Monkey Testing for libraries
When there is a need to test code with random values
Simply put is instead of using pre-determened variables inside test and creating stub data values, you define a sort of joker that can generate pseudo random values and use them through running the test multiple times you have a higher chance to test most of your edge cases.
npm install lib-monkey
There are multiple jokers and runners to use for the library but the basic idea is to send the joker to one of the runners to know where to put the random value
const monkey = require('lib-monkey');
describe('test', () => {
it('test sync or prommise', () =>
monkey.it(() => {
const a = monkey.generate(monkey.int.min(0).max(10));
// Do some tests with a as ranom value
// Will accept sync or prommise based results
})
.do(n_times, concurrency));
it('test callback async', () =>
monkey.it((done) => {
const a = monkey.generate(monkey.int.min(0).max(10));
// Do some tests with a as ranom value
// Call done after async tests
})
.do(n_times, concurrency));
});
There are multipe jokers in library and they have multiple modifyers.
There is the base modifyer nullable(boolean|chance|null == true)
and all the other modifyers are based of of Chance library
monkey.int.min(0).max(10) => chance.integer({ min:0, max: 10 })
monkey.bool.likelihood(70) => chance.bool({ likelihood: 70 })
Currently there are the following jokers:
- bool, char, float, int, letter, natural, string
- date, hammertime, timestamp, timezone
- paragraph, sentence, syllable, word
- pick (from array), array (of jokers), map, object (deep map), typed (of specific type)
It's basicaly every it function in every test framework, except that it's nameless and supports contexts
monkey.it(fn, context)
where fn will have it'sthis
equal context.
And this it will forward and enritch any assertion error that
monkey.it(async () => {
const foo = monkey.generate(monkey.bool); // foo will be 50/50 true false pretty much
// Do Some tests and
})
.do(5)
.catch(err => console.error(err) || process.exit(-1));
To easly re-run the test localy with specific seed or just to test something the library comes with a cli
npm install -g lib-monkey
lib-monkey -s [seed] [CMD]
lib-monkey -j [joker description]
# example
lib-monkey -s 16012018 -j int.min(0).max(10) ===> 7
This project should co-exist with all the avialbe test runners like mocha or jasmin and be used as a tool for random value testing
- Add more Jokers and Runners
-
Add self assertion api - Finish the docs
- Create code coverage for the library
- Maybe change the api to be dot chainable