Mutually exclusive type testing
Want to see pretty graphs? Log in now!
npm install typr
|7||downloads in the last day|
|34||downloads in the last week|
|99||downloads in the last month|
|Last Published By|
|Version||0.2.0 last updated 2 years ago|
The aim of this module is to completely partition the set of JS objects into a set of types T:
T = [Function, Object, Date, Number, String, Boolean, RegExp, Undefined, Arguments, Null, Array]
Prop. 1 For every JS object x, x has exactly one type in T where [typr] returns true for its typecheck.
Prop. 2 Whether or not an Object has keys and can be looped over is determined by
t.hasKeys and is completely independent of the type found in prop 1.
Prop. 3 Number types can be partitioned into 3 sub-types:
- isNumeric i.e. !isNaN && !isInfinite
These properties are tested extensively.
var typr = require('typr'); typr.isFunction(el); // true iff el is a function
Read the very short API.
The reason for prop 2 is that Objects can masquerade othes instances like Functions, Strings, RegExps, Dates. Object.keys() (and for-in) will actually work on some of these. This does not mean they will have interesting keys, but they sometimes do!
A common trick in node modules to export a function, but additionally attach properties to it. Such an object will only be a Function by our partitioning (as that was its original construction), but it is Object-like.
If you would rather test for enumerability of properties, use the extra test for Object-likeness:
Note again that this may not be super intuitive:
var dualThing = new String("wee"); dualThing.prop = "hi"; Object.keys(dualThing); // [ '0', '1', '2', 'prop' ] t.isObject(dualThing); // false t.isString(dualThing); // true t.hasKeys(dualThing); // true
$ npm install typr
Install development dependencies
$ npm install
Run the tests
$ npm test
MIT-Licensed. See LICENSE file for details.