options[:status])
    elsif options[:nothing]
      render_for_text(nil, options[:status])
      render_for_file(default_template, options[:status], layout)
render_to_string(options = nil, &block)

Renders according to the same rules as render, but returns the result in a string instead of sending it as the response body to the browser.

# File actionpack/lib/action_controller/base.rb, line 985
def render_to_string(options = nil, &block) #:doc:
  render(options, &block)
  response.content_type = nil

Resets the session by clearing out all the objects stored within and initializing a new session object.

# File actionpack/lib/action_controller/base.rb, line 1243
def reset_session #:doc:
  @_session = request.session

Sets the etag and/or last_modified on the response and checks it against the client request. If the request doesn't match the options provided, the request is considered stale and should be generated from scratch. Otherwise, it's fresh and we don't need to generate anything and a reply of “304 Not Modified” is sent.


  • :etag

  • :last_modified

  • :publicBy default the Cache-Control header is private, set this to true if you want your application to be cachable by other devices (proxy caches).


def show
  @article = Article.find(params[:id])
  if stale?(:etag => @article, :last_modified => @article.created_at.utc)
    @statistics = @article.really_expensive_call
    respond_to do |format|
      # all the supported formats
# File actionpack/lib/action_controller/base.rb, line 1166
def stale?(options)