perfect-shape 0.3.5 → 0.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: ab1f48d0551ed07a6a2b5f2d51abba6c6596eef82e156e68df57489319104e4c
-  data.tar.gz: bb67e137bbf23df2ed291a574ef45ae97113dd140d0a17269c9b60c27b7261e9
+  metadata.gz: 13f257899f946daccf27fa951ed4208ce8143d2060b24fcb6198ba0e8f738646
+  data.tar.gz: 1f9a0d7a1f210da0ee0e7b88514352b7d90c7893b05fe9b84145672989cd8eae
 SHA512:
-  metadata.gz: 6d365e47635c7daa094f02849df24696784b667c55bffa479bdbaacc41abf1b9ad1114084a4ed7bd0145e6dd18ad233f9b09e6bad8792828417aca49db2076c4
-  data.tar.gz: d8440bb0d35e1539dcf5c8cad79e2930d74e10e32a408391085027d773c0ea7e4272a6011a7244677c035c8f6a7954606065611282ae8cd14aab71cbbe6b2bba
+  metadata.gz: 2af7d835fdaf126b8ebb630fbc08400aab9a7b464c5d18ad3f09035f142940bb1db1fbe4c0a2a1bc33bf4a773f2ce920fb34e36d57a191143020cdd36357b483
+  data.tar.gz: 693a20f1d100e70a8d429c8c02fddcfbdb0b2f2507cff92a0ddbc8c962db7758be5cee94df954ba5a89bba7619bdd9b2ee95b31b17896289e31d90a22d7eeb53

data/CHANGELOG.md

@@ -1,5 +1,23 @@
 # Change Log
 
+## 0.4.0
+
+- `PerfectShape::AffineTransform#new`
+- `PerfectShape::AffineTransform#==`
+- `PerfectShape::AffineTransform#transform_point`
+- `PerfectShape::AffineTransform#transform_points`
+- `PerfectShape::AffineTransform#identity!` (alias: `reset!`)
+- `PerfectShape::AffineTransform#invert!`
+- `PerfectShape::AffineTransform#invertible?`
+- `PerfectShape::AffineTransform#multiply!`
+- `PerfectShape::AffineTransform#translate!`
+- `PerfectShape::AffineTransform#scale!`
+- `PerfectShape::AffineTransform#rotate!`
+- `PerfectShape::AffineTransform#shear!` (alias: `skew!`)
+- `PerfectShape::AffineTransform#clone`
+- `PerfectShape::AffineTransform#inverse_transform_point`
+- `PerfectShape::AffineTransform#inverse_transform_points`
+
 ## 0.3.5
 
 - Check point containment in composite shape outline with distance tolerance (new method signature: `PerfectShape::CompositeShape#contain?(x_or_point, y = nil, outline: false, distance_tolerance: 0)`)

data/README.md

@@ -1,9 +1,9 @@
-# Perfect Shape 0.3.5
+# Perfect Shape 0.4.0
 ## Geometric Algorithms
 [![Gem Version](https://badge.fury.io/rb/perfect-shape.svg)](http://badge.fury.io/rb/perfect-shape)
 [![Test](https://github.com/AndyObtiva/perfect-shape/actions/workflows/ruby.yml/badge.svg)](https://github.com/AndyObtiva/perfect-shape/actions/workflows/ruby.yml)
 
-[`PerfectShape`](https://rubygems.org/gems/perfect-shape) is a collection of pure Ruby geometric algorithms that are mostly useful for GUI (Graphical User Interface) manipulation like checking containment of a mouse click [point](#perfectshapepoint) in popular geometry shapes such as [rectangle](#perfectshaperectangle), [square](#perfectshapesquare), [arc](#perfectshapearc) (open, chord, and pie), [ellipse](#perfectshapeellipse), [circle](#perfectshapecircle), [polygon](#perfectshapepolygon), and [paths](#perfectshapepath) containing [lines](#perfectshapeline), [quadratic bézier curves](#perfectshapequadraticbeziercurve), and [cubic bezier curves](#perfectshapecubicbeziercurve) (including both [Ray Casting Algorithm](https://en.wikipedia.org/wiki/Point_in_polygon#Ray_casting_algorithm), aka [Even-odd Rule](https://en.wikipedia.org/wiki/Even%E2%80%93odd_rule), and [Winding Number Algorithm](https://en.wikipedia.org/wiki/Point_in_polygon#Winding_number_algorithm), aka [Nonzero Rule](https://en.wikipedia.org/wiki/Nonzero-rule)).
+[`PerfectShape`](https://rubygems.org/gems/perfect-shape) is a collection of pure Ruby geometric algorithms that are mostly useful for GUI (Graphical User Interface) manipulation like checking containment of a mouse click [point](#perfectshapepoint) in popular geometry shapes such as [rectangle](#perfectshaperectangle), [square](#perfectshapesquare), [arc](#perfectshapearc) (open, chord, and pie), [ellipse](#perfectshapeellipse), [circle](#perfectshapecircle), [polygon](#perfectshapepolygon), and [paths](#perfectshapepath) containing [lines](#perfectshapeline), [quadratic bézier curves](#perfectshapequadraticbeziercurve), and [cubic bezier curves](#perfectshapecubicbeziercurve), potentially with [affine transforms](#perfectshapeaffinetransform) applied like translation, scale, rotation, shear/skew, and inversion (including both [Ray Casting Algorithm](https://en.wikipedia.org/wiki/Point_in_polygon#Ray_casting_algorithm), aka [Even-odd Rule](https://en.wikipedia.org/wiki/Even%E2%80%93odd_rule), and [Winding Number Algorithm](https://en.wikipedia.org/wiki/Point_in_polygon#Winding_number_algorithm), aka [Nonzero Rule](https://en.wikipedia.org/wiki/Nonzero-rule)).
 
 Additionally, [`PerfectShape::Math`](#perfectshapemath) contains some purely mathematical algorithms, like [IEEE 754-1985 Remainder](https://en.wikipedia.org/wiki/IEEE_754-1985).
 
@@ -14,13 +14,13 @@ To ensure high accuracy, this library does all its mathematical operations with
 Run:
 
 ```
-gem install perfect-shape -v 0.3.5
+gem install perfect-shape -v 0.4.0
 ```
 
 Or include in Bundler `Gemfile`:
 
 ```ruby
-gem 'perfect-shape', '~> 0.3.5'
+gem 'perfect-shape', '~> 0.4.0'
 ```
 
 And, run:
@@ -57,7 +57,6 @@ This is a base class for all shapes. It is not meant to be used directly. Subcla
 - `#center_y`: center y
 - `#bounding_box`: bounding box is a rectangle with x = min x, y = min y, and width/height just as those of shape
 - `#==(other)`: Returns `true` if equal to `other` or `false` otherwise
-- `#normalize_point(x_or_point, y = nil)`: normalizes point into an `Array` of `[x,y]` coordinates
 - `#contain?(x_or_point, y=nil, outline: false, distance_tolerance: 0)`: checks if point is inside if `outline` is `false` or if point is on the outline if `outline` is `true`. `distance_tolerance` can be used as a fuzz factor when `outline` is `true`, for example, to help GUI users mouse-click-select a shape from its outline more successfully
 
 ### `PerfectShape::PointLocation`
@@ -86,6 +85,71 @@ Includes `PerfectShape::PointLocation`
 - `#max_x`: max x
 - `#max_y`: max y
 
+### `PerfectShape::AffineTransform`
+
+Class
+
+Affine transforms have the following matrix:
+
+[ xxp xyp xt ]<br>
+[ yxp yyp yt ]
+
+The matrix is used to transform (x,y) point coordinates as follows:
+
+[ xxp xyp xt ] * [x] = [ xxp * x + xyp * y + xt ]<br>
+[ yxp yyp yt ] * [y] = [ yxp * x + yyp * y + yt ]
+
+`xxp` is the x coordinate x product (`m11`)<br>
+`xyp` is the x coordinate y product (`m12`)<br>
+`yxp` is the y coordinate x product (`m21`)<br>
+`yyp` is the y coordinate y product (`m22`)<br>
+`xt` is the x coordinate translation (`m13`)<br>
+`yt` is the y coordinate translation (`m23`)
+
+Affine transform mutation operations ending with `!` can be chained as they all return `self`.
+
+- `::new(xxp_element = nil, xyp_element = nil, yxp_element = nil, yyp_element = nil, xt_element = nil, yt_element = nil,
+         xxp: nil, xyp: nil, yxp: nil, yyp: nil, xt: nil, yt: nil,
+         m11: nil, m12: nil, m21: nil, m22: nil, m13: nil, m23: nil)`:
+  The constructor accepts either the (x,y)-operation related argument/kwarg names or traditional matrix element kwarg names. If no arguments are supplied, it constructs an identity matrix (i.e. like calling `::new(xxp: 1, xyp: 0, yxp: 0, yyp: 1, xt: 0, yt: 0)`).
+- `#matrix_3d`: Returns Ruby `Matrix` object representing affine transform in 3D (used internally for performing multiplication)
+- `#==(other)`: Returns `true` if equal to `other` or `false` otherwise
+- `#identity!` (alias: `reset!`): Resets to identity matrix (i.e. like calling `::new(xxp: 1, xyp: 0, yxp: 0, yyp: 1, xt: 0, yt: 0)`)
+- `#invertible?` Returns `true` if matrix is invertible and `false` otherwise
+- `#invert!`: Inverts affine transform matrix if invertible or raises an error otherwise
+- `#multiply!(other)`: Multiplies affine transform with another affine transform, storing resulting changes in matrix elements
+- `#translate!(x_or_point, y=nil)`: Translates affine transform with (x, y) translation values
+- `#scale!(x_or_point, y=nil)`: Scales affine transform with (x, y) scale values
+- `#rotate!(degrees)`: Rotates by angle degrees counter-clockwise if angle value is positive or clockwise if angle value is negative. Note that it returns very close approximate results for rotations that are 90/180/270 degrees (good enough for inverse-transform GUI point containment checks needed when checking if mouse-click-point is inside a transformed shape).
+- `#shear!(x_or_point, y=nil)`: Shears by x and y factors
+- `#clone`: Returns a new AffineTransform with the same matrix elements
+- `#transform_point(x_or_point, y=nil)`: returns `[xxp * x + xyp * y + xt, yxp * x + yyp * y + yt]`. Note that result is a close approximation, but should be good enough for GUI mouse-click-point containment checks.
+- `#transform_points(*xy_coordinates_or_points)`: returns `Array` of (x,y) pair `Array`s transformed with `#transform_point` method
+- `#inverse_transform_point(x_or_point, y=nil)`: returns inverse transform of a point (x,y) coordinates (clones self and inverts clone, and then transforms point). Note that result is a close approximation, but should be good enough for GUI mouse-click-point containment checks.
+- `#inverse_transform_points(*xy_coordinates_or_points)`: returns inverse transforms of a point `Array` of (x,y) coordinates
+
+Example:
+
+```ruby
+xxp = 2
+xyp = 3
+yxp = 4
+yyp = 5
+xt = 6
+yt = 7
+affine_transform1 = PerfectShape::AffineTransform.new(xxp: xxp, xyp: xyp, yxp: yxp, yyp: yyp, xt: xt, yt: yt) # (x,y)-operation kwarg names
+affine_transform2 = PerfectShape::AffineTransform.new(m11: xxp, m12: xyp, m21: yxp, m22: yyp, m13: xt, m23: yt) # traditional matrix element kwarg names
+affine_transform3 = PerfectShape::AffineTransform.new(xxp, xyp, yxp, yyp, xt, yt) # standard arguments
+
+affine_transform2.matrix_3d == affine_transform1.matrix_3d # => true
+affine_transform3.matrix_3d == affine_transform1.matrix_3d # => true
+
+affine_transform = PerfectShape::AffineTransform.new.translate!(30, 20).scale!(2, 3)
+
+affine_transform.transform_point(10, 10) # => approximately [50, 50]
+affine_transform.inverse_transform_point(50, 50) # => approximately [10, 10]
+```
+
 ### `PerfectShape::Point`
 
 Class
@@ -99,6 +163,7 @@ Includes `PerfectShape::PointLocation`
 Points are simply represented by an `Array` of `[x,y]` coordinates when used within other shapes, but when needing point-specific operations like `point_distance`, the `PerfectShape::Point` class can come in handy.
 
 - `::point_distance(x, y, px, py)`: Returns the distance from a point to another point
+- `::normalize_point(x_or_point, y = nil)`: Normalizes point args whether two-number `point` `Array` or `x`, `y` args, returning normalized point `Array` of two `BigDecimal`'s
 - `::new(x_or_point=nil, y_arg=nil, x: nil, y: nil)`: constructs a point with (x,y) pair (default: 0,0) whether specified as `Array` of (x,y) pair, flat `x,y` args, or `x:, y:` kwargs.
 - `#min_x`: min x (always x)
 - `#min_y`: min y (always y)
@@ -139,7 +204,7 @@ Includes `PerfectShape::MultiPoint`
 
 ![line](https://raw.githubusercontent.com/AndyObtiva/perfect-shape/master/images/line.png)
 
-- `::relative_counterclockwise(x1, y1, x2, y2, px, py)`: Returns an indicator of where the specified point (px,py) lies with respect to the line segment from (x1,y1) to (x2,y2). The return value can be either 1, -1, or 0 and indicates in which direction the specified line must pivot around its first end point, (x1,y1), in order to point at the specified point (px,py). A return value of 1 indicates that the line segment must turn in the direction that takes the positive X axis towards the negative Y axis. In the default coordinate system used by Java 2D, this direction is counterclockwise. A return value of -1 indicates that the line segment must turn in the direction that takes the positive X axis towards the positive Y axis. In the default coordinate system, this direction is clockwise. A return value of 0 indicates that the point lies exactly on the line segment. Note that an indicator value of 0 is rare and not useful for determining collinearity because of floating point rounding issues. If the point is colinear with the line segment, but not between the end points, then the value will be -1 if the point lies “beyond (x1,y1)” or 1 if the point lies “beyond (x2,y2)”.
+- `::relative_counterclockwise(x1, y1, x2, y2, px, py)`: Returns an indicator of where the specified point (px,py) lies with respect to the line segment from (x1,y1) to (x2,y2). The return value can be either 1, -1, or 0 and indicates in which direction the specified line must pivot around its first end point, (x1,y1), in order to point at the specified point (px,py). A return value of 1 indicates that the line segment must turn in the direction that takes the positive X axis towards the negative Y axis. In the default coordinate system, this direction is counterclockwise. A return value of -1 indicates that the line segment must turn in the direction that takes the positive X axis towards the positive Y axis. In the default coordinate system, this direction is clockwise. A return value of 0 indicates that the point lies exactly on the line segment. Note that an indicator value of 0 is rare and not useful for determining collinearity because of floating point rounding issues. If the point is colinear with the line segment, but not between the end points, then the value will be -1 if the point lies “beyond (x1,y1)” or 1 if the point lies “beyond (x2,y2)”.
 - `::point_distance_square(x1, y1, x2, y2, px, py)`: Returns the square of distance from a point to a line segment.
 - `::point_distance(x1, y1, x2, y2, px, py)`: Returns the distance from a point to a line segment.
 - `::new(points: [])`: constructs a line with two `points` as `Array` of `Array`s of `[x,y]` pairs or flattened `Array` of alternating x and y coordinates
@@ -155,7 +220,7 @@ Includes `PerfectShape::MultiPoint`
 - `#bounding_box`: bounding box is a rectangle with x = min x, y = min y, and width/height of shape
 - `#==(other)`: Returns `true` if equal to `other` or `false` otherwise
 - `#contain?(x_or_point, y=nil, outline: true, distance_tolerance: 0)`: checks if point lies on line, with a distance tolerance (0 by default). Distance tolerance provides a fuzz factor that for example enables GUI users to mouse-click-select a line shape more successfully. `outline` option makes no difference on line
-- `#relative_counterclockwise(x_or_point, y=nil)`: Returns an indicator of where the specified point (px,py) lies with respect to the line segment from (x1,y1) to (x2,y2). The return value can be either 1, -1, or 0 and indicates in which direction the specified line must pivot around its first end point, (x1,y1), in order to point at the specified point (px,py). A return value of 1 indicates that the line segment must turn in the direction that takes the positive X axis towards the negative Y axis. In the default coordinate system used by Java 2D, this direction is counterclockwise. A return value of -1 indicates that the line segment must turn in the direction that takes the positive X axis towards the positive Y axis. In the default coordinate system, this direction is clockwise. A return value of 0 indicates that the point lies exactly on the line segment. Note that an indicator value of 0 is rare and not useful for determining collinearity because of floating point rounding issues. If the point is colinear with the line segment, but not between the end points, then the value will be -1 if the point lies “beyond (x1,y1)” or 1 if the point lies “beyond (x2,y2)”.
+- `#relative_counterclockwise(x_or_point, y=nil)`: Returns an indicator of where the specified point (px,py) lies with respect to the line segment from (x1,y1) to (x2,y2). The return value can be either 1, -1, or 0 and indicates in which direction the specified line must pivot around its first end point, (x1,y1), in order to point at the specified point (px,py). A return value of 1 indicates that the line segment must turn in the direction that takes the positive X axis towards the negative Y axis. In the default coordinate system, this direction is counterclockwise. A return value of -1 indicates that the line segment must turn in the direction that takes the positive X axis towards the positive Y axis. In the default coordinate system, this direction is clockwise. A return value of 0 indicates that the point lies exactly on the line segment. Note that an indicator value of 0 is rare and not useful for determining collinearity because of floating point rounding issues. If the point is colinear with the line segment, but not between the end points, then the value will be -1 if the point lies “beyond (x1,y1)” or 1 if the point lies “beyond (x2,y2)”.
 - `#point_distance(x_or_point, y=nil)`: Returns the distance from a point to a line segment.
 
 Example:
@@ -735,7 +800,7 @@ A composite shape is simply an aggregate of multiple shapes (e.g. square and pol
 - `#center_y`: center y
 - `#bounding_box`: bounding box is a rectangle with x = min x, y = min y, and width/height of shape (bounding box only guarantees that the shape is within it, but it might be bigger than the shape)
 - `#==(other)`: Returns `true` if equal to `other` or `false` otherwise
-- `#contain?(x_or_point, y=nil)`: When `outline` is `false`, it checks if point is inside any of the shapes owned by the composite shape. Otherwise, when `outline` is `true`, it checks if point is on the outline of any of the shapes owned by the composite shape. `distance_tolerance` can be used as a fuzz factor when `outline` is `true`, for example, to help GUI users mouse-click-select a composite shape from its outline more successfully
+- `#contain?(x_or_point, y=nil, outline: false, distance_tolerance: 0)`: When `outline` is `false`, it checks if point is inside any of the shapes owned by the composite shape. Otherwise, when `outline` is `true`, it checks if point is on the outline of any of the shapes owned by the composite shape. `distance_tolerance` can be used as a fuzz factor when `outline` is `true`, for example, to help GUI users mouse-click-select a composite shape from its outline more successfully
 
 Example:
 
@@ -778,7 +843,10 @@ shape.contain?([145, 189], outline: true, distance_tolerance: 1) # => true
 ## Resources
 
 - Rubydoc: https://www.rubydoc.info/gems/perfect-shape
-- AWT Geom Javadoc (inspiration): https://docs.oracle.com/javase/8/docs/api/java/awt/geom/package-summary.html
+- Point in Polygon: https://en.wikipedia.org/wiki/Point_in_polygon
+- Even-Odd Rule: https://en.wikipedia.org/wiki/Even%E2%80%93odd_rule
+- Nonzero Rule: https://en.wikipedia.org/wiki/Nonzero-rule
+- IEEE 754-1985 Remainder: https://en.wikipedia.org/wiki/IEEE_754-1985
 
 ## TODO
 

data/VERSION

@@ -1 +1 @@
-0.3.5
+0.4.0

data/lib/perfect-shape.rb

@@ -23,8 +23,8 @@ $LOAD_PATH.unshift File.expand_path('.', __dir__)
 
 require 'bigdecimal'
 require 'equalizer'
+require 'matrix'
 
-# Perfect Shape algorithms are mostly ported from java.awt.geom: https://docs.oracle.com/javase/8/docs/api/java/awt/geom/package-summary.html
 module PerfectShape
 end
 

data/lib/perfect_shape/affine_transform.rb

@@ -0,0 +1,235 @@
+# Copyright (c) 2021-2022 Andy Maleh
+#
+# Permission is hereby granted, free of charge, to any person obtaining
+# a copy of this software and associated documentation files (the
+# "Software"), to deal in the Software without restriction, including
+# without limitation the rights to use, copy, modify, merge, publish,
+# distribute, sublicense, and/or sell copies of the Software, and to
+# permit persons to whom the Software is furnished to do so, subject to
+# the following conditions:
+#
+# The above copyright notice and this permission notice shall be
+# included in all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+# EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+# NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
+# LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
+# OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
+# WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+
+require 'perfect_shape/math'
+require 'perfect_shape/shape'
+require 'perfect_shape/point'
+require 'perfect_shape/multi_point'
+
+module PerfectShape
+  # Represents an affine transform
+  class AffineTransform
+    include Equalizer.new(:xxp, :xyp, :yxp, :yyp, :xt, :yt)
+  
+    attr_reader :xxp, :xyp, :yxp, :yyp, :xt, :yt
+    alias m11 xxp
+    alias m12 xyp
+    alias m21 yxp
+    alias m22 yyp
+    alias m13 xt
+    alias m23 yt
+  
+    # Creates a new AffineTransform with the following Matrix:
+    #
+    # [ xxp xyp xt ]
+    # [ yxp yyp yt ]
+    #
+    # The Matrix is used to transform (x,y) point coordinates as follows:
+    #
+    # [ xxp xyp xt ] * [x] = [ xxp * x + xyp * y + xt ]
+    # [ yxp yyp yt ] * [y] = [ yxp * x + yyp * y + yt ]
+    #
+    # xxp is the x coordinate x product (m11)
+    # xyp is the x coordinate y product (m12)
+    # yxp is the y coordinate x product (m21)
+    # yyp is the y coordinate y product (m22)
+    # xt is the x coordinate translation (m13)
+    # yt is the y coordinate translation (m23)
+    #
+    # The constructor accepts either the (x,y)-operation related argument/kwarg names
+    # or traditional Matrix element kwarg names
+    #
+    # Example with (x,y)-operation kwarg names:
+    #
+    # AffineTransform.new(xxp: 2, xyp: 3, yxp: 4, yyp: 5, xt: 6, yt: 7)
+    #
+    # Example with traditional Matrix element kwarg names:
+    #
+    # AffineTransform.new(m11: 2, m12: 3, m21: 4, m22: 5, m13: 6, m23: 7)
+    #
+    # Example with standard arguments:
+    #
+    # AffineTransform.new(2, 3, 4, 5, 6, 7)
+    #
+    # If no arguments are supplied, it constructs an identity matrix
+    # (i.e. like calling `::new(xxp: 1, xyp: 0, yxp: 0, yyp: 1, xt: 0, yt: 0)`)
+    def initialize(xxp_element = nil, xyp_element = nil, yxp_element = nil, yyp_element = nil, xt_element = nil, yt_element = nil,
+                   xxp: nil, xyp: nil, yxp: nil, yyp: nil, xt: nil, yt: nil,
+                   m11: nil, m12: nil, m21: nil, m22: nil, m13: nil, m23: nil)
+      self.xxp = xxp_element || xxp || m11 || 1
+      self.xyp = xyp_element || xyp || m12 || 0
+      self.yxp = yxp_element || yxp || m21 || 0
+      self.yyp = yyp_element || yyp || m22 || 1
+      self.xt  = xt_element  || xt  || m13 || 0
+      self.yt  = yt_element  || yt  || m23 || 0
+    end
+    
+    def xxp=(value)
+      @xxp = BigDecimal(value.to_s)
+    end
+    alias m11= xxp=
+    
+    def xyp=(value)
+      @xyp = BigDecimal(value.to_s)
+    end
+    alias m12= xyp=
+    
+    def yxp=(value)
+      @yxp = BigDecimal(value.to_s)
+    end
+    alias m21= yxp=
+    
+    def yyp=(value)
+      @yyp = BigDecimal(value.to_s)
+    end
+    alias m22= yyp=
+    
+    def xt=(value)
+      @xt = BigDecimal(value.to_s)
+    end
+    alias m13= xt=
+    
+    def yt=(value)
+      @yt = BigDecimal(value.to_s)
+    end
+    alias m23= yt=
+    
+    # Resets to identity matrix
+    # Returns self to support fluent interface chaining
+    def identity!
+      self.xxp = 1
+      self.xyp = 0
+      self.yxp = 0
+      self.yyp = 1
+      self.xt  = 0
+      self.yt  = 0
+      
+      self
+    end
+    alias reset! identity!
+
+    # Inverts AffineTransform matrix if invertible
+    # Raises an error if affine transform matrix is not invertible
+    # Returns self to support fluent interface chaining
+    def invert!
+      raise 'Cannot invert (matrix is not invertible)!' if !invertible?
+      
+      self.matrix_3d = matrix_3d.inverse
+      
+      self
+    end
+    
+    def invertible?
+      (m11 * m22 - m12 * m21) != 0
+    end
+    
+    # Multiplies by other AffineTransform
+    def multiply!(other_affine_transform)
+      self.matrix_3d = matrix_3d*other_affine_transform.matrix_3d
+      
+      self
+    end
+    
+    # Translates AffineTransform
+    def translate!(x_or_point, y = nil)
+      x, y = Point.normalize_point(x_or_point, y)
+      return unless x && y
+      
+      translation_affine_transform = AffineTransform.new(xxp: 1, xyp: 0, yxp: 0, yyp: 1, xt: x, yt: y)
+      multiply!(translation_affine_transform)
+      
+      self
+    end
+    
+    # Scales AffineTransform
+    def scale!(x_or_point, y = nil)
+      x, y = Point.normalize_point(x_or_point, y)
+      return unless x && y
+      
+      scale_affine_transform = AffineTransform.new(xxp: x, xyp: 0, yxp: 0, yyp: y, xt: 0, yt: 0)
+      multiply!(scale_affine_transform)
+      
+      self
+    end
+    
+    # Rotates AffineTransform counter-clockwise for positive angle value in degrees
+    # or clockwise for negative angle value in degrees
+    def rotate!(degrees)
+      degrees = Math.normalize_degrees(degrees)
+      radians = Math.degrees_to_radians(degrees)
+      
+      rotation_affine_transform = AffineTransform.new(xxp: Math.cos(radians), xyp: -Math.sin(radians), yxp: Math.sin(radians), yyp: Math.cos(radians), xt: 0, yt: 0)
+      multiply!(rotation_affine_transform)
+      
+      self
+    end
+    
+    # Shears AffineTransform by (x,y) amount
+    def shear!(x_or_point, y = nil)
+      x, y = Point.normalize_point(x_or_point, y)
+      return unless x && y
+      
+      shear_affine_transform = AffineTransform.new(xxp: 1 + x*y, xyp: x, yxp: y, yyp: 1, xt: 0, yt: 0)
+      multiply!(shear_affine_transform)
+      
+      self
+    end
+    alias skew! shear!
+    
+    # Sets elements from a Ruby Matrix representing Affine Transform matrix elements in 3D
+    def matrix_3d=(the_matrix_3d)
+      self.xxp = the_matrix_3d[0, 0]
+      self.xyp = the_matrix_3d[0, 1]
+      self.xt = the_matrix_3d[0, 2]
+      self.yxp = the_matrix_3d[1, 0]
+      self.yyp = the_matrix_3d[1, 1]
+      self.yt = the_matrix_3d[1, 2]
+    end
+    
+    # Returns Ruby Matrix representing Affine Transform matrix elements in 3D
+    def matrix_3d
+      Matrix[[xxp, xyp, xt], [yxp, yyp, yt], [0, 0, 1]]
+    end
+    
+    def transform_point(x_or_point, y = nil)
+      x, y = Point.normalize_point(x_or_point, y)
+      return unless x && y
+      
+      [xxp*x + xyp*y + xt, yxp*x + yyp*y + yt]
+    end
+    
+    def inverse_transform_point(x_or_point, y = nil)
+      clone.invert!.transform_point(x_or_point, y)
+    end
+    
+    def transform_points(*xy_coordinates_or_points)
+      points = xy_coordinates_or_points.first.is_a?(Array) ? xy_coordinates_or_points.first : xy_coordinates_or_points
+      points = MultiPoint.normalize_point_array(points)
+      points.map { |point| transform_point(point) }
+    end
+    
+    def inverse_transform_points(*xy_coordinates_or_points)
+      points = xy_coordinates_or_points.first.is_a?(Array) ? xy_coordinates_or_points.first : xy_coordinates_or_points
+      points = MultiPoint.normalize_point_array(points)
+      points.map { |point| inverse_transform_point(point) }
+    end
+  end
+end

data/lib/perfect_shape/arc.rb

@@ -21,10 +21,10 @@
 
 require 'perfect_shape/shape'
 require 'perfect_shape/rectangular_shape'
+require 'perfect_shape/point'
 require 'perfect_shape/line'
 
 module PerfectShape
-  # Mostly ported from java.awt.geom: https://docs.oracle.com/javase/8/docs/api/java/awt/geom/Arc2D.html
   class Arc < Shape
     include RectangularShape
     include Equalizer.new(:type, :x, :y, :width, :height, :start, :extent)
@@ -145,7 +145,7 @@ module PerfectShape
     # the arc, {@code false} if the point lies outside of the
     # arc's bounds.
     def contain?(x_or_point, y = nil, outline: false, distance_tolerance: 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       if outline
         if type == :pie && x == center_x && y == center_y

data/lib/perfect_shape/composite_shape.rb

@@ -64,7 +64,7 @@ module PerfectShape
     # the composite shape or false if the point lies outside of the
     # path's bounds.
     def contain?(x_or_point, y = nil, outline: false, distance_tolerance: 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       
       shapes.any? { |shape| shape.contain?(x, y, outline: outline, distance_tolerance: distance_tolerance) }

data/lib/perfect_shape/cubic_bezier_curve.rb

@@ -20,10 +20,10 @@
 # WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
 require 'perfect_shape/shape'
+require 'perfect_shape/point'
 require 'perfect_shape/multi_point'
 
 module PerfectShape
-  # Mostly ported from java.awt.geom: https://docs.oracle.com/javase/8/docs/api/java/awt/geom/QuadCurve2D.html
   class CubicBezierCurve < Shape
     class << self
       # Calculates the number of times the cubic bézier curve from (x1,y1) to (x2,y2)
@@ -86,7 +86,7 @@ module PerfectShape
     # the cubic bézier curve, {@code false} if the point lies outside of the
     # cubic bézier curve's bounds.
     def contain?(x_or_point, y = nil, outline: false, distance_tolerance: 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       
       if outline
@@ -120,7 +120,7 @@ module PerfectShape
     # +1 is added for each crossing where the Y coordinate is increasing
     # -1 is added for each crossing where the Y coordinate is decreasing
     def point_crossings(x_or_point, y = nil, level = 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       CubicBezierCurve.point_crossings(points[0][0], points[0][1], points[1][0], points[1][1], points[2][0], points[2][1], points[3][0], points[3][1], x, y, level)
     end
@@ -183,7 +183,7 @@ module PerfectShape
     end
     
     def point_distance(x_or_point, y = nil, minimum_distance_threshold: OUTLINE_MINIMUM_DISTANCE_THRESHOLD)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       
       point = Point.new(x, y)

data/lib/perfect_shape/ellipse.rb

@@ -22,7 +22,6 @@
 require 'perfect_shape/arc'
 
 module PerfectShape
-  # Mostly ported from java.awt.geom: https://docs.oracle.com/javase/8/docs/api/java/awt/geom/Ellipse2D.html
   class Ellipse < Arc
     MESSAGE_CANNOT_UPDATE_ATTRIUBTE = "Ellipse %s cannot be updated. If you want to update type, use Arc instead!"
     
@@ -65,7 +64,7 @@ module PerfectShape
     # ellipse's bounds.
     def contain?(x_or_point, y = nil, outline: false, distance_tolerance: 0)
       # This is implemented again even though super would have just worked to have an optimized algorithm for Ellipse.
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       if outline
         super(x, y, outline: true, distance_tolerance: distance_tolerance)

data/lib/perfect_shape/line.rb

@@ -20,10 +20,10 @@
 # WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
 require 'perfect_shape/shape'
+require 'perfect_shape/point'
 require 'perfect_shape/multi_point'
 
 module PerfectShape
-  # Mostly ported from java.awt.geom: https://docs.oracle.com/javase/8/docs/api/java/awt/geom/Line2D.html
   class Line < Shape
     class << self
       # Returns an indicator of where the specified point (px,py) lies with respect to the line segment from
@@ -213,20 +213,20 @@ module PerfectShape
     # the line, {@code false} if the point lies outside of the
     # line's bounds.
     def contain?(x_or_point, y = nil, outline: true, distance_tolerance: 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       distance_tolerance = BigDecimal(distance_tolerance.to_s)
       point_distance(x, y) <= distance_tolerance
     end
     
     def point_distance(x_or_point, y = nil)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       Line.point_distance(points[0][0], points[0][1], points[1][0], points[1][1], x, y)
     end
     
     def relative_counterclockwise(x_or_point, y = nil)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       Line.relative_counterclockwise(points[0][0], points[0][1], points[1][0], points[1][1], x, y)
     end
@@ -237,7 +237,7 @@ module PerfectShape
     # +1 is returned for a crossing where the Y coordinate is increasing
     # -1 is returned for a crossing where the Y coordinate is decreasing
     def point_crossings(x_or_point, y = nil)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       Line.point_crossings(points[0][0], points[0][1], points[1][0], points[1][1], x, y)
     end

data/lib/perfect_shape/multi_point.rb

@@ -24,6 +24,19 @@ require 'perfect_shape/shape'
 module PerfectShape
   # Represents multi-point shapes like Line, Polygon, and Polyline
   module MultiPoint
+    class << self
+      def normalize_point_array(the_points)
+        if the_points.all? {|the_point| the_point.is_a?(Array)}
+          the_points
+        else
+          the_points = the_points.flatten
+          xs = the_points.each_with_index.select {|n, i| i.even?}.map(&:first)
+          ys = the_points.each_with_index.select {|n, i| i.odd?}.map(&:first)
+          xs.zip(ys)
+        end
+      end
+    end
+  
     attr_reader :points
     
     def initialize(points: [])
@@ -32,11 +45,7 @@ module PerfectShape
     
     # Sets points, normalizing to an Array of Arrays of (x,y) pairs as BigDecimal
     def points=(the_points)
-      unless the_points.first.is_a?(Array)
-        xs = the_points.each_with_index.select {|n, i| i.even?}.map(&:first)
-        ys = the_points.each_with_index.select {|n, i| i.odd?}.map(&:first)
-        the_points = xs.zip(ys)
-      end
+      the_points = MultiPoint.normalize_point_array(the_points)
       @points = the_points.map do |pair|
         [
           pair.first.is_a?(BigDecimal) ? pair.first : BigDecimal(pair.first.to_s),

data/lib/perfect_shape/path.rb

@@ -27,7 +27,6 @@ require 'perfect_shape/cubic_bezier_curve'
 require 'perfect_shape/multi_point'
 
 module PerfectShape
-  # Mostly ported from java.awt.geom: https://docs.oracle.com/javase/8/docs/api/java/awt/geom/Path2D.html
   class Path < Shape
     include MultiPoint
     include Equalizer.new(:shapes, :closed, :winding_rule)
@@ -115,7 +114,7 @@ module PerfectShape
     # the path or false if the point lies outside of the
     # path's bounds.
     def contain?(x_or_point, y = nil, outline: false, distance_tolerance: 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       
       if outline
@@ -149,7 +148,7 @@ module PerfectShape
     # The caller must check for NaN values.
     # The caller may also reject infinite values as well.
     def point_crossings(x_or_point, y = nil)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       return 0 if shapes.count == 0
       movx = movy = curx = cury = endx = endy = 0

data/lib/perfect_shape/point.rb

@@ -33,6 +33,21 @@ module PerfectShape
         py = py.is_a?(BigDecimal) ? py : BigDecimal(py.to_s)
         BigDecimal(Math.sqrt((px - x)**2 + (py - y)**2).to_s)
       end
+      
+      # Normalizes point args whether two-number Array or x, y args returning
+      # normalized point array of two BigDecimal's
+      #
+      # @param x_or_point The point or X coordinate of the point to test.
+      # @param y The Y coordinate of the point to test.
+      #
+      # @return Array of x and y BigDecimal's representing point
+      def normalize_point(x_or_point, y = nil)
+        x = x_or_point
+        x, y = x if y.nil? && x_or_point.is_a?(Array) && x_or_point.size == 2
+        x = x.is_a?(BigDecimal) ? x : BigDecimal(x.to_s)
+        y = y.is_a?(BigDecimal) ? y : BigDecimal(y.to_s)
+        [x, y]
+      end
     end
     
     include PointLocation
@@ -68,14 +83,14 @@ module PerfectShape
     # @return {@code true} if the point is close enough within distance tolerance,
     # {@code false} if the point is too far.
     def contain?(x_or_point, y = nil, outline: true, distance_tolerance: 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       distance_tolerance = BigDecimal(distance_tolerance.to_s)
       point_distance(x, y) <= distance_tolerance
     end
     
     def point_distance(x_or_point, y = nil)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       
       Point.point_distance(self.x, self.y, x, y)

data/lib/perfect_shape/polygon.rb

@@ -20,10 +20,10 @@
 # WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
 require 'perfect_shape/shape'
+require 'perfect_shape/point'
 require 'perfect_shape/multi_point'
 
 module PerfectShape
-  # Mostly ported from java.awt.geom: https://docs.oracle.com/javase/8/docs/api/java/awt/Polygon.html
   class Polygon < Shape
     include MultiPoint
     include Equalizer.new(:points)
@@ -38,7 +38,7 @@ module PerfectShape
     # the polygon, {@code false} if the point lies outside of the
     # polygon's bounds.
     def contain?(x_or_point, y = nil, outline: false, distance_tolerance: 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       if outline
         edges.any? { |edge| edge.contain?(x, y, distance_tolerance: distance_tolerance) }

data/lib/perfect_shape/quadratic_bezier_curve.rb

@@ -20,10 +20,10 @@
 # WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
 
 require 'perfect_shape/shape'
+require 'perfect_shape/point'
 require 'perfect_shape/multi_point'
 
 module PerfectShape
-  # Mostly ported from java.awt.geom: https://docs.oracle.com/javase/8/docs/api/java/awt/geom/QuadCurve2D.html
   class QuadraticBezierCurve < Shape
     class << self
       # Calculates the number of times the quadratic bézier curve from (x1,y1) to (x2,y2)
@@ -80,7 +80,7 @@ module PerfectShape
     # the quadratic bézier curve, {@code false} if the point lies outside of the
     # quadratic bézier curve's bounds.
     def contain?(x_or_point, y = nil, outline: false, distance_tolerance: 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
 
       x1 = points[0][0]
@@ -190,7 +190,7 @@ module PerfectShape
     # +1 is added for each crossing where the Y coordinate is increasing
     # -1 is added for each crossing where the Y coordinate is decreasing
     def point_crossings(x_or_point, y = nil, level = 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       QuadraticBezierCurve.point_crossings(points[0][0], points[0][1], points[1][0], points[1][1], points[2][0], points[2][1], x, y, level)
     end
@@ -247,7 +247,7 @@ module PerfectShape
     end
     
     def point_distance(x_or_point, y = nil, minimum_distance_threshold: OUTLINE_MINIMUM_DISTANCE_THRESHOLD)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       
       point = Point.new(x, y)

data/lib/perfect_shape/rectangle.rb

@@ -21,10 +21,10 @@
 
 require 'perfect_shape/shape'
 require 'perfect_shape/rectangular_shape'
+require 'perfect_shape/point'
 require 'perfect_shape/line'
 
 module PerfectShape
-  # Mostly ported from java.awt.geom: https://docs.oracle.com/javase/8/docs/api/java/awt/geom/Rectangle2D.html
   class Rectangle < Shape
     include RectangularShape
     include Equalizer.new(:x, :y, :width, :height)
@@ -38,7 +38,7 @@ module PerfectShape
     # the rectangle, {@code false} if the point lies outside of the
     # rectangle's bounds.
     def contain?(x_or_point, y = nil, outline: false, distance_tolerance: 0)
-      x, y = normalize_point(x_or_point, y)
+      x, y = Point.normalize_point(x_or_point, y)
       return unless x && y
       
       if outline

data/lib/perfect_shape/shape.rb

@@ -75,21 +75,6 @@ module PerfectShape
       Rectangle.new(x: min_x, y: min_y, width: width, height: height)
     end
   
-    # Normalizes point args whether two-number Array or x, y args returning
-    # normalized point array of two BigDecimal's
-    #
-    # @param x_or_point The point or X coordinate of the point to test.
-    # @param y The Y coordinate of the point to test.
-    #
-    # @return Array of x and y BigDecimal's representing point
-    def normalize_point(x_or_point, y = nil)
-      x = x_or_point
-      x, y = x if y.nil? && x_or_point.is_a?(Array) && x_or_point.size == 2
-      x = x.is_a?(BigDecimal) ? x : BigDecimal(x.to_s)
-      y = y.is_a?(BigDecimal) ? y : BigDecimal(y.to_s)
-      [x, y]
-    end
-    
     # Subclasses must implement
     def contain?(x_or_point, y = nil, outline: false, distance_tolerance: 0)
     end

data/perfect-shape.gemspec

@@ -2,17 +2,17 @@
 # DO NOT EDIT THIS FILE DIRECTLY
 # Instead, edit Juwelier::Tasks in Rakefile, and run 'rake gemspec'
 # -*- encoding: utf-8 -*-
-# stub: perfect-shape 0.3.5 ruby lib
+# stub: perfect-shape 0.4.0 ruby lib
 
 Gem::Specification.new do |s|
   s.name = "perfect-shape".freeze
-  s.version = "0.3.5"
+  s.version = "0.4.0"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0".freeze) if s.respond_to? :required_rubygems_version=
   s.require_paths = ["lib".freeze]
   s.authors = ["Andy Maleh".freeze]
-  s.date = "2022-01-11"
-  s.description = "Perfect Shape is a collection of pure Ruby geometric algorithms that are mostly useful for GUI manipulation like checking containment of a mouse click point in popular geometry shapes such as rectangle, square, arc (open, chord, and pie), ellipse, circle, polygon, and paths containing lines, quadratic b\u00E9zier curves, and cubic bezier curves (including both Ray Casting Algorithm, aka Even-odd Rule, and Winding Number Algorithm, aka Nonzero Rule). Additionally, it contains some purely mathematical algorithms like IEEEremainder (also known as IEEE-754 remainder).".freeze
+  s.date = "2022-01-16"
+  s.description = "Perfect Shape is a collection of pure Ruby geometric algorithms that are mostly useful for GUI manipulation like checking containment of a mouse click point in popular geometry shapes such as rectangle, square, arc (open, chord, and pie), ellipse, circle, polygon, and paths containing lines, quadratic b\u00E9zier curves, and cubic bezier curves, potentially with affine transforms applied like translation, scale, rotation, shear/skew, and inversion (including both Ray Casting Algorithm, aka Even-odd Rule, and Winding Number Algorithm, aka Nonzero Rule). Additionally, it contains some purely mathematical algorithms like IEEEremainder (also known as IEEE-754 remainder).".freeze
   s.email = "andy.am@gmail.com".freeze
   s.extra_rdoc_files = [
     "CHANGELOG.md",
@@ -25,6 +25,7 @@ Gem::Specification.new do |s|
     "README.md",
     "VERSION",
     "lib/perfect-shape.rb",
+    "lib/perfect_shape/affine_transform.rb",
     "lib/perfect_shape/arc.rb",
     "lib/perfect_shape/circle.rb",
     "lib/perfect_shape/composite_shape.rb",

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: perfect-shape
 version: !ruby/object:Gem::Version
-  version: 0.3.5
+  version: 0.4.0
 platform: ruby
 authors:
 - Andy Maleh
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2022-01-11 00:00:00.000000000 Z
+date: 2022-01-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: equalizer
@@ -98,9 +98,11 @@ description: Perfect Shape is a collection of pure Ruby geometric algorithms tha
   are mostly useful for GUI manipulation like checking containment of a mouse click
   point in popular geometry shapes such as rectangle, square, arc (open, chord, and
   pie), ellipse, circle, polygon, and paths containing lines, quadratic bézier curves,
-  and cubic bezier curves (including both Ray Casting Algorithm, aka Even-odd Rule,
-  and Winding Number Algorithm, aka Nonzero Rule). Additionally, it contains some
-  purely mathematical algorithms like IEEEremainder (also known as IEEE-754 remainder).
+  and cubic bezier curves, potentially with affine transforms applied like translation,
+  scale, rotation, shear/skew, and inversion (including both Ray Casting Algorithm,
+  aka Even-odd Rule, and Winding Number Algorithm, aka Nonzero Rule). Additionally,
+  it contains some purely mathematical algorithms like IEEEremainder (also known as
+  IEEE-754 remainder).
 email: andy.am@gmail.com
 executables: []
 extensions: []
@@ -114,6 +116,7 @@ files:
 - README.md
 - VERSION
 - lib/perfect-shape.rb
+- lib/perfect_shape/affine_transform.rb
 - lib/perfect_shape/arc.rb
 - lib/perfect_shape/circle.rb
 - lib/perfect_shape/composite_shape.rb