Tape.md

tape

tape is supported as a testing framework. It cannot be auto-detected.

Setup

tape writes its own TAP output with console.log, which console-fail-test would otherwise report as a failure in every test. Redirect that output to a stream, then call the Node API with the tape module:

// setupTests.js

import { cft } from "console-fail-test";
import tape from "tape";

tape.createStream().pipe(process.stdout);

cft({ testFramework: tape });

cft() must come after tape.createStream(), as it creates tape's default harness if one doesn't exist yet.

Import that setup file at the top of each test file that should be checked for console calls:

// some.test.js

import "./setupTests.js";
import test from "tape";

test("passes when the console is not called", (t) => {
	t.end();
});

Module caching ensures the setup only runs once, even when the tape CLI loads many test files in a single process. Avoid loading the setup file with the CLI's --require flag instead: that runs before tape.wait(), so tests in files that load asynchronously might be skipped.

Subtests

tape runs subtests created with t.test() before their parent test ends. console-fail-test only checks the console for the outermost test, so console calls in a subtest are reported as a failure of its parent test.

Ending Tests

console-fail-test checks the console as soon as a test ends. Calling t.end() ends the test synchronously, so any console calls after it in the same callback are not reported.

Plans

console-fail-test reports each console call as a failed assertion on the test. A test that uses t.plan() will therefore also fail with plan != count.