Random data generator for when test data is insignificant. Tailoring data too closely for the domain can end up being confusing in tests because it distracts from the behavior that is actually important in the test. This is why I prefer to make it very clear when test data is insignificant by using a data generator. This library captures most of the patterns I use frequently.
To save myself from needing to maintain the actual data generators, this library leverages Chance.js for generating the data. So why not just use Chance.js directly?
- I find it annoying that an instance of Chance has to be created. Rather than initialize an instance in every module, I wrap a single instance for reuse throughout the tests.
- I can set my default options to align with my typical conventions rather than having to repeatedly configure Chance.js's very flexible API
- Lots of my tests need objects to be generated, but the structure of those objects typically does not matter. Chance.js does not have a simple-object generator, so I've included that on my own.
- Many of my tests need lists generated. Like objects, Chance.js does not generate lists, so I've included this on my own.
$ npm install @travi/any --save-dev
$ nvm install
$ npm install
$ npm test
When making changes to the API, these changes must be reflected in the types
file (index.d.ts
). This library uses tsd
to test the types. These tests
can be found at index.d-test.ts
. As you make changes to the API, first update
index.d-test.ts
with the expected results, then update index.d.ts
with
the appropriate types to satisfy the tests.