Dynatype

A dynamic datatype is a meta construct that abstracts the data represented by a datatype. It is possible to define datatypes and define relations between those.

Usage

The API is separated into descriptors and translators.

Descriptor

The descriptor API is usable via

var descriptor = require('dynatype').descriptor

It offers the following functions

| function | description | |----------|-------------| | create(name, data) | Create a new dynatype descriptor. It always must have a unique name. The data object contains information about the structure of the type. It is a JS object. | | typename(dynatype) | Get the name of a dynatype descriptor instance | | hasProperty(dynatype, property) | Returns whether the dynatype defines the property property | | datatype(dynatype, property) | Returns the datatype of the dynatype property. |

Data Format

The data object for a descriptor has a special structure. It is a mapping between the name of the property and its type. They type can be one of the following

• "number"
• "bool"
• "string"
• "array" this requires the extra field arraytype for the array datatype. We don't support mixed arrays.
• Other descriptors. Recursive definitions are not supported.

A descriptor could look like this:

var pointXY = descriptor.create('PointXY', {
  'x': 'number',
  'y': 'number'
})

Translator

A translator defines a method of translating one datatype into another. Consider this example:

var pointXY = /*...*/
var pointRPhi = descriptor.create('PointRPhi', {
  'r': 'number',
  'phi': 'number'
})
var translator = translator.create(pointXY, pointRPhi, {/*...*/})

this creates a translator that defines some way of translating a point given in XY coordinates into a point given in polar coordinates. The data that describes the translation is specific to the system it is used in. The dynatype libraries do only store this information but never touch it.