bandage

Simple generator based test library

Stats

StarsIssuesVersionUpdatedCreatedSize
bandage
450.5.06 years ago6 years agoMinified + gzip package size for bandage in KB

Readme

bandage

npm status build status coverage status dependency status

Generator best test harness and test runner.

Usage

Use like if tape supported generator functions. Just always pass a generator function to test and stop calling end/done. The test is always done when the function is done.

var test = require('bandage');

test('yielding test', function *(t) {
  t.equal(5, 2+3, 'addition works');
  var v = yield Promise.resolve(true);
  t.ok(v, 'promise resolved correctly');
});

test('plain test', function *(t) {
  t.equal(6, 2*3, 'multiplication works');
});

Which you can run with the bundled (vowel-free) bndg executable:

$ bndg test.js
TAP version 13
# yielding test
ok 1 addition works
ok 2 promise resolved correctly
# plain test
ok 3 multiplication works

1..3
# tests 3
# pass  3
# fail  0

Error handling

By default, bandage catches errors and reports it as a test failure with a stack trace if you do not catch it, or use t.throws. Here's an example:

var test = require('bandage');

var rejecting = function () {
  return new Promise(function (resolve, reject) {
    setTimeout(function () {
      reject(new Error('async reject'));
    });
  });
};

test('error is caught', function *T(t) {
  yield rejecting();
  t.fail('we do not reach this');
});

Which will output:

TAP version 13
# error is caught
not ok 1 Error: async reject
  ---
    operator: error
    expected: undefined
    actual: [Error: async reject]
    at: null._onTimeout (/path/to/test.js:6:14)
    stack:
      Error: async reject
        at null._onTimeout (/path/to/test.js:6:14)
        at Timer.listOnTimeout (timers.js:92:15)
  ...

1..1
# tests 1
# pass  0
# fail  1

If you would like to test that errors are correctly passed through, just catch them yourself:

var test = require('bandage');

var rejecting = function () {
  return new Promise(function (resolve, reject) {
    setTimeout(function () {
      reject(new Error('async reject'));
    });
  });
};

test('error is caught', function *T(t) {
  t.plan(2);
  try {
    yield rejecting();
  }
  catch (err) {
    t.equal(err.message, 'async reject', 'caught async reject');
  }
  t.pass('we do reach this');
});

Which will output:

TAP version 13
# error is caught
ok 1 caught async reject
ok 2 we do reach this

1..2
# tests 2
# pass  2
# fail  0

If something unexpectedly throws in some callback stack that isn't matched by anything, that will still provide a TAP breaking stack trace. That is still a failed test though - there's only so much we can do at this point.

Subtests

Subtests work with bandage. The only difference is you need to yield it:

var test = require('bandage');

var slow = function () {
  return new Promise(function (res) {
    setTimeout(function () {
      res(true);
    }, 50);
  });
};

test('nested tests', function *(t) {
  var a = yield slow();
  t.ok(a, 'waited for slow');

  yield t.test('subtest', function *(st) {
    var b = yield slow();
    st.ok(b, 'waited for slow in subtest');
  });

  t.pass('pass after waiting for subtest')
});

Which outputs

TAP version 13
# nested tests
ok 1 waited for slow
ok 2 subtest
ok 3 waited for slow in subtest
ok 4 pass after waiting for subtest

1..4
# tests 4
# pass  4
# fail  0

Setup and Teardown

Because tests are executed sequentially in the order of the file, you can create setup tests at the top of your file, and teardown tests at the bottom.

If you find any bugs or have a feature request, please open an issue on github!

The npm package download data comes from npm's download counts api and package details come from npms.io.