- provide
it.only
anddescribe.only
which add only those tests to the tests to run, really handy - add a
--bail
option, which stops test execution after the first failure - move
buildSpy()
into it's own npm module (maybe "kavun-spy") - IDEA: write a codemodder that fixes
beforeEach
etc. and converts them into proper tests!? to become more compatible with mocha - move the time tracking into the reporter, if the reporter wants to do anything fancy with the time, it needs to be in control and a test runner just runs tests does NOT measure or alike
- show a stacktrace that contains the failing test (currently its always just
runner.js
) (why? e.g. so it can be clicked on in IDEs) - Keep improving tenets!
- Move the changelog-checker into own repo (it is https://github.com/wolframkriesing/to-do-list-checker)
- Improvement: Simplified the
npm test
command a bit, and use a (pure) linux command, not a certain shell - ??? something useful :)
- move tests inside lib directory, close to the source
- BUG: make the check-todo-list (using CHANGELOG.md) script work properly (by testing it :) )
- write tenets for kavun (like: no unnecessary dependency, as fast as possible, )
- add release scripts
- remove sinon by internal, simple spy, see #6
- refactor test file loaders, so we can pass filter criteria as params
- pass any number of file names of tests that shall be loaded
- provide
it
anddescribe
, and use it at least internally everywhere - provide
xit
andxdescribe
- move to pure
import
(using esm) to not needrequire
and be browser compatible - remove the dependency on uuid, and fix the timetracking for each test (it used to use the same timer for multiple tests)
- make the reporter configurable
- build a minimal reporter
- BUG: kavun was not installable anymore, fixed