angelo v0.0.20
Angelo - A (rerunnable) Mocha test runner
Useful for running (and rerunning!) Mocha test suites.
Simple API. Load and play -- that's all you need to know!
Install
npm install angelo
Example usage
Angelo = require('angelo').Angelo
angelo = new Angelo()
angelo.load("/path/to/test.js")
angelo.play()
FAQ
Why "Angelo"?
Angelo Moriondo is the inventor of the first espresso machine.
Okay, but why use this library?
You might like Angelo if:
1) You use mocha programmatically.
and
2) You want to rerun a test suite multiple times.
"IMO this should be done with complete process isolation, you could use
json-stream or the json reporter and exec(), plus then that keeps mocha
lighter. win-win situation :D" - TJ
And that's what Angelo does!
Mocha has an issue with re-running test suites within the same process. This is because Mocha uses a call to require
to load test files. When require
is called, however, files are cached and only loaded once. A side-effect of this reliance on require
is that tests are only ever run once.
As a workaround, you could clear the cache of loaded modules before reloading, but this isn't foolproof. Instead, Angelo does what TJ recommends -- it runs Mocha in a child process and receives results via the JSON stream reporter.
Angelo uses spawn
, instead of exec
, so that results can be streamed as they happen. Using exec
would buffer all results until all the tests were completed.
Lasly, another benefit of Angelo is the ability to run tests in parallel. Now that test execution is running in a separate child process, you can run multiple test files in parallel by creating multiple Angelo objects.