package Test::Output; use warnings; use strict; use Test::Builder; use Capture::Tiny qw/capture capture_stdout capture_stderr capture_merged/; use Exporter qw(import); our %EXPORT_TAGS = ( stdout => [ qw( stdout_is stdout_isnt stdout_like stdout_unlike ) ], stderr => [ qw( stderr_is stderr_isnt stderr_like stderr_unlike ) ], output => [ qw( output_is output_isnt output_like output_unlike ) ], combined => [ qw( combined_is combined_isnt combined_like combined_unlike ) ], functions => [ qw( output_from stderr_from stdout_from combined_from ) ], tests => [ qw( output_is output_isnt output_like output_unlike stderr_is stderr_isnt stderr_like stderr_unlike stdout_is stdout_isnt stdout_like stdout_unlike combined_is combined_isnt combined_like combined_unlike ) ], all => [ qw( output_is output_isnt output_like output_unlike stderr_is stderr_isnt stderr_like stderr_unlike stdout_is stdout_isnt stdout_like stdout_unlike combined_is combined_isnt combined_like combined_unlike output_from stderr_from stdout_from combined_from ) ], ); our @EXPORT = keys %{ { map { $_ => 1 } map { @{ $EXPORT_TAGS{$_} } } keys %EXPORT_TAGS } }; my $Test = Test::Builder->new; =encoding utf8 =head1 NAME Test::Output - Utilities to test STDOUT and STDERR messages. =cut our $VERSION = '1.031'; =head1 SYNOPSIS use Test::More tests => 4; use Test::Output; sub writer { print "Write out.\n"; print STDERR "Error out.\n"; } stdout_is(\&writer,"Write out.\n",'Test STDOUT'); stderr_isnt(\&writer,"No error out.\n",'Test STDERR'); combined_is( \&writer, "Write out.\nError out.\n", 'Test STDOUT & STDERR combined' ); output_is( \&writer, "Write out.\n", "Error out.\n", 'Test STDOUT & STDERR' ); # Use bare blocks. stdout_is { print "test" } "test", "Test STDOUT"; stderr_isnt { print "bad test" } "test", "Test STDERR"; output_is { print 'STDOUT'; print STDERR 'STDERR' } "STDOUT", "STDERR", "Test output"; =head1 DESCRIPTION Test::Output provides a simple interface for testing output sent to C or C. A number of different utilities are included to try and be as flexible as possible to the tester. Likewise, L provides a much more robust capture mechanism without than the original L. =cut =head1 TESTS =cut =head2 STDOUT =over 4 =item B =item B stdout_is ( $coderef, $expected, 'description' ); stdout_is { ... } $expected, 'description'; stdout_isnt( $coderef, $expected, 'description' ); stdout_isnt { ... } $expected, 'description'; C captures output sent to C from C<$coderef> and compares it against C<$expected>. The test passes if equal. C passes if C is not equal to C<$expected>. =cut sub stdout_is (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; my $stdout = stdout_from($test); my $ok = ( $stdout eq $expected ); $Test->ok( $ok, $description ) || $Test->diag("STDOUT is:\n$stdout\nnot:\n$expected\nas expected"); return $ok; } sub stdout_isnt (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; my $stdout = stdout_from($test); my $ok = ( $stdout ne $expected ); $Test->ok( $ok, $description ) || $Test->diag("STDOUT:\n$stdout\nmatching:\n$expected\nnot expected"); return $ok; } =item B =item B stdout_like ( $coderef, qr/$expected/, 'description' ); stdout_like { ... } qr/$expected/, 'description'; stdout_unlike( $coderef, qr/$expected/, 'description' ); stdout_unlike { ... } qr/$expected/, 'description'; C captures the output sent to C from C<$coderef> and compares it to the regex in C<$expected>. The test passes if the regex matches. C passes if STDOUT does not match the regex. =back =cut sub stdout_like (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; unless ( my $regextest = _chkregex( 'stdout_like' => $expected ) ) { return $regextest; } my $stdout = stdout_from($test); my $ok = ( $stdout =~ $expected ); $Test->ok( $ok, $description ) || $Test->diag("STDOUT:\n$stdout\ndoesn't match:\n$expected\nas expected"); return $ok; } sub stdout_unlike (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; unless ( my $regextest = _chkregex( 'stdout_unlike' => $expected ) ) { return $regextest; } my $stdout = stdout_from($test); my $ok = ( $stdout !~ $expected ); $Test->ok( $ok, $description ) || $Test->diag("STDOUT:\n$stdout\nmatches:\n$expected\nnot expected"); return $ok; } =head2 STDERR =over 4 =item B =item B stderr_is ( $coderef, $expected, 'description' ); stderr_is {... } $expected, 'description'; stderr_isnt( $coderef, $expected, 'description' ); stderr_isnt {... } $expected, 'description'; C is similar to C, except that it captures C. The test passes if C from C<$coderef> equals C<$expected>. C passes if C is not equal to C<$expected>. =cut sub stderr_is (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; my $stderr = stderr_from($test); my $ok = ( $stderr eq $expected ); $Test->ok( $ok, $description ) || $Test->diag("STDERR is:\n$stderr\nnot:\n$expected\nas expected"); return $ok; } sub stderr_isnt (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; my $stderr = stderr_from($test); my $ok = ( $stderr ne $expected ); $Test->ok( $ok, $description ) || $Test->diag("STDERR:\n$stderr\nmatches:\n$expected\nnot expected"); return $ok; } =item B =item B stderr_like ( $coderef, qr/$expected/, 'description' ); stderr_like { ...} qr/$expected/, 'description'; stderr_unlike( $coderef, qr/$expected/, 'description' ); stderr_unlike { ...} qr/$expected/, 'description'; C is similar to C except that it compares the regex C<$expected> to C captured from C<$codref>. The test passes if the regex matches. C passes if C does not match the regex. =back =cut sub stderr_like (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; unless ( my $regextest = _chkregex( 'stderr_like' => $expected ) ) { return $regextest; } my $stderr = stderr_from($test); my $ok = ( $stderr =~ $expected ); $Test->ok( $ok, $description ) || $Test->diag("STDERR:\n$stderr\ndoesn't match:\n$expected\nas expected"); return $ok; } sub stderr_unlike (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; unless ( my $regextest = _chkregex( 'stderr_unlike' => $expected ) ) { return $regextest; } my $stderr = stderr_from($test); my $ok = ( $stderr !~ $expected ); $Test->ok( $ok, $description ) || $Test->diag("STDERR:\n$stderr\nmatches:\n$expected\nnot expected"); return $ok; } =head2 COMBINED OUTPUT =over 4 =item B =item B combined_is ( $coderef, $expected, 'description' ); combined_is {... } $expected, 'description'; combined_isnt ( $coderef, $expected, 'description' ); combined_isnt {... } $expected, 'description'; C directs C to C then captures C. This is equivalent to UNIXs C<< 2>&1 >>. The test passes if the combined C and C from $coderef equals $expected. C passes if combined C and C are not equal to C<$expected>. =cut sub combined_is (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; my $combined = combined_from($test); my $ok = ( $combined eq $expected ); $Test->ok( $ok, $description ) || $Test->diag( "STDOUT & STDERR are:\n$combined\nnot:\n$expected\nas expected"); return $ok; } sub combined_isnt (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; my $combined = combined_from($test); my $ok = ( $combined ne $expected ); $Test->ok( $ok, $description ) || $Test->diag( "STDOUT & STDERR:\n$combined\nmatching:\n$expected\nnot expected"); return $ok; } =item B =item B combined_like ( $coderef, qr/$expected/, 'description' ); combined_like { ...} qr/$expected/, 'description'; combined_unlike ( $coderef, qr/$expected/, 'description' ); combined_unlike { ...} qr/$expected/, 'description'; C is similar to C except that it compares a regex (C<$expected)> to C and C captured from C<$codref>. The test passes if the regex matches. C passes if the combined C and C does not match the regex. =back =cut sub combined_like (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; unless ( my $regextest = _chkregex( 'combined_like' => $expected ) ) { return $regextest; } my $combined = combined_from($test); my $ok = ( $combined =~ $expected ); $Test->ok( $ok, $description ) || $Test->diag( "STDOUT & STDERR:\n$combined\ndon't match:\n$expected\nas expected"); return $ok; } sub combined_unlike (&$;$$) { my $test = shift; my $expected = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; unless ( my $regextest = _chkregex( 'combined_unlike' => $expected ) ) { return $regextest; } my $combined = combined_from($test); my $ok = ( $combined !~ $expected ); $Test->ok( $ok, $description ) || $Test->diag( "STDOUT & STDERR:\n$combined\nmatching:\n$expected\nnot expected"); return $ok; } =head2 OUTPUT =over 4 =item B =item B output_is ( $coderef, $expected_stdout, $expected_stderr, 'description' ); output_is {... } $expected_stdout, $expected_stderr, 'description'; output_isnt( $coderef, $expected_stdout, $expected_stderr, 'description' ); output_isnt {... } $expected_stdout, $expected_stderr, 'description'; The C function is a combination of the C and C functions. For example: output_is(sub {print "foo"; print STDERR "bar";},'foo','bar'); is functionally equivalent to stdout_is(sub {print "foo";},'foo') && stderr_is(sub {print STDERR "bar";'bar'); except that C<$coderef> is only executed once. Unlike C and C which ignore STDERR and STDOUT respectively, C requires both C and C to match in order to pass. Setting either C<$expected_stdout> or C<$expected_stderr> to C ignores C or C respectively. output_is(sub {print "foo"; print STDERR "bar";},'foo',undef); is the same as stdout_is(sub {print "foo";},'foo') C provides the opposite function of C. It is a combination of C and C. output_isnt(sub {print "foo"; print STDERR "bar";},'bar','foo'); is functionally equivalent to stdout_is(sub {print "foo";},'bar') && stderr_is(sub {print STDERR "bar";'foo'); As with C, setting either C<$expected_stdout> or C<$expected_stderr> to C ignores the output to that facility. output_isnt(sub {print "foo"; print STDERR "bar";},undef,'foo'); is the same as stderr_is(sub {print STDERR "bar";},'foo') =cut sub output_is (&$$;$$) { my $test = shift; my $expout = shift; my $experr = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; my ( $stdout, $stderr ) = output_from($test); my $ok = 1; my $diag; if ( defined($experr) && defined($expout) ) { unless ( $stdout eq $expout ) { $ok = 0; $diag .= "STDOUT is:\n$stdout\nnot:\n$expout\nas expected"; } unless ( $stderr eq $experr ) { $diag .= "\n" unless ($ok); $ok = 0; $diag .= "STDERR is:\n$stderr\nnot:\n$experr\nas expected"; } } elsif ( defined($expout) ) { $ok = ( $stdout eq $expout ); $diag .= "STDOUT is:\n$stdout\nnot:\n$expout\nas expected"; } elsif ( defined($experr) ) { $ok = ( $stderr eq $experr ); $diag .= "STDERR is:\n$stderr\nnot:\n$experr\nas expected"; } else { unless ( $stdout eq '' ) { $ok = 0; $diag .= "STDOUT is:\n$stdout\nnot:\n\nas expected"; } unless ( $stderr eq '' ) { $diag .= "\n" unless ($ok); $ok = 0; $diag .= "STDERR is:\n$stderr\nnot:\n\nas expected"; } } $Test->ok( $ok, $description ) || $Test->diag($diag); return $ok; } sub output_isnt (&$$;$$) { my $test = shift; my $expout = shift; my $experr = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; my ( $stdout, $stderr ) = output_from($test); my $ok = 1; my $diag; if ( defined($experr) && defined($expout) ) { if ( $stdout eq $expout ) { $ok = 0; $diag .= "STDOUT:\n$stdout\nmatching:\n$expout\nnot expected"; } if ( $stderr eq $experr ) { $diag .= "\n" unless ($ok); $ok = 0; $diag .= "STDERR:\n$stderr\nmatching:\n$experr\nnot expected"; } } elsif ( defined($expout) ) { $ok = ( $stdout ne $expout ); $diag = "STDOUT:\n$stdout\nmatching:\n$expout\nnot expected"; } elsif ( defined($experr) ) { $ok = ( $stderr ne $experr ); $diag = "STDERR:\n$stderr\nmatching:\n$experr\nnot expected"; } else { if ( $stdout eq '' ) { $ok = 0; $diag = "STDOUT:\n$stdout\nmatching:\n\nnot expected"; } if ( $stderr eq '' ) { $diag .= "\n" unless ($ok); $ok = 0; $diag .= "STDERR:\n$stderr\nmatching:\n\nnot expected"; } } $Test->ok( $ok, $description ) || $Test->diag($diag); return $ok; } =item B =item B output_like ( $coderef, $regex_stdout, $regex_stderr, 'description' ); output_like { ... } $regex_stdout, $regex_stderr, 'description'; output_unlike( $coderef, $regex_stdout, $regex_stderr, 'description' ); output_unlike { ... } $regex_stdout, $regex_stderr, 'description'; C and C follow the same principles as C and C except they use a regular expression for matching. C attempts to match C<$regex_stdout> and C<$regex_stderr> against C and C produced by $coderef. The test passes if both match. output_like(sub {print "foo"; print STDERR "bar";},qr/foo/,qr/bar/); The above test is successful. Like C, setting either C<$regex_stdout> or C<$regex_stderr> to C ignores the output to that facility. output_like(sub {print "foo"; print STDERR "bar";},qr/foo/,undef); is the same as stdout_like(sub {print "foo"; print STDERR "bar";},qr/foo/); C test pass if output from C<$coderef> doesn't match C<$regex_stdout> and C<$regex_stderr>. =back =cut sub output_like (&$$;$$) { my $test = shift; my $expout = shift; my $experr = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; my ( $stdout, $stderr ) = output_from($test); my $ok = 1; unless ( my $regextest = _chkregex( 'output_like_STDERR' => $experr, 'output_like_STDOUT' => $expout ) ) { return $regextest; } my $diag; if ( defined($experr) && defined($expout) ) { unless ( $stdout =~ $expout ) { $ok = 0; $diag .= "STDOUT:\n$stdout\ndoesn't match:\n$expout\nas expected"; } unless ( $stderr =~ $experr ) { $diag .= "\n" unless ($ok); $ok = 0; $diag .= "STDERR:\n$stderr\ndoesn't match:\n$experr\nas expected"; } } elsif ( defined($expout) ) { $ok = ( $stdout =~ $expout ); $diag .= "STDOUT:\n$stdout\ndoesn't match:\n$expout\nas expected"; } elsif ( defined($experr) ) { $ok = ( $stderr =~ $experr ); $diag .= "STDERR:\n$stderr\ndoesn't match:\n$experr\nas expected"; } else { unless ( $stdout eq '' ) { $ok = 0; $diag .= "STDOUT is:\n$stdout\nnot:\n\nas expected"; } unless ( $stderr eq '' ) { $diag .= "\n" unless ($ok); $ok = 0; $diag .= "STDERR is:\n$stderr\nnot:\n\nas expected"; } } $Test->ok( $ok, $description ) || $Test->diag($diag); return $ok; } sub output_unlike (&$$;$$) { my $test = shift; my $expout = shift; my $experr = shift; my $options = shift if ( ref( $_[0] ) ); my $description = shift; my ( $stdout, $stderr ) = output_from($test); my $ok = 1; unless ( my $regextest = _chkregex( 'output_unlike_STDERR' => $experr, 'output_unlike_STDOUT' => $expout ) ) { return $regextest; } my $diag; if ( defined($experr) && defined($expout) ) { if ( $stdout =~ $expout ) { $ok = 0; $diag .= "STDOUT:\n$stdout\nmatches:\n$expout\nnot expected"; } if ( $stderr =~ $experr ) { $diag .= "\n" unless ($ok); $ok = 0; $diag .= "STDERR:\n$stderr\nmatches:\n$experr\nnot expected"; } } elsif ( defined($expout) ) { $ok = ( $stdout !~ $expout ); $diag .= "STDOUT:\n$stdout\nmatches:\n$expout\nnot expected"; } elsif ( defined($experr) ) { $ok = ( $stderr !~ $experr ); $diag .= "STDERR:\n$stderr\nmatches:\n$experr\nnot expected"; } $Test->ok( $ok, $description ) || $Test->diag($diag); return $ok; } =head1 EXPORTS By default, all subroutines are exported by default. =over 4 =item * :stdout - the subs with C in the name. =item * :stderr - the subs with C in the name. =item * :functions - the subs with C<_from> at the end. =item * :output - the subs with C in the name. =item * :combined - the subs with C in the name. =item * :tests - everything that outputs TAP =item * :all - everything (which is the same as the default) =back =head1 FUNCTIONS =cut =head2 stdout_from my $stdout = stdout_from($coderef) my $stdout = stdout_from { ... }; stdout_from() executes $coderef and captures STDOUT. =cut sub stdout_from (&) { my $test = shift; my $stdout = capture_stdout { select( ( select(STDOUT), $| = 1 )[0] ); $test->() }; return $stdout; } =head2 stderr_from my $stderr = stderr_from($coderef) my $stderr = stderr_from { ... }; C executes C<$coderef> and captures C. =cut sub stderr_from (&) { my $test = shift; # XXX why is this here and not in output_from or combined_from -- xdg, 2012-05-13 local $SIG{__WARN__} = sub { print STDERR @_ } if $] < 5.008; my $stderr = capture_stderr { select( ( select(STDERR), $| = 1 )[0] ); $test->() }; return $stderr; } =head2 output_from my ($stdout, $stderr) = output_from($coderef) my ($stdout, $stderr) = output_from {...}; C executes C<$coderef> one time capturing both C and C. =cut sub output_from (&) { my $test = shift; my ($stdout, $stderr) = capture { select( ( select(STDOUT), $| = 1 )[0] ); select( ( select(STDERR), $| = 1 )[0] ); $test->(); }; return ( $stdout, $stderr ); } =head2 combined_from my $combined = combined_from($coderef); my $combined = combined_from {...}; C executes C<$coderef> one time combines C and C, and captures them. C is equivalent to using C<< 2>&1 >> in UNIX. =cut sub combined_from (&) { my $test = shift; my $combined = capture_merged { select( ( select(STDOUT), $| = 1 )[0] ); select( ( select(STDERR), $| = 1 )[0] ); $test->(); }; return $combined; } sub _chkregex { my %regexs = @_; foreach my $test ( keys(%regexs) ) { next unless ( defined( $regexs{$test} ) ); my $usable_regex = $Test->maybe_regex( $regexs{$test} ); unless ( defined($usable_regex) ) { my $ok = $Test->ok( 0, $test ); $Test->diag("'$regexs{$test}' doesn't look much like a regex to me."); # unless $ok; return $ok; } } return 1; } =head1 AUTHOR Currently maintained by brian d foy, C. Shawn Sorichetti, C<< >> =head1 SOURCE AVAILABILITY This module is in Github: http://github.com/briandfoy/test-output/tree/master =head1 BUGS Please report any bugs or feature requests to C, or through the web interface at L. I will be notified, and then you'll automatically be notified of progress on your bug as I make changes. =head1 ACKNOWLEDGEMENTS Thanks to chromatic whose TieOut.pm was the basis for capturing output. Also thanks to rjbs for his help cleaning the documentation, and pushing me to L. (This feature has been removed since it uses none of L's strengths). Thanks to David Wheeler for providing code block support and tests. Thanks to Michael G Schwern for the solution to combining C and C. =head1 COPYRIGHT & LICENSE Copyright 2005-2013 Shawn Sorichetti, All Rights Reserved. This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself. =cut 1; # End of Test::Output