Geomotion for RubyMotion
iOS Geometry in idiomatic Ruby. Exhaustively tested. What's not to love?
- Adds methods to return useful information, like whether a
rect.contains?(a_point), orpoint.distance_to(another_point) - Easily modify CGRects with methods like
shrink_left,grow_down,below, and many many more. - Easy conversion to and from
NSValue(#to_ns_valueand##from_ns_value) - Adds nice
inspectmethods - Many operators (
+, -, *) CATransform3DandCGAffineTransformmethods to create and concatenate transforms- Read on for in-depth examples!
# Initializers
rect = CGRect.new([10, 100], [50, 20])
rect = CGRect.make(x: 10, y: 100, width: 50, height: 20)
rect = CGRect.make(origin: CGPoint(0, 0), size: CGSize(0, 0))
# there are, for convenience, function versions of these:
rect = CGRect(10, 100, 50, 20)
rect = CGRect([10, 100], [50, 20])
rect = CGRect([[10, 100], [50, 20]])
rect = CGRect(x: 10, y: 100, w: 50, h: 20)
rect = CGRect(origin: [10, 100], size: [50, 20])
# Getters
[rect.x, rect.y, rect.width, rect.height]
=> [10, 100, 50, 20]
goofy_rect = CGRect.make(x: 10.1, y: 100.9, width: 50.2, height: 20.8)
goofy_rect.integral
=> CGRect([10.0, 100.0], [51.0, 22.0])
rect_zero = CGRect.zero
rect_zero = CGRect.empty # alias for CGRect.zero
=> CGRect(0, 0, 0, 0)
rect_zero.empty?
=> true
# to get the center of the frame, relative to the origin or in absolute coordinates
rect.center
=> CGPoint(25, 10) # center relative to bounds
rect.center(true)
=> CGPoint(35, 110) # center relative to frame
# length of the diagonals
rect = CGRect.new([0, 0], [30, 40])
rect.diagonal # => 50
# Other points in the rect can be returned as well, and the same
# relative/absolute return values are supported (defaults to relative)
top_left top_center
| |
o--o--o top_right
| |
center_left o x o center_right
| |
o--o--o bottom_right
| |
bottom_left bottom_center
# Operator Overloading
-rect
=> CGRect(-10, -100, -50, -20)
# union of rects
rect + CGRect.make(x: 9, y: 99, width: 10, height: 10)
=> rect.union_with(CGRect.make(x: 9, y: 99, width: 10, height: 10))
=> CGRect(9, 99, 50, 20)
# increases the size, but keeps the origin
rect + CGSize.make(width: 11, height: 1)
=> CGRect(10, 100, 61, 21)
# not the same as `grow`, which grows the rect in all directions
# move the rect via a point
rect + CGPoint.make(x: 10, y: 10)
=> rect.offset(CGPoint.make(x: 10, y: 10))
=> CGRect(20, 110, 50, 20)
# move the rect via an offset
rect + UIOffsetMake(10, 10)
rect.offset(UIOffsetMake(10, 10))
rect.offset(10, 10)
=> CGRect(20, 110, 50, 20)
a_point + a_size
=> CGRect(a_point, a_size) # a point and a size make a rectangle. makes sense, right?
# Union and Intersection
rect.union_with CGRect.make(x: 9, y: 99, width: 10, height: 10)
=> CGRect(9, 99, 50, 20)
rect.intersection_with CGRect.make(x: 9, y: 99, width: 10, height: 10)
=> CGRect(10, 100, 10, 10)
rect.intersects?(another_rect)
=> true/false, whether they overlap at all or not
rect.contains?(a_point or a_rect)
=> true/false, whether the point or rect is *completely contained* in the receiving rect
# Growing and shrinking
# The center stays the same. Think margins!
rect.grow(CGSize.make(width: 10, height: 20))
=> CGRect(5, 90, 60, 40)
rect.shrink(10)
=> CGRect(15, 105, 40, 10)
# Powerful layout adjustments with chainable methods
view = UIView.alloc.initWithFrame rect.below.width(100).height(10)
view.frame
=> CGRect(10, 120, 100, 10)
view2 = UIView.alloc.initWithFrame rect.beside(10)
view2.frame
=> CGRect(70, 100, 50, 20.0)
# More examples of adjustments
rect = CGRect.make(x: 10, y: 100, width: 50, height: 20)
[rect.right(20).x, rect.left(20).x, rect.up(20).y, rect.down(20).y]
=> [30, -10, 80, 120]
# Layout "above" and "before" rectangles
# (default offset is the rectangle's width or height)
rect.before(5)
=> CGRect(-45, 100, 50, 20)
rect.before(5, width:20)
=> CGRect(-15, 100, 20, 20)
rect.above(5)
=> CGRect(10, 75, 50, 20)
rect.above(5, height:10)
=> CGRect(10, 85, 50, 10)
# Layout a rect relative to others
rect2 = CGRect.make(x: 50, y: 50, width: 100, height: 100)
rect3 = CGRect.make(x:100, y: 200, width: 20, height: 20)
CGRect.layout(rect, above: rect2, right_of: rect3)
=> CGRect(120, 30, 50, 20)
# Also supports margins
CGRect.layout(rect, above: rect2, right_of: rect3, margins: [0, 0, 10, 15])
=> CGRect(135, 20, 50, 20)When you are positioning frames, you'll be doing so in one of two ways:
- Two frames relative to each other, within a common parent frame
- A frame being added as a child of another frame
(generally speaking)
geomotion is optimized for both cases, but the arsenal of methods is different.
Any of the location methods (up, down, left, right, beside, before, above, below)
will return a frame that is in the same coordinate system of the receiver, and
this behavior cannot be changed.
frame = CGRect.make(x: 10, y: 10, width:10, height: 10)
frame.beside
# => [[20, 10], [10, 10]]
frame.right(30).down(5).taller(100)
# => [[10+30, 10+5], [10, 10+100]]
# aka
# => [[40, 15], [10, 110]]Any methods that include the x or y variable in their name will be absolute.
frame.x
frame.min_x
frame.max_x
frame.mid_x
frame.y
frame.min_y
frame.max_y
frame.mid_yAny of the position methods that do NOT include the x or y variable will ignore the x and y values unless explicitly told to use absolute coordinates.
Note: These methods will "normalize" the width and height, so even if the width or height is negative, these methods will always return positive numbers. If you specify absolute coordinates, the values might be negative, but they will also be sorted (x == min, min < mid, mid < max, x + width == max).
frame = CGRect.make(x: 10, y: 10, width:10, height: 10)
frame.top_left # => [0, 0]
frame.top_center # => [5, 0]
frame.bottom_right # => [10, 10]
# use absolute coordinates
frame.top_left(true) # => [10, 10]
frame.top_center(true) # => [15, 10]
frame.bottom_right(true) # => [20, 20]
# negative widths and heights are "corrected" when using absolute coordinates
frame = CGRect.make(x: 20, y: 20, width:-10, height: -10)
frame.top_center(true) # => [15, 10]
frame.bottom_right(true) # => [20, 20]Most of the frame-manipulation methods delegate to the apply method. You can
use this method to perform batch changes.
frame = view.frame.apply(left: 10, y: 0, wider: 50, grow_height: 10)All of the methods that return a new frame (left, shrink, below and friends)
also accept a hash in which you can apply more changes. You can accomplish the
same thing using method chaining; this is an implementation detail that might
also clean your code up by grouping changes.
frame = CGRect.make(x: 10, y: 10, width:10, height: 10)
frame.beside.width(20).down(10).height(20)
# => [[20, 20], [20, 20]]
# using the options hash / apply method
frame.beside(width: 20, down: 10, height: 20)
# => [[20, 20], [20, 20]]
frame.below(grow_width: 10, grow_up: 5)
# => [[0, 15], [40, 25]]
# convert to NSValue, for use in NSCoding or where an Objective-C object is
# needed. CGRect is a "boxed" object in RubyMotion, and in Objective-C it is a
# C-struct and so can't be stored in an NSArray, for example.
NSValue.valueWithCGRect(CGRect.new([0, 10], [10, 20]))
# =>
value = CGRect.new([0, 10], [10, 20]).to_ns_value
rect = CGRect.from_ns_value(value)# Initializers
size = CGSize.new(50, 20)
size = CGSize.make(width: 50, height: 20)
# there are, for convenience, function versions of these:
size = CGSize(50, 20)
size = CGSize([50, 20])
size = CGSize(width: 50, height: 20)
# Getters
size_zero = CGSize.empty
=> CGSize(0, 0)
size_zero.empty?
=> true
# length of the diagonals
size = CGSize.new([30, 40])
size.diagonal # => 50
# modify width, height, or both
# bigger
size_zero = CGSize.empty
size_zero.grow(5) # => CGSize(5, 5)
size_zero.wider(10) # => CGSize(10, 0)
size_zero.taller(10) # => CGSize(0, 10)
# smaller
size_ten = CGSize.new(10, 10)
size_ten.shrink(5) # => CGSize(5, 5)
size_ten.shorter(10) # => CGSize(10, 0)
size_ten.thinner(10) # => CGSize(0, 10)
# Operator Overloading
-size
=> CGSize(-50, -20)
size + CGSize.make(width: 100, height: 50)
=> CGSize(150, 70)
size + CGPoint.make(x: 10, y: 30)
=> CGRect(10, 30, 50, 20)
# Combine with CGPoint
size.rect_at_point CGPoint.make(x: 10, y: 30)
=> CGRect(10, 30, 50, 20)
# convert to NSValue, for use in NSCoding or where an Objective-C object is
# needed.
NSValue.valueWithCGSize(CGSize.new(0, 10))
# =>
value = CGSize.new(0, 10).to_ns_value
size = CGSize.from_ns_value(value)# Initializers
point = CGPoint.new(10, 100)
point = CGPoint.make(x: 10, y: 100)
# there are, for convenience, function versions of these:
point = CGPoint(10, 100)
point = CGPoint([10, 100])
point = CGPoint(x: 10, y: 100)
# Return a modified copy
point.up(50).left(5)
=> CGPoint(5, 50)
# original is not modified, a new point is returned
point.down(50).right(5)
=> CGPoint(15, 150)
# Operator Overloading
-point
=> CGPoint(-10, -100)
point + CGPoint.make(x: 20, y: 40)
=> CGPoint(30, 140)
point + CGSize.make(width: 50, height: 20)
=> CGRect(10, 100, 50, 20)
# Combine with CGSize
point.rect_of_size CGSize.make(width: 50, height: 20)
=> CGRect(10, 100, 50, 20)
# Compare with CGRect
point.inside? CGRect.make(x: 0, y: 0, width: 20, height: 110)
=> true
# Compare with origin
# length
CGPoint.new(3, 4).length
=> 5
# angle
CGPoint.new(1, 1).angle * 180 / Math::PI
=> 45.0
# if you only need to *compare* lengths, use rough_length. It is faster, since
# it doesn't perform the sqrt part of pythagorean's theorem.
CGPoint.new(3, 4).rough_length
=> 25
# Distance to point
point = CGPoint.new(10, 100)
point.distance_to(CGPoint.make(x: 13, y:104))
=> 5
# If you just need to know whether the points are within a certain distance, it
# is faster to use distance_within? (it uses rough_length to compare the distances)
point.distance_within?(5, to: CGPoint.make(x: 13, y: 104))
=> true
# Angle between target and receiver
# (hint: our answer should be 45°)
point = CGPoint.new(10, 100)
point.angle_to(CGPoint.make(x: 20, y:110))
=> 0.785398163397 (pi/4)
# convert to NSValue, for use in NSCoding or where an Objective-C object is
# needed.
NSValue.valueWithCGPoint(CGPoint.new(0, 10))
# =>
value = CGPoint.new(0, 10).to_ns_value
point = CGPoint.from_ns_value(value)These are assigned to the UIView#transform parameter. See CATransform3D for
the transforms that are designed for CALayer object.
# you *can* create it manually
transform = CGAffineTransform.make(a: 1, b: 0, c: 0, d: 1, tx: 0, ty: 0)
transform = CGAffineTransform(1, 0, 0, 1, 0, 0)
# but don't! the `make` method accepts `translate`, `scale`, and `rotate` args
transform = CGAffineTransform.make(scale: 2, translate: [10, 10], rotate: Math::PI)
# identity transform is easy
CGAffineTransform.identity
# just to be sure
CGAffineTransform.identity.identity? # => true
# Operator Overloading
transform1 = CGAffineTransform.make(scale: 2)
transform2 = CGAffineTransform.make(translate: [10, 10])
# concatenate transforms
transform1 + transform2
transform1 << transform2 # alias
transform1 - transform2
# => transform1 + -transform2
# => transform1 + transform2.invert
transform1 - transform1 # => CGAffineTransform.identity
# create new transforms by calling `translate`, `scale`, or `rotate` as factory
# methods
CGAffineTransform.translate(10, 10)
CGAffineTransform.scale(2) # scale x and y by 2
CGAffineTransform.scale(2, 4) # scale x by 2 and y by 4
CGAffineTransform.rotate(Math::PI / 4)
# "shearing" turns a rectangle into a parallelogram
# see sceenshot below or run geomotion app
CGAffineTransform.shear(0.5, 0) # in x direction
CGAffineTransform.shear(0, 0.5) # in y direction
# you can combine these, but it looks kind of strange. better to pick one
# direction
# or you can chain these methods
CGAffineTransform.identity.translate(10, 10).scale(2).rotate(Math::PI / 4)
# convert to NSValue, for use in NSCoding or where an Objective-C object is
# needed.
NSValue.valueWithCGAffineTransform(CGAffineTransform.translate(0, 10))
# =>
value = CGAffineTransform.translate(0, 10).to_ns_value
transform = CGAffineTransform.from_ns_value(value)
CALayers can take on full 3D transforms.
# these are really gnarly
transform = CATransform3D.make(
m11: 1, m12: 0, m13: 0, m14: 0,
m21: 0, m22: 1, m23: 0, m24: 0,
m31: 0, m32: 0, m33: 1, m34: 0,
m41: 0, m42: 0, m43: 0, m44: 1,)
transform = CATransform3D(same thing works here)
# accepts transforms like CGAffineTransform, but many take 3 instead of 2 args
transform = CATransform3D.make(scale: [2, 2, 1], translate: [10, 10, 10], rotate: Math::PI)
# identity transform
CATransform3D.identity
CATransform3D.identity.identity? # => true
# Operator Overloading
transform1 = CATransform3D.make(scale: 2)
transform2 = CATransform3D.make(translate: [10, 10])
# concatenate transforms
transform1 + transform2
transform1 << transform2 # alias
transform1 - transform2
# => transform1 + -transform2
# => transform1 + transform2.invert
transform1 - transform1 # => CATransform3D.identity
# create new transforms by calling factory methods
CATransform3D.translate(10, 10, 10)
CATransform3D.scale(2) # scale x and y by 2
CATransform3D.scale(2, 4, 3) # scale x by 2, y by 4, z by 3
CATransform3D.rotate(Math::PI / 4)
# "shearing" works the same as CGAffineTransform
CATransform3D.shear(0.5, 0) # in x direction
CATransform3D.shear(0, 0.5) # in y direction
# "perspective" changes are better than rotation because they make one side
# bigger and one side smaller
# see sceenshot below or run geomotion app
CATransform3D.perspective(0.002, 0) # similar to rotating around x-axis
CATransform3D.perspective(0, 0.002) # "rotates" around the y-axis
# or you can chain these methods
CATransform3D.identity.translate(10, 10, 10).scale(2).rotate(Math::PI / 4)
# convert to NSValue, for use in NSCoding or CAKeyframeAnimation#values
NSValue.valueWithCATransform3D(CATransform3D.translate(0, 10, 0))
# =>
value = CATransform3D.translate(0, 10, 0).to_ns_value
transform = CATransform3D.from_ns_value(value)
-
gem install geomotion -
Add
require 'geomotion'in yourRakefile.
If you have cool/better ideas, pull-request away!