package pixel import "github.com/go-gl/mathgl/mgl32" // Transform holds space transformation information. Concretely, a transformation is specified by position, // anchor, scale and rotation. // // All points are first rotated around the anchor. Then they are multiplied by the scale. If the // scale factor is 2, the object becomes 2x bigger. Finally, all points are moved, so that the original // anchor is located precisely at the position. // // Create a Transform object with the Position function. This sets the position variable, which is the // most important. Then use methods, like Scale and Rotate to change scale, rotation and achor. The order // in which you apply these methods is irrelevant. // // pixel.Position(pixel.V(100, 100)).Rotate(math.Pi / 3).Scale(1.5) type Transform struct { pos, anc, sca Vec rot float64 } // Position creates a Transformation object with specified position. Anchor is (0, 0), rotation is 0 and scale is 1. func Position(position Vec) Transform { return Transform{ pos: position, sca: V(1, 1), } } // Position sets position. func (t Transform) Position(position Vec) Transform { t.pos = position return t } // Move adds delta to position. func (t Transform) Move(delta Vec) Transform { t.pos += delta return t } // Anchor sets anchor. Anchor is the rotation center and will be moved to the position. func (t Transform) Anchor(anchor Vec) Transform { t.anc = anchor return t } // MoveAnchor adds delta to anchor. func (t Transform) MoveAnchor(delta Vec) Transform { t.anc += delta return t } // Scale scales the transform by the supplied factor. // // Note, that subsequent calls to this method accumulate the final scale factor. Scaling two times by 2 is equivalent // to scaling once by 4. func (t Transform) Scale(scale float64) Transform { t.sca = t.sca.Scaled(scale) return t } // ScaleXY scales the transform by the supplied X and Y factor. Note, that scale is applied before rotation. // // Note, that subsequent calls to this method accumulate the final scale factor. Scaling two times by 2 is equivalent // to scaling once by 4. func (t Transform) ScaleXY(scale Vec) Transform { t.sca = V(t.sca.X()*scale.X(), t.sca.Y()*scale.Y()) return t } // Rotate rotates the transform by the supplied angle in radians. // // Note, that subsequent calls to this method accumulate the final rotation. Rotating two times by Pi/2 is // equivalent to rotating once by Pi. func (t Transform) Rotate(angle float64) Transform { t.rot += angle return t } // Mat3 returns a transformation matrix that satisfies previously set transform properties. func (t Transform) Mat3() mgl32.Mat3 { mat := mgl32.Ident3() mat = mat.Mul3(mgl32.Translate2D(float32(t.pos.X()), float32(t.pos.Y()))) mat = mat.Mul3(mgl32.Rotate3DZ(float32(t.rot))) mat = mat.Mul3(mgl32.Scale2D(float32(t.sca.X()), float32(t.sca.Y()))) mat = mat.Mul3(mgl32.Translate2D(float32(-t.anc.X()), float32(-t.anc.Y()))) return mat } // Camera is a convenience function, that returns a Transform that acts like a camera. // Center is the position in the world coordinates, that will be projected onto the center of the screen. // One unit in world coordinates will be projected onto zoom pixels. // // It is possible to apply additional rotations, scales and moves to the returned transform. func Camera(center, zoom, screenSize Vec) Transform { return Position(0).Anchor(center).ScaleXY(2 * zoom).ScaleXY(V(1/screenSize.X(), 1/screenSize.Y())) }