Tests can be reported in a variety of different ways.
When you run a test script directly, it'll always output TAP. The tap runner will interpret this output, and can format it in a variety of different ways.
Node-tap includes 2 reporting engines, and you can extend either one, or consume the TAP formatted output in custom modules of your own.
The newer React-based reporter is called treport. It uses ink to provide live feedback about tests in progress.
The older streams-based bundled reporting engine is tap-mocha-reporter, so named because it ports many of the report styles built into mocha.
The --reporter or -R argument on the command line can specify:
--reporter-arg=<arg> or -r<arg> option can be specified one or more times
to provide a list of arguments to pass to CLI reporters.The built-in reports are:
The default when stdout is a terminal and colors are enabled. Also the class to extend to create new treport reporters. Does all the things, handles all the edge cases, and ends with a pleasant surprise.
A lot like Base, but says a lot less. No timer, no list of tests concurrently running, nothing printed on test passing. Just the failures and the terse summary.
A spec style reporter with the current running jobs and Terse summary and
footer.
Setting --reporter=tap will dump the output as a raw TAP stream.
This is the default when stdout is not a terminal, or colors are disabled.
The old default. Show a summary of each test file being run, along with reporting each failure and pending test.
Output absolutely nothing
Output based on rspec, with hierarchical indentation and unicode red and green checks and X's.
XML output popular in .NET land.
Output results in one big JSON object.
Output results as a stream of \n-delimited JSON.
Output a dot for each pass and failure.
List out each test as it's run.
Just the post-test summary of failures and test count.
A magical cat who is also a toaster pastry.
Mostly for debugging tap-mocha-reporter, dumping out the TAP output and the way that its being interpreted.