Capturing the stdout

Hi all,

I am running a test case and need it capture the output it gives me.

require 'stringio'
old_stdout = $stdout
$stdout = StringIO.new
............
@reqout = $stdout.string
$stdout = old_stdout

the above code captures the stdout if something written explicily (like puts).
but it doesnot capture the output or running a testcase.
i want to capture the results if the testcases.
is there a way it could be done.

Thanks,
Navya.

···

---------------------------------
Yahoo! Mail
Use Photomail to share photos without annoying attachments.

Navya Amerineni wrote:

Hi all,

I am running a test case and need it capture the output it gives me.

require 'stringio'
old_stdout = $stdout
$stdout = StringIO.new
...........
@reqout = $stdout.string
$stdout = old_stdout

the above code captures the stdout if something written explicily (like puts).
but it doesnot capture the output or running a testcase.
i want to capture the results if the testcases.
is there a way it could be done.

You probably must redirect $defout also. Other than that doing it externally (i.e. in the calling process) might be an option, too.

Kind regards

  robert

IIRC $defout is deprecated.

You can separate the output (from the code being tested) from the information
displayed by Test::Unit as follows:

include Test::Unit::UI
out = StringIO.new ""
test_out = StringIO.new ""

$stdout = out
ret = Console::TestRunner.new(my_test_suite, NORMAL, test_out).start
$stdout = STDOUT
# "actual" stdout in out.string
# Test::Unit info in test_out.string

If you also want to capture the stdout from subprocesses, you need something
like
  begin
    old_stdout = STDOUT.dup
    STDOUT.reopen("mytempfile", "w")
    ...
    # create the TestRunner, possibly redirecting its output as shown above;
    # fork, system, ``, etc. in the tested code will be redirected to the
    # tempfile
    ...
  ensure
    STDOUT.reopen(old_stdout)
  end

Silly example of the former:

RUBY_VERSION # => "1.8.4"
require 'stringio'
require 'test/unit/ui/console/testrunner'

class Foo
  def foo(a,b)
    if b > 3
      puts a + b
    else
      puts a - b
    end
   
    a * b
  end
end

class TC_Foo < Test::Unit::TestCase
  def setup; @foo = Foo.new end
  
  def test_foo
    assert_equal(8, @foo.foo(2, 4))
  end
end

include Test::Unit::UI
out = StringIO.new ""
test_out = StringIO.new ""

$stdout = out
ret = Console::TestRunner.new(TC_Foo.suite, NORMAL, test_out).start
$stdout = STDOUT

puts "Output:"
out.string.each{|x| puts "-> #{x}"}
puts "----"
puts "Test (#{ret.passed? ? "passed" : "failed"}):"
test_out.string.each{|x| puts "->> #{x}"}
# >> Output:
# >> -> 6
# >> ----
# >> Test (passed):
# >> ->> Loaded suite TC_Foo
# >> ->> Started
# >> ->> .
# >> ->> Finished in 0.000601 seconds.
# >> ->>
# >> ->> 1 tests, 1 assertions, 0 failures, 0 errors

···

On Mon, Mar 20, 2006 at 07:39:02PM +0900, Robert Klemme wrote:

>I am running a test case and need it capture the output it gives me.
>
>require 'stringio'
>old_stdout = $stdout
>$stdout = StringIO.new
>...........
>@reqout = $stdout.string
>$stdout = old_stdout
>
>
>the above code captures the stdout if something written explicily (like
>puts).
>but it doesnot capture the output or running a testcase.
>i want to capture the results if the testcases.
>is there a way it could be done.

You probably must redirect $defout also. Other than that doing it
externally (i.e. in the calling process) might be an option, too.

--
Mauricio Fernandez - http://eigenclass.org - singular Ruby