Getting started testing Perl module with Test2
Before you can use it you need to install it on your computer. Look for one of the articles explaining how to install Perl Modules.
cpanm Test2
Project directory structure
Our files use the following directory structure. This is the standard layout for Perl code. At least code that reaches CPAN.
├── lib │ └── App.pm └── t ├── 01-test.t ├── 02-test.t └── 03-test.t
The application under test is very simple:
examples/test2-first/lib/App.pm
package App; use strict; use warnings; sub add { my ($x, $y) = @_; return $x * $y; } 1;
If you look closely you might notice a bug in the code. That's on purpose, so we can easily show failing tests.
Anyway the first test we write looks like this:
examples/test2-first/t/01-test.t
use Test2::V0; use App; ok( App::add(2, 2) == 4 ); done_testing();
The heart of the test is calling the function of the application under test and comparing it to the expected value. The ok function provided by the Test2::V0 module accepts a boolean value and displays ok or not ok according to the truth value of the expression.
done_testing indicates that we successfully reached the end of our tests. This helps us avoid an early exit from the test script that might hide many failing tests.
We can run the tests using the prove command:
$ prove -l t/01-test.t
The output will look like this:
t/01-test.t .. t/01-test.t .. ok All tests successful. Files=1, Tests=1, 0 wallclock secs ( 0.02 usr 0.00 sys + 0.06 cusr 0.01 csys = 0.09 CPU) Result: PASS
Showing errors
The observer might notice that the above success was a happy (or not so happy?) coincidence. With almost any other pair of input values the test would fail.
Let's see what happens when we add another test case. (I've created a separate test file so you can see it in full.
examples/test2-first/t/02-test.t
use Test2::V0; use App; ok( App::add(2, 2) == 4 ); ok( App::add(3, 4) == 7 ); done_testing();
Let's run it:
$ prove -l t/02-test.t
The result looks like this:
t/02-test.t .. 1/? # Failed test at t/02-test.t line 5. # Seeded srand with seed '20190419' from local date. t/02-test.t .. Dubious, test returned 1 (wstat 256, 0x100) Failed 1/2 subtests Test Summary Report ------------------- t/02-test.t (Wstat: 256 Tests: 2 Failed: 1) Failed test: 2 Non-zero exit status: 1 Files=1, Tests=2, 0 wallclock secs ( 0.02 usr 0.00 sys + 0.07 cusr 0.00 csys = 0.09 CPU) Result: FAIL
We can see that something has failed. We can even see the number of the failed test is 2, but we can do better than that.
Test failure with details
Instead of using the ok function that accepts a boolean we used the is function provided by Test2::V0. It accepts two values. We pass to it the the real result of an operation followed by expected value.
examples/test2-first/t/03-test.t
use Test2::V0; use App; is( App::add(2, 2), 4 ); is( App::add(3, 4), 7 ); done_testing();
Let's run it:
prove -l t/03-test.t
The result looks like this:
t/03-test.t .. 1/? # Failed test at t/03-test.t line 5. # +-----+----+-------+ # | GOT | OP | CHECK | # +-----+----+-------+ # | 12 | eq | 7 | # +-----+----+-------+ # Seeded srand with seed '20190419' from local date. t/03-test.t .. Dubious, test returned 1 (wstat 256, 0x100) Failed 1/2 subtests Test Summary Report ------------------- t/03-test.t (Wstat: 256 Tests: 2 Failed: 1) Failed test: 2 Non-zero exit status: 1 Files=1, Tests=2, 0 wallclock secs ( 0.02 usr 0.00 sys + 0.07 cusr 0.00 csys = 0.09 CPU) Result: FAIL
Not only can you see the failure, it also tells us what was the received value (GOT) and what was the expected value (CHECK). It also tells us that the operator used to compare the two values was eq. While it is not ideal for numbers, it is good in most cases even when we are comparing numbers.
In any case this is pretty nice.
About Test2
For more information see also the Test2-Suite by Chad 'Exodist' Granum.
Comments
Even better would be to make use of the optional parameter to `is()` (and `ok()`), namey the description of the test, e.g.
is( App::add(3, 4), 7 , "add(3, 4) = 3+4 = 7");
Published on 2019-04-19