HEX
Server: Apache
System: Linux s198.coreserver.jp 5.15.0-151-generic #161-Ubuntu SMP Tue Jul 22 14:25:40 UTC 2025 x86_64
User: nagasaki (10062)
PHP: 7.1.33
Disabled: NONE
Upload Files
File: //usr/local/rvm/src/ruby-2.7.4/spec/ruby/library/bigdecimal/BigDecimal_spec.rb
require_relative '../../spec_helper'
require 'bigdecimal'

describe "BigDecimal" do
  it "is not defined unless it is required" do
    ruby_exe('puts Object.const_defined?(:BigDecimal)').should == "false\n"
  end
end

describe "Kernel#BigDecimal" do

  it "creates a new object of class BigDecimal" do
    BigDecimal("3.14159").should be_kind_of(BigDecimal)
    (0..9).each {|i|
      BigDecimal("1#{i}").should == 10 + i
      BigDecimal("-1#{i}").should == -10 - i
      BigDecimal("1E#{i}").should == 10**i
      BigDecimal("1000000E-#{i}").should == 10**(6-i).to_f
      # ^ to_f to avoid Rational type
    }
    (1..9).each {|i|
      BigDecimal("100.#{i}").to_s.should =~ /\A0\.100#{i}E3\z/i
      BigDecimal("-100.#{i}").to_s.should =~ /\A-0\.100#{i}E3\z/i
    }
  end

  it "BigDecimal(Rational) with bigger-than-double numerator" do
    rational = 99999999999999999999/100r
    rational.numerator.should > 2**64
    BigDecimal(rational, 100).to_s.should == "0.99999999999999999999e18"
  end

  it "accepts significant digits >= given precision" do
    BigDecimal("3.1415923", 10).precs[1].should >= 10
  end

  it "determines precision from initial value" do
    pi_string = "3.14159265358979323846264338327950288419716939937510582097494459230781640628620899862803482534211706798214808651328230664709384460955058223172535940812848111745028410270193852110555964462294895493038196442881097566593014782083152134043"
    BigDecimal(pi_string).precs[1].should >= pi_string.size-1
  end

  it "ignores leading and trailing whitespace" do
    BigDecimal("  \t\n \r1234\t\r\n ").should == BigDecimal("1234")
    BigDecimal("  \t\n \rNaN   \n").nan?.should == true
    BigDecimal("  \t\n \rInfinity   \n").infinite?.should == 1
    BigDecimal("  \t\n \r-Infinity   \n").infinite?.should == -1
  end

  ruby_version_is ""..."2.6" do
    it "ignores trailing garbage" do
      BigDecimal("123E45ruby").should == BigDecimal("123E45")
      BigDecimal("123x45").should == BigDecimal("123")
      BigDecimal("123.4%E5").should == BigDecimal("123.4")
      BigDecimal("1E2E3E4E5E").should == BigDecimal("100")
    end
  end

  ruby_version_is "2.6" do
    it "does not ignores trailing garbage" do
      -> { BigDecimal("123E45ruby") }.should raise_error(ArgumentError)
      -> { BigDecimal("123x45") }.should raise_error(ArgumentError)
      -> { BigDecimal("123.4%E5") }.should raise_error(ArgumentError)
      -> { BigDecimal("1E2E3E4E5E") }.should raise_error(ArgumentError)
    end
  end

  it "raises ArgumentError for invalid strings" do
    -> { BigDecimal("ruby") }.should raise_error(ArgumentError)
    -> { BigDecimal("  \t\n \r-\t\t\tInfinity   \n") }.should raise_error(ArgumentError)
  end

  it "allows omitting the integer part" do
    BigDecimal(".123").should == BigDecimal("0.123")
  end

  ruby_version_is ""..."2.6" do
    it "allows for underscores in all parts" do
      reference = BigDecimal("12345.67E89")

      BigDecimal("12_345.67E89").should == reference
      BigDecimal("1_2_3_4_5_._6____7_E89").should == reference
      BigDecimal("12345_.67E_8__9_").should == reference
    end
  end

  ruby_version_is "2.6" do
    it "process underscores as Float()" do
      reference = BigDecimal("12345.67E89")

      BigDecimal("12_345.67E89").should == reference
      -> { BigDecimal("1_2_3_4_5_._6____7_E89") }.should raise_error(ArgumentError)
      -> { BigDecimal("12345_.67E_8__9_") }.should raise_error(ArgumentError)
    end
  end

  it "accepts NaN and [+-]Infinity" do
    BigDecimal("NaN").nan?.should == true

    pos_inf = BigDecimal("Infinity")
    pos_inf.finite?.should == false
    pos_inf.should > 0
    pos_inf.should == BigDecimal("+Infinity")

    neg_inf = BigDecimal("-Infinity")
    neg_inf.finite?.should == false
    neg_inf.should < 0
  end

  describe "accepts NaN and [+-]Infinity as Float values" do
    it "works without an explicit precision" do
      BigDecimal(Float::NAN).nan?.should == true

      pos_inf = BigDecimal(Float::INFINITY)
      pos_inf.finite?.should == false
      pos_inf.should > 0
      pos_inf.should == BigDecimal("+Infinity")

      neg_inf = BigDecimal(-Float::INFINITY)
      neg_inf.finite?.should == false
      neg_inf.should < 0
    end

    it "works with an explicit precision" do
      BigDecimal(Float::NAN, Float::DIG).nan?.should == true

      pos_inf = BigDecimal(Float::INFINITY, Float::DIG)
      pos_inf.finite?.should == false
      pos_inf.should > 0
      pos_inf.should == BigDecimal("+Infinity")

      neg_inf = BigDecimal(-Float::INFINITY, Float::DIG)
      neg_inf.finite?.should == false
      neg_inf.should < 0
    end
  end

  it "allows for [eEdD] as exponent separator" do
    reference = BigDecimal("12345.67E89")

    BigDecimal("12345.67e89").should == reference
    BigDecimal("12345.67E89").should == reference
    BigDecimal("12345.67d89").should == reference
    BigDecimal("12345.67D89").should == reference
  end

  it "allows for varying signs" do
    reference = BigDecimal("123.456E1")

    BigDecimal("+123.456E1").should == reference
    BigDecimal("-123.456E1").should == -reference
    BigDecimal("123.456E+1").should == reference
    BigDecimal("12345.6E-1").should == reference
    BigDecimal("+123.456E+1").should == reference
    BigDecimal("+12345.6E-1").should == reference
    BigDecimal("-123.456E+1").should == -reference
    BigDecimal("-12345.6E-1").should == -reference
  end

  it "raises ArgumentError when Float is used without precision" do
    -> { BigDecimal(1.0) }.should raise_error(ArgumentError)
  end

  it "returns appropriate BigDecimal zero for signed zero" do
    BigDecimal(-0.0, Float::DIG).sign.should == -1
    BigDecimal(0.0, Float::DIG).sign.should == 1
  end

  it "pre-coerces long integers" do
    BigDecimal(3).add(1 << 50, 3).should == BigDecimal('0.113e16')
  end

  describe "when interacting with Rational" do
    before :each do
      @a = BigDecimal('166.666666666')
      @b = Rational(500, 3)
      @c = @a - @b
    end

    # Check the input is as we understand it

    it "has the LHS print as expected" do
      @a.to_s.should == "0.166666666666e3"
      @a.to_f.to_s.should == "166.666666666"
      Float(@a).to_s.should == "166.666666666"
    end

    it "has the RHS print as expected" do
      @b.to_s.should == "500/3"
      @b.to_f.to_s.should == "166.66666666666666"
      Float(@b).to_s.should == "166.66666666666666"
    end

    it "has the expected precision on the LHS" do
      @a.precs[0].should == 18
    end

    it "has the expected maximum precision on the LHS" do
      @a.precs[1].should == 27
    end

    it "produces the expected result when done via Float" do
      (Float(@a) - Float(@b)).to_s.should == "-6.666596163995564e-10"
    end

    it "produces the expected result when done via to_f" do
      (@a.to_f - @b.to_f).to_s.should == "-6.666596163995564e-10"
    end

    # Check underlying methods work as we understand

    it "BigDecimal precision is the number of digits rounded up to a multiple of nine" do
      1.upto(100) do |n|
        b = BigDecimal('4' * n)
        precs, _ = b.precs
        (precs >= 9).should be_true
        (precs >= n).should be_true
        (precs % 9).should == 0
      end
      BigDecimal('NaN').precs[0].should == 9
    end

    it "BigDecimal maximum precision is nine more than precision except for abnormals" do
      1.upto(100) do |n|
        b = BigDecimal('4' * n)
        precs, max = b.precs
        max.should == precs + 9
      end
      BigDecimal('NaN').precs[1].should == 9
    end

    it "BigDecimal(Rational, 18) produces the result we expect" do
      BigDecimal(@b, 18).to_s.should == "0.166666666666666667e3"
    end

    it "BigDecimal(Rational, BigDecimal.precs[0]) produces the result we expect" do
      BigDecimal(@b, @a.precs[0]).to_s.should == "0.166666666666666667e3"
    end

    # Check the top-level expression works as we expect

    it "produces a BigDecimal" do
      @c.class.should == BigDecimal
    end

    it "produces the expected result" do
      @c.should == BigDecimal("-0.666667e-9")
      @c.to_s.should == "-0.666667e-9"
    end

    it "produces the correct class for other arithmetic operators" do
      (@a + @b).class.should == BigDecimal
      (@a * @b).class.should == BigDecimal
      (@a / @b).class.should == BigDecimal
      (@a % @b).class.should == BigDecimal
    end
  end
end