vector2d 2.3.0 → 3.0.0

data/spec/lib/vector2d/comparison_spec.rb

@@ -0,0 +1,358 @@
+# frozen_string_literal: true
+
+require "spec_helper"
+
+describe Vector2d::Comparison do
+  subject(:vector) { Vector2d.new(2, 3) }
+
+  describe "#approx_equal?" do
+    subject { vector.approx_equal?(other) }
+
+    context "when the vectors are equal" do
+      let(:other) { Vector2d.new(2, 3) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when the vectors differ by a rounding error" do
+      let(:vector) { Vector2d.new(0.1, 0.2) + Vector2d.new(0.2, 0.4) }
+      let(:other) { Vector2d.new(0.3, 0.6) }
+
+      it { is_expected.to be(true) }
+
+      it "is more than #== accepts" do
+        expect(vector).not_to eq(other)
+      end
+    end
+
+    context "when the vectors differ" do
+      let(:other) { Vector2d.new(2, 4) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "when the other vector is coercible" do
+      let(:other) { [2, 3] }
+
+      it { is_expected.to be(true) }
+    end
+
+    it "is symmetric" do
+      other = Vector2d.new(0.1, 0.2) + Vector2d.new(0.2, 0.4)
+
+      expect(Vector2d.new(0.3, 0.6)).to be_approx_equal(other)
+    end
+
+    it "scales the tolerance with the magnitudes" do
+      big = Vector2d.new(1e8, 2e8)
+
+      expect(big).to be_approx_equal(big.rotate(2 * Math::PI))
+    end
+
+    it "doesn't accept the same drift at unit scale" do
+      expect(Vector2d.new(1, 2))
+        .not_to be_approx_equal(Vector2d.new(1, 2 + 5.5e-8))
+    end
+
+    context "with a tolerance" do
+      it "takes it as an absolute distance" do
+        expect(vector).to be_approx_equal(Vector2d.new(2, 4), 1.5)
+      end
+
+      it "is false outside the tolerance" do
+        expect(vector).not_to be_approx_equal(Vector2d.new(2, 4), 0.5)
+      end
+
+      it "includes the tolerance itself" do
+        expect(vector).to be_approx_equal(Vector2d.new(2, 4), 1.0)
+      end
+
+      it "raises an error on a complex tolerance" do
+        expect { vector.approx_equal?(vector, Complex(1, 2)) }.to(
+          raise_error(ArgumentError, "not a valid coordinate: (1+2i)")
+        )
+      end
+    end
+  end
+
+  describe "#parallel?" do
+    subject { vector.parallel?(other) }
+
+    context "when the other vector points the same way" do
+      let(:other) { Vector2d.new(4, 6) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when the other vector points the opposite way" do
+      let(:other) { Vector2d.new(-4, -6) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when the vectors aren't parallel" do
+      let(:other) { Vector2d.new(3, 2) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "with the zero vector" do
+      let(:other) { Vector2d.new(0, 0) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when both vectors are zero" do
+      let(:vector) { Vector2d.new(0, 0) }
+      let(:other) { Vector2d.new(0, 0) }
+
+      it { is_expected.to be(true) }
+    end
+
+    it "coerces the argument" do
+      expect(vector.parallel?("4x6")).to be(true)
+    end
+
+    it "is true for any vector and a multiple of it" do
+      vectors = (1..50).flat_map do |x|
+        (1..50).map { |y| Vector2d.new(x, y) }
+      end
+
+      expect(vectors).to all(satisfy { |v| v.parallel?(v * 3.7) })
+    end
+
+    context "with large vectors" do
+      let(:vector) { Vector2d.new(1.3e8, 7.7e8) }
+
+      it "is true for a multiple of itself" do
+        expect(vector.parallel?(vector * Math::PI)).to be(true)
+      end
+
+      it "is false for a vector that is barely off" do
+        expect(vector.parallel?(Vector2d.new(1.3e8, 7.7e8 + 1))).to be(false)
+      end
+    end
+  end
+
+  describe "#independent?" do
+    subject { vector.independent?(other) }
+
+    context "when the vectors aren't parallel" do
+      let(:other) { Vector2d.new(3, 2) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when the other vector points the same way" do
+      let(:other) { Vector2d.new(4, 6) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "with the zero vector" do
+      let(:other) { Vector2d.new(0, 0) }
+
+      it { is_expected.to be(false) }
+    end
+  end
+
+  describe "#opposite?" do
+    subject { vector.opposite?(other) }
+
+    context "when the other vector points the opposite way" do
+      let(:other) { Vector2d.new(-2, -3) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when the other vector is a negative multiple" do
+      let(:other) { Vector2d.new(-4, -6) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when the other vector points the same way" do
+      let(:other) { Vector2d.new(4, 6) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "when the vectors aren't parallel" do
+      let(:other) { Vector2d.new(3, 2) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "when the other vector is perpendicular" do
+      let(:other) { Vector2d.new(-3, 2) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "with the zero vector" do
+      let(:other) { Vector2d.new(0, 0) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "when the receiver is zero" do
+      let(:vector) { Vector2d.new(0, 0) }
+      let(:other) { Vector2d.new(2, 3) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "when both vectors are zero" do
+      let(:vector) { Vector2d.new(0, 0) }
+      let(:other) { Vector2d.new(0, 0) }
+
+      it { is_expected.to be(false) }
+    end
+
+    it "coerces the argument" do
+      expect(vector.opposite?("-2x-3")).to be(true)
+    end
+
+    it "ignores the magnitude of the vectors" do
+      expect(vector.opposite?(vector * -1e9)).to be(true)
+    end
+
+    it "is true for any vector and a negative multiple of it" do
+      vectors = (1..50).flat_map do |x|
+        (1..50).map { |y| Vector2d.new(x, y) }
+      end
+
+      expect(vectors).to all(satisfy { |v| v.opposite?(v * -3.7) })
+    end
+
+    it "is true for the reverse of the vector" do
+      expect(vector.opposite?(vector.reverse)).to be(true)
+    end
+
+    context "with large vectors" do
+      let(:vector) { Vector2d.new(1.3e8, 7.7e8) }
+
+      it "is true for a negative multiple of itself" do
+        expect(vector.opposite?(vector * -Math::PI)).to be(true)
+      end
+
+      it "is false for a vector that is barely off" do
+        expect(vector.opposite?(Vector2d.new(-1.3e8, -7.7e8 - 1))).to be(false)
+      end
+    end
+  end
+
+  describe "#perpendicular?" do
+    subject { vector.perpendicular?(other) }
+
+    context "when the other vector is perpendicular" do
+      let(:other) { Vector2d.new(-3, 2) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when the other vector is perpendicular the other way" do
+      let(:other) { Vector2d.new(3, -2) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when the vectors aren't perpendicular" do
+      let(:other) { Vector2d.new(3, 2) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "with the zero vector" do
+      let(:other) { Vector2d.new(0, 0) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when both vectors are zero" do
+      let(:vector) { Vector2d.new(0, 0) }
+      let(:other) { Vector2d.new(0, 0) }
+
+      it { is_expected.to be(true) }
+    end
+
+    it "coerces the argument" do
+      expect(vector.perpendicular?("-3x2")).to be(true)
+    end
+
+    it "is true for any vector rotated a quarter turn" do
+      vectors = (1..50).flat_map do |x|
+        (1..50).map { |y| Vector2d.new(x, y) }
+      end
+
+      expect(vectors)
+        .to all(satisfy { |v| v.perpendicular?(v.rotate(Math::PI / 2)) })
+    end
+
+    context "with large vectors" do
+      let(:vector) { Vector2d.new(1.3e8, 7.7e8) }
+
+      it "is true for a quarter turn" do
+        expect(vector.perpendicular?(vector.rotate(Math::PI / 2)))
+          .to be(true)
+      end
+
+      it "is false for a vector that is barely off" do
+        expect(vector.perpendicular?(Vector2d.new(-7.7e8, 1.3e8 + 1)))
+          .to be(false)
+      end
+    end
+  end
+
+  describe "#finite?" do
+    subject { vector.finite? }
+
+    context "when both coordinates are finite" do
+      it { is_expected.to be(true) }
+    end
+
+    context "when x is infinite" do
+      let(:vector) { Vector2d.new(Float::INFINITY, 3) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "when y is infinite" do
+      let(:vector) { Vector2d.new(2, -Float::INFINITY) }
+
+      it { is_expected.to be(false) }
+    end
+
+    context "when a coordinate is NaN" do
+      let(:vector) { Vector2d.new(2, Float::NAN) }
+
+      it { is_expected.to be(false) }
+    end
+  end
+
+  describe "#nan?" do
+    subject { vector.nan? }
+
+    context "when neither coordinate is NaN" do
+      it { is_expected.to be(false) }
+    end
+
+    context "when x is NaN" do
+      let(:vector) { Vector2d.new(Float::NAN, 3) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when y is NaN" do
+      let(:vector) { Vector2d.new(2, Float::NAN) }
+
+      it { is_expected.to be(true) }
+    end
+
+    context "when a coordinate is infinite" do
+      let(:vector) { Vector2d.new(2, Float::INFINITY) }
+
+      it { is_expected.to be(false) }
+    end
+  end
+end

data/spec/lib/vector2d/componentwise_spec.rb

@@ -0,0 +1,300 @@
+# frozen_string_literal: true
+
+require "spec_helper"
+
+describe Vector2d::Componentwise do
+  subject(:vector) { Vector2d.new(2, 3) }
+
+  describe "#abs" do
+    it "returns the absolute value of each axis" do
+      expect(Vector2d.new(-2, 3).abs).to eq(Vector2d.new(2, 3))
+    end
+
+    it "leaves a positive vector alone" do
+      expect(vector.abs).to eq(vector)
+    end
+
+    it "is component-wise, not the length" do
+      expect(Vector2d.new(-2, -3).abs).to eq(Vector2d.new(2, 3))
+    end
+  end
+
+  describe "#ceil" do
+    subject(:vector) { Vector2d.new(2.3, 3.3) }
+
+    it "rounds the vector up" do
+      expect(vector.ceil).to eq(Vector2d.new(3, 4))
+    end
+
+    it "rounds up to the given precision" do
+      expect(Vector2d.new(2.441, 3.666).ceil(2))
+        .to eq(Vector2d.new(2.45, 3.67))
+    end
+
+    it "rounds up to a negative precision" do
+      expect(Vector2d.new(24.4, 36.6).ceil(-1)).to eq(Vector2d.new(30, 40))
+    end
+  end
+
+  describe "#floor" do
+    subject(:vector) { Vector2d.new(2.7, 3.6) }
+
+    it "rounds the vector down" do
+      expect(vector.floor).to eq(Vector2d.new(2, 3))
+    end
+
+    it "rounds down to the given precision" do
+      expect(Vector2d.new(2.444, 3.669).floor(2))
+        .to eq(Vector2d.new(2.44, 3.66))
+    end
+
+    it "rounds down to a negative precision" do
+      expect(Vector2d.new(24.4, 36.6).floor(-1)).to eq(Vector2d.new(20, 30))
+    end
+  end
+
+  describe "#round" do
+    let(:vector) { Vector2d.new(2.3333, 3.666) }
+
+    context "without argument" do
+      subject { vector.round }
+
+      it { is_expected.to eq(Vector2d.new(2, 4)) }
+    end
+
+    context "with precision" do
+      subject { vector.round(2) }
+
+      it { is_expected.to eq(Vector2d.new(2.33, 3.67)) }
+    end
+  end
+
+  describe "#truncate" do
+    subject(:vector) { Vector2d.new(2, 3).truncate(2.5) }
+
+    it_behaves_like "a deprecated method", "truncate", "#limit_length"
+
+    it "limits the length, as #limit_length does" do
+      expect(vector.length).to be_within(0.0001).of(2.5)
+    end
+  end
+
+  describe "#sign" do
+    it "returns the sign of each axis" do
+      expect(Vector2d.new(-2, 3).sign).to eq(Vector2d.new(-1, 1))
+    end
+
+    it "is zero for an axis that is zero" do
+      expect(Vector2d.new(0, -3).sign).to eq(Vector2d.new(0, -1))
+    end
+
+    it "returns integers for a float vector" do
+      expect(Vector2d.new(-2.5, 0.0).sign.to_a).to eq([-1, 0])
+    end
+
+    it "treats negative zero as zero" do
+      expect(Vector2d.new(-0.0, 1).sign).to eq(Vector2d.new(0, 1))
+    end
+
+    context "when a coordinate is NaN" do
+      it "raises an error" do
+        expect { Vector2d.new(Float::NAN, 3).sign }.to(
+          raise_error(ArgumentError, "NaN has no sign")
+        )
+      end
+    end
+  end
+
+  describe "#snap" do
+    subject(:vector) { Vector2d.new(23, 47) }
+
+    it "snaps each axis to the nearest multiple" do
+      expect(vector.snap(10)).to eq(Vector2d.new(20, 50))
+    end
+
+    it "takes a step per axis" do
+      expect(vector.snap(Vector2d.new(10, 5))).to eq(Vector2d.new(20, 45))
+    end
+
+    it "coerces the step" do
+      expect(vector.snap("10x5")).to eq(Vector2d.new(20, 45))
+    end
+
+    it "leaves a vector on the grid alone" do
+      expect(Vector2d.new(20, 50).snap(10)).to eq(Vector2d.new(20, 50))
+    end
+
+    it "snaps negative coordinates away from zero at the halfway point" do
+      expect(Vector2d.new(-15, -25).snap(10)).to eq(Vector2d.new(-20, -30))
+    end
+
+    it "handles a negative step" do
+      expect(vector.snap(-10)).to eq(Vector2d.new(20, 50))
+    end
+
+    context "with a fractional step" do
+      subject(:vector) { Vector2d.new(2.3, 3.7) }
+
+      it "snaps to the fractional grid" do
+        expect(vector.snap(0.5)).to eq(Vector2d.new(2.5, 3.5))
+      end
+
+      it "keeps the type of the step" do
+        expect(vector.snap(1)).to eql(Vector2d.new(2, 4))
+      end
+    end
+
+    context "when the step is zero" do
+      it "leaves both axes unchanged" do
+        expect(vector.snap(0)).to eq(vector)
+      end
+
+      it "leaves only the unstepped axis unchanged" do
+        expect(vector.snap(Vector2d.new(10, 0))).to eq(Vector2d.new(20, 47))
+      end
+    end
+  end
+
+  describe "#max" do
+    subject(:vector) { Vector2d.new(2, 8) }
+
+    it "returns the larger value of each axis" do
+      expect(vector.max(Vector2d.new(5, 5))).to eq(Vector2d.new(5, 8))
+    end
+
+    it "coerces the argument" do
+      expect(vector.max(5)).to eq(Vector2d.new(5, 8))
+    end
+
+    it "is symmetric" do
+      other = Vector2d.new(5, 5)
+      expect(other.max(vector)).to eq(vector.max(other))
+    end
+  end
+
+  describe "#min" do
+    subject(:vector) { Vector2d.new(2, 8) }
+
+    it "returns the smaller value of each axis" do
+      expect(vector.min(Vector2d.new(5, 5))).to eq(Vector2d.new(2, 5))
+    end
+
+    it "coerces the argument" do
+      expect(vector.min(5)).to eq(Vector2d.new(2, 5))
+    end
+
+    it "is symmetric" do
+      other = Vector2d.new(5, 5)
+      expect(other.min(vector)).to eq(vector.min(other))
+    end
+  end
+
+  describe "#clamp" do
+    subject(:vector) { Vector2d.new(2, 8) }
+
+    it "clamps each axis" do
+      expect(vector.clamp(Vector2d.new(3, 3), Vector2d.new(6, 6)))
+        .to eq(Vector2d.new(3, 6))
+    end
+
+    it "coerces the bounds" do
+      expect(vector.clamp(3, 6)).to eq(Vector2d.new(3, 6))
+    end
+
+    it "leaves a vector within the bounds alone" do
+      expect(vector.clamp(0, 10)).to eq(vector)
+    end
+
+    it "raises ArgumentError without an upper bound" do
+      expect { vector.clamp(3) }.to raise_error(ArgumentError)
+    end
+
+    context "with a range" do
+      it "clamps each axis" do
+        expect(vector.clamp(3..6)).to eq(Vector2d.new(3, 6))
+      end
+
+      it "coerces the bounds" do
+        expect(vector.clamp("3x3".."6x6")).to eq(Vector2d.new(3, 6))
+      end
+
+      it "leaves a vector within the bounds alone" do
+        expect(vector.clamp(0..10)).to eq(vector)
+      end
+
+      it "matches the two argument form" do
+        expect(vector.clamp(3..6)).to eq(vector.clamp(3, 6))
+      end
+
+      it "clamps only the upper bound of a beginless range" do
+        expect(vector.clamp(..6)).to eq(Vector2d.new(2, 6))
+      end
+
+      it "clamps only the lower bound of an endless range" do
+        expect(vector.clamp(3..)).to eq(Vector2d.new(3, 8))
+      end
+
+      it "leaves a vector bounded at neither end alone" do
+        expect(vector.clamp(nil..nil)).to eq(vector)
+      end
+
+      it "matches #min for a beginless range" do
+        expect(vector.clamp(..6)).to eq(vector.min(6))
+      end
+
+      it "matches #max for an endless range" do
+        expect(vector.clamp(3..)).to eq(vector.max(3))
+      end
+
+      it "raises ArgumentError for an exclusive range" do
+        expect { vector.clamp(3...6) }.to raise_error(ArgumentError)
+      end
+
+      it "raises ArgumentError when given a second argument" do
+        expect { vector.clamp(3..6, 6) }.to raise_error(ArgumentError)
+      end
+    end
+  end
+
+  describe "#with_x" do
+    it "replaces x" do
+      expect(vector.with_x(5)).to eq(Vector2d.new(5, 3))
+    end
+
+    it "keeps the type of the new coordinate" do
+      expect(vector.with_x(5.0).x).to be_a(Float)
+    end
+
+    it "leaves the receiver alone" do
+      expect { vector.with_x(5) }.not_to change(vector, :to_a)
+    end
+
+    it "raises an error on a string" do
+      expect { vector.with_x("5") }.to(
+        raise_error(ArgumentError, 'not a valid coordinate: "5"')
+      )
+    end
+
+    it "raises an error on nil" do
+      expect { vector.with_x(nil) }.to(
+        raise_error(ArgumentError, "not a valid coordinate: nil")
+      )
+    end
+  end
+
+  describe "#with_y" do
+    it "replaces y" do
+      expect(vector.with_y(5)).to eq(Vector2d.new(2, 5))
+    end
+
+    it "keeps the type of the new coordinate" do
+      expect(vector.with_y(5.0).y).to be_a(Float)
+    end
+
+    it "raises an error on a vector" do
+      expect { vector.with_y(Vector2d.new(2, 3)) }.to(
+        raise_error(ArgumentError, "not a valid coordinate: Vector2d(2,3)")
+      )
+    end
+  end
+end