should is an expressive, readable, test framework agnostic, assertion library for node.
It extends the Object prototype with a single non-enumerable getter that allows you to express how that object should behave.
var should = require('should');
var user = {
name: 'tj'
, pets: ['tobi', 'loki', 'jane', 'bandit']
};
user.should.have.property('name', 'tj');
user.should.have.property('pets').with.lengthOf(4);
// if the object was created with Object.create(null)
// then it doesn't inherit `Object` and have the `should` getter
// so you can do:
should(user).have.property('name', 'tj');
should(true).ok;
someAsyncTask(foo, function(err, result){
should.not.exist(err);
should.exist(result);
result.bar.should.equal(foo);
});
$ npm install should --save-dev
If you want to use should in browser, use version that is in root of repository. It is build with browserify (see Makefile about how it is build). To build fresh version:
# you should have browserify
npm install -g browserify
make browser
Every assertion will return a should.js-wraped Object, so assertions can be chained. For example:
user.should.be.an.instanceOf(Object).and.have.property('name', 'tj');
user.pets.should.be.instanceof(Array).and.have.lengthOf(4);
All assertions return themselves but not property
and ownProperty
. They two would return the property value.
for example if a property is volatile we can first assert property existence:
// `user.should.have.property('pets')` returned `should(user.pets)` but not `should(user)`
user.should.have.property('pets').with.lengthOf(4)
which is essentially equivalent to below, however the property may not exist:
user.pets.should.have.lengthOf(4)
our dummy getters such as a, an, be, of and have make tests more readable while the getters and and with helps express chaining:
user.should.be.an.instanceOf(Object).and.have.property('name', 'tj')
user.should.have.property('pets').with.a.lengthOf(4)
Those getters don't have any actual effect, just for readability.
For some rare cases should can be used statically, without Object.prototype
.
It can be replacement for node assert module (and it uses the same AssertionError
):
assert.fail(actual, expected, message, operator) // just write wrong should assertion
assert(value, message), assert.ok(value, [message]) // should(value).ok
assert.equal(actual, expected, [message]) // should(actual).eql(expected, [message])
assert.notEqual(actual, expected, [message]) // should(actual).not.eql(expected, [message])
assert.deepEqual(actual, expected, [message]) // should(actual).eql(expected, [message])
assert.notDeepEqual(actual, expected, [message]) // should(actual).not.eql(expected, [message])
assert.strictEqual(actual, expected, [message]) // should(actual).equal(expected, [message])
assert.notStrictEqual(actual, expected, [message]) // should(actual).not.equal(expected, [message])
assert.throws(block, [error], [message]) // should(block).throw([error])
assert.doesNotThrow(block, [message]) // should(block).not.throw([error])
assert.ifError(value) // should(value).Error (to check if it is error) or should(value).not.ok (to check that it is falsy)
Assert truthfulness:
true.should.be.ok
'yay'.should.be.ok
(1).should.be.ok
or negated:
false.should.not.be.ok
''.should.not.be.ok
(0).should.not.be.ok
Assert === true:
true.should.be.true
'1'.should.not.be.true
Assert === false:
false.should.be.false
(0).should.not.be.false
Assert Arguments
:
var args = (function(){ return arguments; })(1,2,3);
args.should.be.arguments;
[].should.not.be.arguments;
Asserts that given object is empty
[].should.be.empty
''.should.be.empty
({}).should.be.empty
(function() {
arguments.should.be.empty;
})()
equality:
({ foo: 'bar' }).should.eql({ foo: 'bar' })
[1,2,3].should.eql([1,2,3])
strict equality:
should.strictEqual(undefined, value)
should.strictEqual(false, value)
(4).should.equal(4)
'test'.should.equal('test')
[1,2,3].should.not.equal([1,2,3])
(4).should.be.exactly(4)
Assert inclusive numeric range:
user.age.should.be.within(5, 50)
Assert floating point number:
(99.99).should.be.approximately(100, 0.1);
Assert typeof:
user.should.be.type('object')
'test'.should.be.type('string')
Assert instanceof or instanceOf:
user.should.be.an.instanceof(User)
[].should.be.an.instanceOf(Array)
Assert numeric value above the given value:
user.age.should.be.above(5)
user.age.should.not.be.above(100)
Assert numeric value below the given value:
user.age.should.be.below(100)
user.age.should.not.be.below(5)
Assert numeric valus is NaN:
(undefined + 0).should.be.NaN;
Assert numeric valus is Infinity:
(1/0).should.be.Infinity;
Assert regexp match:
username.should.match(/^\w+$/)
Assert length property exists and has a value of the given number:
user.pets.should.have.length(5)
user.pets.should.have.a.lengthOf(5)
({ length: 10}).should.have.length(10);
Aliases: lengthOf
Assert property exists and has optional value(compare using ===
):
user.should.have.property('name')
user.should.have.property('age', 15)
user.should.not.have.property('rawr')
user.should.not.have.property('age', 0)
Assert given properties exists:
user.should.have.properties('name', 'age');
user.should.have.properties(['name', 'age']);
Assert own property (on the immediate object):
({ foo: 'bar' }).should.have.ownProperty('foo')
Asserts that .statusCode
is code
:
res.should.have.status(200);
Asserts that a .headers
object with field
and optional value
are present:
res.should.have.header('content-length');
res.should.have.header('Content-Length', '123');
res.should.have.header('content-length', '123');
Assert that Content-Type is "application/json; charset=utf-8"
res.should.be.json
Assert that Content-Type is "text/html; charset=utf-8"
res.should.be.html
Assert that the given obj
is present via indexOf()
, so this works for strings, arrays, or custom objects implementing indexOf. Also it can assert if given object will have some sub-object.
Assert array value:
[1,2,3].should.include(3)
[1,2,3].should.include(2)
[1,2,3].should.not.include(4)
Assert substring:
'foo bar baz'.should.include('foo')
'foo bar baz'.should.include('bar')
'foo bar baz'.should.include('baz')
'foo bar baz'.should.not.include('FOO')
Assert object includes another object:
var tobi = { name: 'Tobi', age: 1 };
var jane = { name: 'Jane', age: 5 };
var user = { name: 'TJ', pet: tobi };
user.should.include({ pet: tobi });
user.should.include({ pet: tobi, name: 'TJ' });
user.should.not.include({ pet: jane });
user.should.not.include({ name: 'Someone' });
Assert that an object equal to the given obj
is present in an Array:
[[1],[2],[3]].should.includeEql([3])
[[1],[2],[3]].should.includeEql([2])
[[1],[2],[3]].should.not.includeEql([4])
Assert an exception is thrown:
(function(){
throw new Error('fail');
}).should.throw();
Assert an exception is not thrown:
(function(){
}).should.not.throw();
Assert exception message matches string:
(function(){
throw new Error('fail');
}).should.throw('fail');
Assert exepection message matches regexp:
(function(){
throw new Error('failed to foo');
}).should.throw(/^fail/);
An alias of throw
, its purpose is to be an option for those who run
jshint in strict mode.
(function(){
throw new Error('failed to baz');
}).should.throwError(/^fail.*/);
Assert that string starts with str
.
'foobar'.should.startWith('foo')
'foobar'.should.not.startWith('bar')
Assert that string ends with str
.
'foobar'.should.endWith('bar')
'foobar'.should.not.endWith('foo')
Assert own object keys, which must match exactly, and will fail if you omit a key or two:
var obj = { foo: 'bar', baz: 'raz' };
obj.should.have.keys('foo', 'baz');
obj.should.have.keys(['foo', 'baz']);
({}).should.be.an.Object;
(1).should.be.an.Number;
[].should.be.an.Array;
(true).should.be.a.Boolean;
''.should.be.a.String;
As it can often be difficult to ascertain exactly where failed assertions are coming from in your tests, an optional description parameter can be passed to several should matchers. The description will follow the failed assertion in the error:
(1).should.eql(0, 'some useful description')
AssertionError: expected 1 to equal 0 | some useful description
at Object.eql (/Users/swift/code/should.js/node_modules/should/lib/should.js:280:10)
...
The methods that support this optional description are: eql
, equal
, within
, a
, instanceof
, above
, below
, match
, length
, property
, ownProperty
, include
, and includeEql
.
For example you can use should with the Mocha test framework by simply including it:
var should = require('should');
var mylib = require('mylib');
describe('mylib', function () {
it('should have a version with the format #.#.#', function() {
lib.version.should.match(/^\d+\.\d+\.\d+$/);
}
});
To run the tests for should simply run:
$ make test
Yes, yes it does, with a single getter should, and no it won't break your code, because it does this properly with a non-enumerable property.
(The MIT License)
Copyright (c) 2010-2011 TJ Holowaychuk <[email protected]>
Copyright (c) 2011 Aseem Kishore <[email protected]>
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the 'Software'), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.