/ 
TAP-Parser-0.51
River stage zero No dependents
/ TAP::Harness

NAME

TAP::Harness - Run Perl test scripts with statistics

VERSION

Version 0.51

DESCRIPTION

This is a simple test harness which allows tests to be run and results automatically aggregated and output to STDOUT.

SYNOPSIS

use TAP::Harness;
my $harness = TAP::Harness->new( \%args );
$harness->runtests(@tests);

METHODS

Class methods

new

my %args = (
   verbose => 1,
   lib     => [ 'lib', 'blib/lib' ],
)
my $harness = TAP::Harness->new( \%args );

The constructor returns a new TAP::Harness object. It accepts an optional hashref whose allowed keys are:

  • verbose

    Print individual test results to STDOUT.

  • timer

    Append run time for each test to output. Uses Time::HiRes if available.

  • failures

    Only show test failures (this is a no-op if verbose is selected).

  • lib

    Accepts a scalar value or array ref of scalar values indicating which paths to allowed libraries should be included if Perl tests are executed. Naturally, this only makes sense in the context of tests written in Perl.

  • switches

    Accepts a scalar value or array ref of scalar values indicating which switches should be included if Perl tests are executed. Naturally, this only makes sense in the context of tests written in Perl.

  • quiet

    Suppress some test output (mostly failures while tests are running).

  • really_quiet

    Suppress everything but the tests summary.

  • exec

    Typically, Perl tests are run through this. However, anything which spits out TAP is fine. You can use this argument to specify the name of the program (and optional switches) to run your tests with:

    exec => '/usr/bin/ruby -w'

  • execrc

    Location of 'execrc' file. See "USING EXECRC" below.

  • errors

    If parse errors are found in the TAP output, a note of this will be made in the summary report. To see all of the parse errors, set this argument to true:

    errors => 1

  • directives

    If set to a true value, only test results with directives will be displayed. This overrides other settings such as verbose or failures.

Instance Methods

runtests

$harness->runtests(@tests);

Accepts and array of @tests to be run. This should generally be the names of test files, but this is not required. Each element in @tests will be passed to TAP::Parser::new() as a source. See TAP::Parser for more information.

Tests will be run in the order found.

If the environment variable PERL_TEST_HARNESS_DUMP_TAP is defined it should name a directory into which a copy of the raw TAP for each test will be written. TAP is written to files named for each test. Subdirectories will be created as needed.

aggregate_tests

$harness->aggregate_tests( $aggregate, @tests );

Tests will be run in the order found.

SUBCLASSING

TAP::Harness is designed to be (mostly) easy to subclass. If you don't like how a particular feature functions, just override the desired methods.

Methods

The following methods are one's you may wish to override if you want to subclass TAP::Harness.

summary

$harness->summary( \%args );

summary prints the summary report after all tests are run. The argument is a hashref with the following keys:

  • start

    This is created with Benchmark->new and it the time the tests started. You can print a useful summary time, if desired, with:

    $self->output(timestr( timediff( Benchmark->new, $start_time ), 'nop' ));

  • aggregate

    This is the TAP::Parser::Aggregate object for all of the tests run.

  • tests

    This is an array reference of all test names. To get the TAP::Parser object for individual tests:

    my $aggregate = $args->{aggregate};
my $tests     = $args->{tests};

foreach my $name ( @$tests ) {
    my ($parser) = $aggregate->parsers($test);
    ... do something with $parser
}

    This is a bit clunky and will be cleaned up in a later release.

output

$harness->output(@list_of_strings_to_output);

All output from TAP::Harness is driven through this method. If you would like to redirect output somewhere else, just override this method.

failure_output

$harness->failure_output(@list_of_strings_to_output);

Identical to output, this method is called for any output which represents a failure.

balanced_range

my @ranges = $harness->balanced_range( $limit, @numbers );

Given a limit in the number of characters and a list of numbers, this method first creates a range of numbers with range and then groups them into individual strings which are roughly the length of $limit. Returns an array of strings.

range

my @range = $harness->range(@list_of_numbers);

Taks a list of numbers, sorts them, and returns a list of ranged strings:

print join ', ' $harness->range( 2, 7, 1, 3, 10, 9  );
# 1-3, 7, 9-10

output_test_failure

$harness->output_test_failure($parser);

As individual test programs are run, if a test program fails, this method is called to spit out the list of failed tests.

USING EXECRC

WARNING: this functionality is still experimental. While we intend to support it, the file format may change.

Sometimes you want to use different executables to run different tests. If that's the case, you'll need to create an execrc file. This file should be a YAML file. This should be representating a hash with (at present) three keys, exact, regex, and default.

---
exact: 
  # whoops!  We have a ruby test here!
  -
    - /usr/bin/ruby
    - t/ruby.t
regex:
  # let's test some web pages
  -
    - /usr/bin/perl
    - -w
    - bin/test_html.pl
    - ^https?://
default:
  -
    - /usr/bin/perl
    - -wT

  • exact

    The exact key should be an array reference with each element being an array reference whose items are an exact list of what need to be passed to the shell to execute the test. So for the 'exact' item of t/ruby.t above, we attempt to execute in the shell:

    /usr/bin/ruby t/ruby.t

  • regex

    This is the same as exact, except that the final element in each array reference should be a Perl regular expression. For the ^https?:// regular expression above, when the harness sees a test for http://www.example.com, it will pass the following to the shell:

    /usr/bin/perl -w bin/test_html.pl http://www.example.com/

  • default

    Any item which the harness does not match to another execrc entry will automatically be executed with the default.

Blank lines are allowed. Lines beginning with a '#' are comments (the '#' may have spaces in front of it).

So for the above execrc file, if it's named 'my_execrc' (as it is in the examples/ directory which comes with this distribution), then you could potentially run it like this, if you're using the runtests utility:

runtests --execrc my_execrc t/ - < list_of_urls.txt

Then for a test named t/test_is_written_in_ruby.t, it will be executed with:

/usr/bin/ruby -w t/test_is_written_in_ruby.t

If the list of urls contains "http://www.google.com/", it will be executed as follows:

/usr/bin/perl test_html.pl http://www.google.com/

Of course, if test_html.pl outputs anything other than TAP, this will fail.

See the README in the examples directory for a ready-to-run example.

REPLACING

If you like the runtests utility and TAP::Parser but you want your own harness, all you need to do is write one and provide new and runtests methods. Then you can use the runtests utility like so:

runtests --harness My::Test::Harness

Note that while runtests accepts a list of tests (or things to be tested), new has a fairly rich set of arguments. You'll probably want to read over this code carefully to see how all of them are being used.

SEE ALSO

Test::Harness