About
test-cmd.rb provides an object-oriented interface for spawning a command.
Examples
Callbacks
The success and failure callbacks provide hooks for when a command exits successfully or unsuccessfully. The callback is passed an instance of Test::Cmd:
require "test-cmd"
cmd("ruby", "-e", "exit 0")
.success { print "The command [#{_1.pid}] was successful", "\n" }
.failure { print "The command [#{_1.pid}] was unsuccessful", "\n" }
Test::Unit
The following example demonstrates how tests might be written with
test-unit from the standard library. The
cmd
method takes the name or path of a command, alongside any arguments:
require "test/unit"
require "test/cmd"
class CmdTest < Test::Unit::TestCase
def test_ruby_stdout
assert_equal "42\n", ruby("puts 42").stdout
end
def test_ruby_stderr
assert_equal "42\n", ruby("warn 42").stderr
end
def test_ruby_success_exit_status
assert_equal 0, ruby("exit 0").exit_status
end
def test_ruby_failure_exit_status
assert_equal 1, ruby("exit 1").exit_status
end
private
def ruby(code)
cmd("ruby", "-e", code)
end
end
IO#sync
Sometimes it can be neccessary to bypass Ruby's internal buffer and flush
output to the operating system immediately, otherwise there can be unexpected
results. Consider the following example, where the output will be
bar\nfoo\n
rather than foo\nbar\n
:
##
# test.rb
fork do
sleep(1)
puts "bar"
end
puts "foo"
Process.wait
##
# cmd.rb
p cmd("ruby", "test.rb").stdout # => "bar\nfoo\n"
And with output flushed to the operating system immediately:
##
# test.rb
$stdout.sync = true
fork do
sleep(1)
puts "bar"
end
puts "foo"
Process.wait
##
# cmd.rb
p cmd("ruby", "test.rb").stdout # => "foo\nbar\n"
Documentation
A complete API reference is available at 0x1eef.github.io/x/test-cmd.rb
Install
Rubygems.org
test-cmd.rb can be installed via rubygems.org:
gem install test-cmd.rb