CODEWITHSUNDEEP
rubytestingunicodeutf-8minitest
fl00r
fl00r

Reputation: 83680

MiniTest and non-Latin test description

Ruby 1.9 has got cool Unicode support, yeah?

# encoding: utf-8
require 'minitest/spec'
require 'minitest/autorun'

describe "test" do
  it "α β γ δ & a b c d" do
    (1+1).must_equal 3
  end
end

# 1) Failure:
# test_0001__a_b_c_d(TestSpec) [test.rb:7]:
# Expected 3, not 2.

Where are my non-Latin letters? I should always write my tests in my horrible English?

Because I can define methods with any Unicode symbol:

def α_β_γ_δ_a_b_c_d
  puts "geeeek"
end

α_β_γ_δ_a_b_c_d
#=> "geeeek"

PS My question seems to be not clear. I want to ask how to make minitest's failure description to show my non latin definitions.

Upvotes: 2

Views: 242

Answers (1)

Selman Ulug
Selman Ulug

Reputation: 867

it is about regexp used here. it shows utf-8 characters after monkey patching /\W+/ with /\s+/.

# encoding: utf-8
require 'minitest/spec'
require 'minitest/autorun'

class MiniTest::Spec < MiniTest::Unit::TestCase
  def self.it desc = "anonymous", &block
    block ||= proc { skip "(no tests defined)" }

    @specs ||= 0
    @specs += 1

    # regexp /\W+/ replaced with /\s+/
    name = "test_%04d_%s" % [ @specs, desc.gsub(/\s+/, '_').downcase ]

    define_method name, &block

    self.children.each do |mod|
      mod.send :undef_method, name if mod.public_method_defined? name
    end
  end
end

describe "test" do
  it "α β γ δ & a b c D" do
    (1+1).must_equal 3
  end
end

#   1) Failure:
# test_0001_α_β_γ_δ_&_a_b_c_d(test) [forwarding.rb:24]:
# Expected: 3
#   Actual: 2

Upvotes: 2

Related Questions