This guide focuses on the interaction between Controller and View in the Model-View-Controller triangle. As you know, the Controller is responsible for orchestrating the whole process of handling a request in Rails, though it normally hands off any heavy code to the Model. But then, when it's time to send a response back to the user, the Controller hands things off to the View. It's that handoff that is the subject of this guide.

In broad strokes, this involves deciding what should be sent as the response and calling an appropriate method to create that response. If the response is a full-blown view, Rails also does some extra work to wrap the view in a layout and possibly to pull in partial views. You'll see all of those paths later in this guide.

From the controller's point of view, there are three ways to create an HTTP response:

I'll cover each of these methods in turn. But first, a few words about the very easiest thing that the controller can do to create a response: nothing at all.

You've heard that Rails promotes "convention over configuration." Default rendering is an excellent example of this. By default, controllers in Rails automatically render views with names that correspond to actions. For example, if you have this code in your BooksController class:

Rails will automatically render app/views/books/show.html.erb after running the method. In fact, if you have the default catch-all route in place (map.connect :controller/:action/:id), Rails will even render views that don't have any code at all in the controller. For example, if you have the default route in place and a request comes in for /books/sale_list, Rails will render app/views/books/sale_list.html.erb in response.

In most cases, the ActionController::Base#render method does the heavy lifting of rendering your application's content for use by a browser. There are a variety of ways to customize the behavior of render. You can render the default view for a Rails template, or a specific template, or a file, or inline code, or nothing at all. You can render text, JSON, or XML. You can specify the content type or HTTP status of the rendered response as well.

Perhaps the simplest thing you can do with render is to render nothing at all:

This will send an empty response to the browser (though it will include any status headers you set with the :status option, discussed below).

If you want to render the view that corresponds to a different action within the same template, you can use render with the :action option:

If the call to update_attributes_ fails, calling the +update action in this controller will render the edit.html.erb template belonging to the same controller.

What if you want to render a template from an entirely different controller from the one that contains the action code? You can do that with the :template option to render, which accepts the full path (relative to app/views) of the template to render. For example, if you're running code in an AdminProductsController that lives in app/controllers/admin, you can render the results of an action to a template in app/views/products this way:

If you want to use a view that's entirely outside of your application (perhaps you're sharing views between two Rails applications), you can use the :file option to render:

The :file option takes an absolute file-system path. Of course, you need to have rights to the view that you're using to render the content.

The render method can do without a view completely, if you're willing to use the :inline option to supply ERB as part of the method call. This is perfectly valid:

By default, inline rendering uses ERb. You can force it to use Builder instead with the :type option:

You can also render javascript-based page updates inline using the :update option to render:

You can send plain text - with no markup at all - back to the browser by using the :text option to render:

JSON is a javascript data format used by many AJAX libraries. Rails has built-in support for converting objects to JSON and rendering that JSON back to the browser:

Rails also has built-in support for converting objects to XML and rendering that XML back to the caller:

Rails can render vanilla JavaScript (as an alternative to using update with n .rjs file):

This will send the supplied string to the browser with a MIME type of text/javascript.

Calls to the render method generally accept four options:

By default, Rails will serve the results of a rendering operation with the MIME content-type of text/html (or application/json if you use the :json option, or application/xml for the :xml option.). There are times when you might like to change this, and you can do so by setting the :content_type option:

With most of the options to render, the rendered content is displayed as part of the current layout. You'll learn more about layouts and how to use them later in this guide.

You can use the :layout option to tell Rails to use a specific file as the layout for the current action:

You can also tell Rails to render with no layout at all:

Rails will automatically generate a response with the correct HTML status code (in most cases, this is 200 OK). You can use the :status option to change this:

Rails understands either numeric status codes or symbols for status codes. You can find its list of status codes in actionpack/lib/action_controller/status_codes.rb. You can also see there how it maps symbols to status codes in that file.

You can use the :location option to set the HTTP Location header:

To find the current layout, Rails first looks for a file in app/views/layouts with the same base name as the controller. For example, rendering actions from the PhotosController class will use /app/views/layouts/photos.html.erb. If there is no such controller-specific layout, Rails will use /app/views/layouts/application.html.erb. If there is no .erb layout, Rails will use a .builder layout if one exists. Rails also provides several ways to more precisely assign specific layouts to individual controllers and actions.

You can override the automatic layout conventions in your controllers by using the layout declaration in the controller. For example:

With this declaration, all methods within ProductsController will use app/views/layouts/inventory.html.erb for their layout.

To assign a specific layout for the entire application, use a declaration in your ApplicationController class:

With this declaration, all views in the entire application will use app/views/layouts/main.html.erb for their layout.

You can use a symbol to defer the choice of layout until a request is processed:

Now, if the current user is a special user, they'll get a special layout when viewing a product. You can even use an inline method to determine the layout:

Layouts specified at the controller level support :only and :except options that take either a method name or an array of method names:

With those declarations, the inventory layout would be used only for the index method, the product layout would be used for everything else except the rss method, and the rss method will have its layout determined by the automatic layout rules.

Layouts are shared downwards in the hierarchy, and more specific layouts always override more general ones. For example:

application.rb:

posts_controller.rb:

special_posts_controller.rb:

old_posts_controller.rb:

In this application:

Sooner or later, most Rails developers will see the error message "Can only render or redirect once per action". While this is annoying, it's relatively easy to fix. Usually it happens because of a fundamental misunderstanding of the way that render works.

For example, here's some code that will trigger this error:

If @book.special? evaluates to true, Rails will start the rendering process to dump the @book variable into the special_show view. But this will not stop the rest of the code in the show action from running, and when Rails hits the end of the action, it will start to render the show view - and throw an error. The solution is simple: make sure that you only have one call to render or redirect in a single code path. One thing that can help is and return. Here's a patched version of the method:

Another way to handle returning responses to a HTTP request is with redirect_to. As you've seen, render tells Rails which view (or other asset) to use in constructing a response. The redirect_to method does something completely different: it tells the browser to send a new request for a different URL. For example, you could redirect from wherever you are in your code to the index of photos in your application with this call:

You can use redirect_to with any arguments that you could use with link_to or url_for. In addition, there's a special redirect that sends the user back to the page they just came from:

Rails uses HTTP status code 302 (permanent redirect) when you call redirect_to. If you'd like to use a different status code (perhaps 301, temporary redirect), you can do so by using the :status option:

Just like the :status option for render, :status for redirect_to accepts both numeric and symbolic header designations.

Sometimes inexperienced developers conceive of redirect_to as a sort of goto command, moving execution from one place to another in your Rails code. This is not correct. Your code stops running and waits for a new request for the browser. It just happens that you've told the browser what request it should make next, by sending back a HTTP 302 status code.

Consider these actions to see the difference:

With the code in this form, there will be likely be a problem if the @book variable is nil. Remember, a render :action doesn't run any code in the target action, so nothing will set up the @books variable that the index view is presumably depending on. One way to fix this is to redirect instead of rendering:

With this code, the browser will make a fresh request for the index page, the code in the index method will run, and all will be well.

The head method exists to let you send back responses to the browser that have only headers. It provides a more obvious alternative to calling render :nothing. The head method takes one response, which is interpreted as a hash of header names and values. For example, you can return only an error header:

Or you can use other HTTP headers to convey additional information:

When Rails renders a view as a response, it does so by combining the view with the current layout (using the rules for finding the current layout that were covered earlier in this guide). Within a layout, you have access to three tools for combining different bits of output to form the overall response:

I'll discuss each of these in turn.

Asset tags provide methods for generating HTML that links views to assets like images, javascript, stylesheets, and feeds. There are four types of include tag:

You can use these tags in layouts or other views, although the tags other than image_tag are most commonly used in the <head> section of a layout.

The auto_discovery_link_tag helper builds HTML that most browsers and newsreaders can use to detect the presences of RSS or ATOM feeds. It takes the type of the link (:rss+ or :atom), a hash of options that are passed through to url_for, and a hash of options for the tag:

There are three tag options available for auto_discovery_link_tag:

The javascript_include_tag helper returns an HTML <script> tag for each source provided. Rails looks in public/javascripts for these files by default, but you can specify a full path relative to the document root, or a URL, if you prefer. For example, to include public/javascripts/main.js:

To include public/javascripts/main.js and public/javascripts/columns.js:

To include public/javascripts/main.js and public/photos/columns.js:

To include http://example.com/main.js:

The defaults option loads the Prototype and Scriptaculous libraries:

The all option loads every javascript file in public/javascripts, starting with the Prototype and Scriptaculous libraries:

You can supply the :recursive option to load files in subfolders of public/javascripts as well:

If you're loading multiple javascript files, you can create a better user experience by combining multiple files into a single download. To make this happen in production, specify :cache ⇒ true in your javascript_include_tag:

By default, the combined file will be delivered as javascripts/all.js. You can specify a location for the cached asset file instead:

The stylesheet_link_tag helper returns an HTML <link> tag for each source provided. Rails looks in public/stylesheets for these files by default, but you can specify a full path relative to the document root, or a URL, if you prefer. For example, to include public/stylesheets/main.cs:

To include public/stylesheets/main.css and public/stylesheets/columns.css:

To include public/stylesheets/main.css and public/photos/columns.css:

To include http://example.com/main.cs:

By default, stylesheet_link_tag creates links with media="screen" rel="stylesheet" type="text/css". You can override any of these defaults by specifying an appropriate option (:media, :rel, or :type):

The all option links every CSS file in public/stylesheets:

You can supply the :recursive option to link files in subfolders of public/stylesheets as well:

If you're loading multiple CSS files, you can create a better user experience by combining multiple files into a single download. To make this happen in production, specify :cache ⇒ true in your stylesheet_link_tag:

By default, the combined file will be delivered as stylesheets/all.css. You can specify a location for the cached asset file instead:

The image_tag helper builds an HTML <image> tag to the specified file. By default, files are loaded from public/images. If you don't specify an extension, .png is assumed by default:

You can supply a path to the image if you like:

You can supply a hash of additional HTML options:

There are also three special options you can use with image_tag:

Within the context of a layout, yield identifies a section where content from the view should be inserted. The simplest way to use this is to have a single yield, into which the entire contents of the view currently being rendered is inserted:

You can also create a layout with multiple yielding regions:

The main body of the view will always render into the unnamed yield. To render content into a named yield, you use the content_for method.

The content_for method allows you to insert content into a yield block in your layout. You only use content_for to insert content in named yields. For example, this view would work with the layout that you just saw:

The result of rendering this page into the supplied layout would be this HTML:

The content_for method is very helpful when your layout contains distinct regions such as sidebars and footers that should get their own blocks of content inserted. It's also useful for inserting tags that load page-specific javascript or css files into the header of an otherwise-generic layout.

Partial templates - usually just called "partials" - are another device for breaking apart the rendering process into more manageable chunks. With a partial, you can move the code for rendering a particular piece of a response to its own file.

To render a partial as part of a view, you use the render method within the view, and include the :partial option:

This will render a file named _menu.html.erb at that point within the view being rendered. Note the leading underscore character: partials are named with a leading underscore to distinguish them from regular views, even though they are referred to without the underscore. This holds true even when you're pulling in a partial from another folder:

That code will pull in the partial from app/views/shared/_menu.html.erb.

One way to use partials is to treat them as the equivalent of subroutines: as a way to move details out of a view so that you can grasp what's going on more easily. For example, you might have a view that looked like this:

Here, the _ad_banner.html.erb and _footer.html.erb partials could contain content that is shared among many pages in your application. You don't need to see the details of these sections when you're concentrating on a particular page.

A partial can use its own layout file, just as a view can use a layout. For example, you might call a partial like this:

This would look for a partial named _link_area.html.erb and render it using the layout _graybar.html.erb. Note that layouts for partials follow the same leading-underscore naming as regular partials, and are placed in the same folder with the partial that they belong to (not in the master layouts folder).

You can also pass local variables into partials, making them even more powerful and flexible. For example, you can use this technique to reduce duplication between new and edit pages, while still keeping a bit of distinct content:

new.html.erb:

edit.html.erb:

_form.html.erb:

Although the same partial will be rendered into both views, the label on the submit button is controlled by a local variable passed into the partial.

Every partial also has a local variable with the same name as the partial (minus the underscore). You can pass an object in to this local variable via the :object option:

Within the customer partial, the @customer variable will refer to @new_customer from the parent view.

If you have an instance of a model to render into a partial, you can use a shorthand syntax:

Assuming that the @customer instance variable contains an instance of the Customer model, this will use _customer.html.erb to render it.

Partials are very useful in rendering collections. When you pass a collection to a partial via the :collection option, the partial will be inserted once for each member in the collection:

index.html.erb:

_product.html.erb:

When a partial is called with a pluralized collection, then the individual instances of the partial have access to the member of the collection being rendered via a variable named after the partial. In this case, the partial is _product, and within the +_product partial, you can refer to product to get the instance that is being rendered. To use a custom local variable name within the partial, specify the :as option in the call to the partial:

With this change, you can access an instance of the @products collection as the item local variable within the partial.

You can also specify a second partial to be rendered between instances of the main partial by using the :spacer_template option:

Rails will render the _product_ruler partial (with no data passed in to it) between each pair of _product partials.

There's also a shorthand syntax available for rendering collections. For example, if @products is a collection of products, you can render the collection this way:

index.html.erb:

_product.html.erb:

Rails determines the name of the partial to use by looking at the model name in the collection. In fact, you can even create a heterogeneous collection and render it this way, and Rails will choose the proper partial for each member of the collection:

index.html.erb:

_customer.html.erb:

_employee.html.erb:

In this case, Rails will use the customer or employee partials as appropriate for each member of the collection.

Lighthouse ticket