Backport #4200: minitest 2.0.2 on trunk

• Status: Closed
• Assignee: yugui (Yuki Sonoda)

[zenspider (Ryan Davis)]

=begin
Please merge minitest 2.0.2 on trunk to 1_9_2. This requires the test/unit wrapper to me merged as well, if it hasn't been already.

Paths:

lib/minitest/*
lib/test/unit.rb
lib/test/unit/*
test/minitest/*
=end

[zenspider (Ryan Davis)]

=begin
May I do this?
=end

[nobu (Nobuyoshi Nakada)]

• Status changed from Open to Assigned
• Assignee set to yugui (Yuki Sonoda)

=begin

=end

[yugui (Yuki Sonoda)]

=begin
On Wed, Jan 5, 2011 at 2:01 PM, Ryan Davis [e-mail] wrote:

May I do this?

I don't want merge any enhancement into ruby_1_9_2. So I hesitated to
merge minitest 2.0.2. Sorry for replying so late.
I will merge only bug fixes later. Or you can do it by yourself.

--
Yuki Sonoda (Yugui)
[e-mail]
http://yugui.jp

=end

[zenspider (Ryan Davis)]

=begin

On Jan 5, 2011, at 00:41 , Yugui wrote:

On Wed, Jan 5, 2011 at 2:01 PM, Ryan Davis [e-mail] wrote:

May I do this?

I don't want merge any enhancement into ruby_1_9_2. So I hesitated to
merge minitest 2.0.2. Sorry for replying so late.
I will merge only bug fixes later. Or you can do it by yourself.

Lucas? Is that you??

Enhancements are mostly in separate files. Leave out minitest/benchmark.rb and that's most of it. I did add some assertions since 1.6.0 (2010-03-27) but it is mostly bug fixes and refactorings.

Here are relevant changes:

=== 2.0.2 / 2010-12-24

• 1 minor enhancement:

  • Completed doco on minitest/benchmark for specs.

• 1 bug fix:

  • Benchmarks in specs that didn't call bench_range would die. (zzak).

=== 2.0.1 / 2010-12-15

• 4 minor enhancements:

  • Do not filter backtrace if $DEBUG
  • Exit autorun via nested at_exit handler, in case other libs call exit
  • Make options accesor lazy.
  • Split printing of test name and its time. (nurse)

• 1 bug fix:

  • Fix bug when ^T is hit before runner start

=== 2.0.0 / 2010-11-11

• 3 major enhancements:

  • Added minitest/benchmark! Assert your performance! YAY!
  • Refactored runner to allow for more extensibility. See minitest/benchmark.
  • This makes the runner backwards incompatible in some ways!

• 9 minor enhancements:

  • Added MiniTest::Unit.after_tests { ... }
  • Improved output by adding test rates and a more sortable verbose format
  • Improved readme based on feedback from others
  • Added io method to TestCase. If used, it'll supplant '.EF' output.
  • Refactored IO in MiniTest::Unit.
  • Refactored _run_anything to _run_suite to make it easier to wrap (ngauthier)
  • Spec class names are now the unmunged descriptions (btakita)
  • YAY for not having redundant rdoc/readmes!
  • Help output is now generated from the flags you passed straight up.

• 4 bug fixes:

  • Fixed scoping issue on minitest/mock (srbaker/prosperity)
  • Fixed some of the assertion default messages
  • Fixes autorun when on windows with ruby install on different drive (larsch)
  • Fixed rdoc output bug in spec.rb

=== 1.7.2 / 2010-09-23

• 3 bug fixes:

  • Fixed doco for expectations and Spec.
  • Fixed test_capture_io on 1.9.3+ (sora_h)
  • assert_raises now lets MiniTest::Skip through. (shyouhei)

=== 1.7.1 / 2010-09-01

• 1 bug fix:

  • 1.9.2 fixes for spec tests

=== 1.7.0 / 2010-07-15

• 5 minor enhancements:

  • Added assert_output (mapped to must_output).
  • Added assert_silent (mapped to must_be_silent).
  • Added examples to readme (Mike Dalessio)
  • Added options output at the top of the run, for fatal run debugging (tenderlove)
  • Spec's describe method returns created class

=end

[yugui (Yuki Sonoda)]

=begin
Hi Ryan,

I have merged bug fixes in MiniTest. It is harder than what I thought.

• I did not import lib/minitest/benchmark.rb because it is a new feature.

• I suspect the change in lib/minitest/mock.rb is to merge.
  I agree that the minitest/mock.rb before 5916 had had a wrong
  design. But 5916 changes minitest/mock's behavior anyway.

• I merged a part of changes for test/minitest/test_mini_mock.rb.
  I did not merge the rest of the changes which depends on 5916.

• I did not import lib/minitest/pride.rb because it seems a new feature.

• I did not import lib/minitest/spec.rb because its change consits of:

  • minor feature change in Module#infect_with_assetions (sorting)
  • debug print? (Module#infect_with_assertions)
  • refactoring (relocation of calling #infect_with_assertions)
  • following a new feature (calling #infact_with_assertions for assert_silent)
  • feature change (Kernel#describe)
  • adding rdocs (Module#infect_with_assertions, MiniTest::Spec, ...)

  I do not merge any rdoc change into ruby_1_9_2 except fixing a
  terribly confusing description.

• MiniTest::Unit

  • I did not merge 5783 because it only adds new features.
  • I did not merge 5785 because it only updated the version. I think
    we should not update the version because I never add new feature into
    ruby_1_9_2 so minitest in ruby_1_9_2 is not 1.7.0.
  • I did not merge 5835 because of the same reason.
  • I merged a part of 5875 because it is a bug. Once I decided not to
    merge shyouhei's fix because I regarded it as a limitation of minitest
    1.6.0 that MiniTest::Skip broke the assersion. It was a wrong
    decision.
  • I did not merge 5877 because it just updated the version.
  • I did not merge 5884 because it changed behavior.
  • I did not merge 5886 because it is an enhancement and a refactoring.
  • I did not merge 5889 because it is a fix for a bug in 5886.
  • I did not merge 5905 because it seemed something related to 5886.
  • I did not merge 5906 because it updated the version.
  • I did not merge 5908 because it added a new feature.
  • I did not merge 5911 because it was a refactoring related to 5908.
  • I did not merge 5913 because it was a refactoring related to 5908.
  • I did not merge 5918 because it only affects Ruby 1.8 or eariler.
    By the way, I think the condition should be RUBY_VERSION >= "1.9".
  • I did not merge 5922 because it was a refactoring.
  • I did not merge 5923 because it contained only style changes.
  • I did not merge 5924 because it was a refactoring.
  • I did not merge 5926 because it was a refactoring + adding after_tests{ }.
  • I merged only a part of 5937 which fixed some messages.
  • I did not merge 5939 because it was a typo.
  • I did not merge 5944 because there is not README.txt for MiniTest
    in Ruby's repository.
  • I did not merge 5945 because it was a rdoc fix.
  • I did not merge 5948 because I did not merge the latter half of 5937.
  • I did not merge 5951 because it just updated the version.
  • I did not merge 6056 becuase it is a minor feature change and only
    affects on $DEBUG == true
  • I merged 5907
  • I did not merge 6058 because I did not merge 5901.
  • I did not merge 6059 because I did not merge 5901.
  • I did not merge 6072 because it was a minor improvement, not a bug.
  • I did not merge 6079 because it just updated the version.
  • I did not merge 6093 becuase of the same reason.
  • I did not merge 6095 because it is for Ruby 1.9.3.

--
Yuki Sonoda (Yugui)
[e-mail]
http://yugui.jp

=end

[yugui (Yuki Sonoda)]

• Status changed from Assigned to Closed

=begin

=end

[jonforums (Jon Forums)]

=begin
Could the differences between MiniTest 2.0.2 merged into ruby_1_9_2 and MiniTest 2.0.2 available as a standalone RubyGem cause issues similar to RubyGems 1.3.7?

Jon
=end