REST

REST = REpresentational State Transfer. The concept was developed by Roy Fielding in his Ph.D. dissertation. REST describes a set of architectural principles for designing distributed systems.

REST principles could be applied to any distributed system, but in practice it is used most often for web applications and services, where clients and servers communicate using the HTTP protocol.

The basic idea is pretty simple:

• There are resources. Each resource has a unique name. In the context of the web (and the HTTP protocol), each resource is named by a URL. The URLs specify the server and "path" of the resource. The server part of the URL specifies which server to contact to access the resource. The path part of the URL names the resource to distinguish it from other resources on the same server.
• There are representations of resources. For any resource, a web application might present it in a variety of formats. For example, the same resource could be represented using HTML and CSS for presentation in a web browser, and also by XML or JSON for use by programs.
• There are verbs which clients can use to create, access, and modify resources. In the context of the web, the verbs are the HTTP request methods. The most common HTTP methods are GET, PUT, POST, DELETE. Each HTTP method

We'll use a concrete example: a wholesale fruit business, called ycpfruit.com.

Naming resources

One important kind of resource is the kinds of fruit that are available. For example, the business might make the list of the kinds of fruit it sells available at the URL

http://ycpfruit.com/fruit

More information about specific kinds of fruit could be made available through URLs like

http://ycpfruit.com/fruit/Apples

http://ycpfruit.com/fruit/Oranges

It is common to organize resource URLs in a directory-like scheme: for example, we can think of the path /fruit as naming fruit in general, and more specific paths such as /fruit/Apples as naming specific kinds of fruit.

Verbs

In a RESTful web service, the HTTP actions (a.k.a. methods) are used as verbs to create, access, and modify resources.

The GET verb accesses a resource: in other words, it requests that the web service return a representation of the named resource.

The POST verb creates a resource. For example, a POST request to the path /fruit, where the body of the request contains a representation of a fruit resource (perhaps in XML or JSON format), would create a new fruit resource (which could then be accessed by subsequent requests).

The PUT verb creates a new resource or replaces an existing resource. For example, a PUT request to the path /fruit/Apples would replace the Apples resource with the one contained in the request body (assuming that a resource named Apples existed previously).

The DELETE verb deletes the resource named by a resource. For example, a DELETE request specifying the path /fruit/Kumquats would delete Kumquats as a fruit, and no further requests to access that resource would succeed.

TODO: persistence layer (implementation using in-memory data structures)