Ripple | Core

A simple extensible in-memory data structure of resources.

var ripple = core()

ripple(name, body) // setter
ripple(name)       // getter

You can also use the method-chained API:

ripple                
  .resource(name, body)
  .resource(name, body)
  ...

The resources it registers are accesible under ripple.resources.

A canonical resource is an object with the following shape and three properties:

{ name: 'foo'
, body: 'bar'
, headers: { 'content-type': 'text/plain' }
}

That is, it can be uniquely identified (name), the resource itself (body) and some arbitrary metadata (headers). Core only deals with the content-type header, however other modules may add and interpret their own per-resource metadata.

Core only comes with one type (text/plain) out of the box, so will fail to register anything other than a string. This is to make it very extensible and future-proof, such that you could for example create other exotic types like application/jsx, text/jade or data/immutable.

Note that you do not have to register a canonical resource, you will most likely use shortcuts in your application code (see API for more).

ripple('foo', 'bar')

// will result in ripple.resources ===
{
  foo: { name: 'foo'
       , body: 'bar'
       , headers: { 'content-type': 'text/plain }
       }
}

Resource Interpretation

When an content-type header is not explicitly given, core will loop through it's registered types and see if any of them understand this particular resource (by passing { name, body, headers } to each type check function). If any of them do:

• The content-type header will be set
• The type parse function will be run on the resource
• The resource will be stored internally
• A change event will be emitted

You'll need to extend ripple.types to tell it how to interpret other resources. Each type object should have the following:

{ header: 'the content type you are registering'
, check: function // necessary
, parse: function // optional
}

The parse function is a chance to initialise the resource, set default headers, etc. Some examples:

• application/data - proxies change events so you can do per-resource change listeners ripple('data').on('change, fn)
• application/javascript - turns a function as a string into a real function (useful since streamed over WS).

Other modules can also extend existing parse functions. For example, sync extends every type parse function to add the ability to define server/client transformation functions.

See other existing vanilla types for more examples: Data, Versioned Data, Functions, HTML, CSS.

Event

The core instance is emitterified. Whenever a resource is registered, a change event will be emitted.

ripple.on('change', doSomething)

API

ripple('name')                   // - returns the resource body if it exists
ripple('name', body)             // - creates & returns resource, with specified name and body
ripple('name', body, headers })  // - creates & returns resource, with specified name, body and headers
ripple({ name, body, headers })  // - creates & returns resource, with specified name, body and headers
ripple([ ... ])                  // - calls ripple on each item - registers an array of resources
ripple.resources                 // - returns raw resources
ripple.resource                  // - alias for ripple, returns ripple instead of resource for method chaining
ripple.register                  // - alias for ripple
ripple.on                        // - event listener for changes - all resources