Testing — use Test

The built-in Test module — plan, ok, is, isnt, is-deeply — emitting TAP.

use Test brings in Raku's standard testing routines. Each assertion prints a TAP line (ok N - description or not ok N …), and plan declares how many you expect. This is the same module Raku's own test suite uses, and it runs unchanged in the playground.

A small test suite #

plan states the count; ok asserts a truthy value; is checks a value against an expected one; isnt checks inequality.

use Test;
plan 3;
ok 2 > 1, "two is greater than one";
is "ab".uc, "AB", "uppercasing";
isnt "cat", "dog", "different words";

1..3
ok 1 - two is greater than one
ok 2 - uppercasing
ok 3 - different words

Each passing assertion is numbered in order, and the 1..3 plan line lets a test harness confirm the whole suite ran.

Comparing structures — is-deeply #

is compares stringified values; is-deeply compares structure and type exactly, so it's the right check for lists, hashes, and nested data.

use Test;
plan 2;
is-deeply (1, 2, 3).grep(* > 1).List, (2, 3), "grep keeps the tail";
is-deeply { a => 1, b => 2 }, { b => 2, a => 1 }, "hash order-independent";

1..2
ok 1 - grep keeps the tail
ok 2 - hash order-independent

Notes #

• Core assertions: ok/nok (truthy/falsy), is/isnt (string-equal or not), is-deeply (structural), like/unlike (regex match), cmp-ok (compare with a named op), and dies-ok/lives-ok/throws-like for exceptions.
• A failing assertion prints not ok N followed by diagnostic lines showing the expected and received values, so a red test tells you what went wrong.
• plan N up front fixes the expected count; alternatively call done-testing at the end to let the count be inferred.
• subtest "name", { … } groups related assertions and reports them as one.
• The output format is TAP (Test Anything Protocol), consumable by prove and other standard harnesses.