Documentation | Report Issue | Source Code
Minitap is a Minitest plugin that reports test results in TAP-Y or TAP-J. Those are the formats that Tapout reads, so with Minitap a Minitest suite can be shown in any of Tapout's report formats.
$ ruby -Ilib test/some_test.rb --tapy | tapout progress
Minitap works with Minitest 5.11 and newer, including Minitest 6, on Ruby 2.7 and newer.
$ gem install minitap
Or in a Gemfile:
gem 'minitap'
Tapout is installed along with it.
Load Minitap in your test helper, after Minitest:
require 'minitest/autorun'
require 'minitap'
That adds two options to the test command line, --tapy and --tapj.
Without one of them Minitest reports as usual.
$ ruby -Ilib test/some_test.rb --tapy
The output is a TAP-Y stream, which isn't meant to be read as it is.
Pipe it to tapout, with the name of the report you want:
$ ruby -Ilib test/some_test.rb --tapy | tapout progress
tapout --help lists the reports.
Rake's TestTask and Minitest's own Minitest::TestTask pass the
TESTOPTS variable on to Minitest:
$ rake test TESTOPTS=--tapy | tapout progress
Rake's -q option keeps Rake's own output out of the stream.
To always report in TAP-Y or TAP-J, say so in the test helper:
require 'minitest/autorun'
require 'minitap'
Minitap.tapy! # or Minitap.tapj!
An option on the command line still takes precedence. Minitap.format = :tapj
does the same as Minitap.tapj!, which is handy with a variable:
Minitap.format = ENV['TAP_FORMAT'] if ENV['TAP_FORMAT']
As a shorthand, require 'minitap/autorun' requires minitest/autorun
and chooses TAP-Y:
require 'minitap/autorun'
Minitap is an ordinary Minitest plugin, so Minitest's own ways of
loading one work as well as require 'minitap':
Minitest.load :minitap
Minitest 5 loads every installed plugin by itself, so there the options are available without anything in the test helper. Minitest 6 no longer does that, and one of the lines above is needed.
- Each test class is a case, and each test a test, labelled by name. A spec is labelled as it was written.
- A failure or error comes with its message, the file and line in your code, the lines of source around it, and a backtrace without Minitest in it.
- A skipped test is reported as omitted, with the reason.
- What a test prints to
$stdoutand$stderris captured and reported with the test, since it would otherwise break the stream.
Version 0.6 uses Minitest's plugin system only, and no longer needs the
minitest-reporter-api gem.
Minitest.reporter = Minitap::TapYno longer does anything. UseMinitap.tapy!instead.- The reporters are
Minitap::TapYandMinitap::TapJ. They are no longer also defined asMinitest::TapYandMinitest::TapJ. - Skipped tests have the status
omit, as TAP-Y/J specifies. They used to have the statusskip.
For Minitest 4 and older, use Minitap 0.4.
Because Tapout reads Minitest's output through a pipe, anything in the
stream that isn't TAP-Y/J may make it halt. A debugger needs the
terminal, though. For that there is a pause code (16.chr), which
tells tapout to stop reading test results and pass everything
straight through, and a resume code (23.chr), which starts it
reading again.
With binding.pry, for example:
require 'tapout/utility'
tapout.pause{ binding.pry }
Or, more convenient still:
require 'tapout/utility'
tapout.pry(binding)
Which amounts to:
STDOUT.puts 16.chr
binding.pry
STDOUT.puts 23.chr
Copyright (c) 2011 Rubyworks
Made available under the terms of the BSD-2-Clause license.
Portions of this program were drawn from Alexander Kern's MiniTest Reporter (c) 2011 Alexander Kern. Thanks Alexander for making my endeavor a little bit easier.
See LICENSE.txt and NOTICE.md for copyright and licensing details.