@awayjs/core
Version:
AwayJS core classes
682 lines (626 loc) • 22.3 kB
text/typescript
import { ArgumentError } from '../errors/ArgumentError';
import { Point } from './Point';
import { Vector3D } from './Vector3D';
/**
* The Matrix export class represents a transformation matrix that determines how to
* map points from one coordinate space to another. You can perform various
* graphical transformations on a display object by setting the properties of
* a Matrix object, applying that Matrix object to the <code>matrix</code>
* property of a Transform object, and then applying that Transform object as
* the <code>transform</code> property of the display object. These
* transformation functions include translation(<i>x</i> and <i>y</i>
* repositioning), rotation, scaling, and skewing.
*
* <p>Together these types of transformations are known as <i>affine
* transformations</i>. Affine transformations preserve the straightness of
* lines while transforming, so that parallel lines stay parallel.</p>
*
* <p>To apply a transformation matrix to a display object, you create a
* Transform object, set its <code>matrix</code> property to the
* transformation matrix, and then set the <code>transform</code> property of
* the display object to the Transform object. Matrix objects are also used as
* parameters of some methods, such as the following:</p>
*
* <ul>
* <li>The <code>draw()</code> method of a BitmapData object</li>
* <li>The <code>beginBitmapFill()</code> method,
* <code>beginGradientFill()</code> method, or
* <code>lineGradientStyle()</code> method of a Graphics object</li>
* </ul>
*
* <p>A transformation matrix object is a 3 x 3 matrix with the following
* contents:</p>
*
* <p>In traditional transformation matrixes, the <code>u</code>,
* <code>v</code>, and <code>w</code> properties provide extra capabilities.
* The Matrix export class can only operate in two-dimensional space, so it always
* assumes that the property values <code>u</code> and <code>v</code> are 0.0,
* and that the property value <code>w</code> is 1.0. The effective values of
* the matrix are as follows:</p>
*
* <p>You can get and set the values of all six of the other properties in a
* Matrix object: <code>a</code>, <code>b</code>, <code>c</code>,
* <code>d</code>, <code>tx</code>, and <code>ty</code>.</p>
*
* <p>The Matrix export class supports the four major types of transformations:
* translation, scaling, rotation, and skewing. You can set three of these
* transformations by using specialized methods, as described in the following
* table: </p>
*
* <p>Each transformation function alters the current matrix properties so
* that you can effectively combine multiple transformations. To do this, you
* call more than one transformation function before applying the matrix to
* its display object target(by using the <code>transform</code> property of
* that display object).</p>
*
* <p>Use the <code>new Matrix()</code> constructor to create a Matrix object
* before you can call the methods of the Matrix object.</p>
*/
export class Matrix {
public rawData: Float32Array = new Float32Array(6);
/**
* The value that affects the positioning of pixels along the <i>x</i> axis
* when scaling or rotating an image.
*/
public get a(): number {
return this.rawData[0];
}
public set a(value: number) {
this.rawData[0] = value;
}
/**
* The value that affects the positioning of pixels along the <i>y</i> axis
* when rotating or skewing an image.
*/
public get b(): number {
return this.rawData[1];
}
public set b(value: number) {
this.rawData[1] = value;
}
/**
* The value that affects the positioning of pixels along the <i>x</i> axis
* when rotating or skewing an image.
*/
public get c(): number {
return this.rawData[2];
}
public set c(value: number) {
this.rawData[2] = value;
}
/**
* The value that affects the positioning of pixels along the <i>y</i> axis
* when scaling or rotating an image.
*/
public get d(): number {
return this.rawData[3];
}
public set d(value: number) {
this.rawData[3] = value;
}
/**
* The distance by which to translate each point along the <i>x</i> axis.
*/
public get tx(): number {
return this.rawData[4];
}
public set tx(value: number) {
this.rawData[4] = value;
}
/**
* The distance by which to translate each point along the <i>y</i> axis.
*/
public get ty(): number {
return this.rawData[5];
}
public set ty(value: number) {
this.rawData[5] = value;
}
/**
* Creates a new Matrix object with the specified parameters. In matrix
* notation, the properties are organized like this:
*
* <p>If you do not provide any parameters to the <code>new Matrix()</code>
* constructor, it creates an <i>identity matrix</i> with the following
* values:</p>
*
* <p>In matrix notation, the identity matrix looks like this:</p>
*
* @param a The value that affects the positioning of pixels along the
* <i>x</i> axis when scaling or rotating an image.
* @param b The value that affects the positioning of pixels along the
* <i>y</i> axis when rotating or skewing an image.
* @param c The value that affects the positioning of pixels along the
* <i>x</i> axis when rotating or skewing an image.
* @param d The value that affects the positioning of pixels along the
* <i>y</i> axis when scaling or rotating an image..
* @param tx The distance by which to translate each point along the <i>x</i>
* axis.
* @param ty The distance by which to translate each point along the <i>y</i>
* axis.
*/
constructor(rawData?: Float32Array);
constructor(a?: number, b?: number, c?: number, d?: number, tx?: number, ty?: number);
constructor(a: number | Float32Array = 1,
b: number = 0,
c: number = 0,
d: number = 1,
tx: number = 0,
ty: number = 0) {
if (a instanceof Float32Array) {
this.copyRawDataFrom(a);
} else {
const raw: Float32Array = this.rawData;
raw[0] = Number(a);
raw[1] = b;
raw[2] = c;
raw[3] = d;
raw[4] = tx;
raw[5] = ty;
}
}
public copyRawDataFrom(vector: Float32Array, offset: number = 0): void {
const raw: Float32Array = this.rawData;
raw[0] = vector[offset + 0];
raw[1] = vector[offset + 1];
raw[2] = vector[offset + 2];
raw[3] = vector[offset + 3];
raw[4] = vector[offset + 4];
raw[5] = vector[offset + 5];
}
/**
* Returns a new Matrix object that is a clone of this matrix, with an exact
* copy of the contained object.
*
* @return A Matrix object.
*/
public clone(): Matrix {
const raw: Float32Array = this.rawData;
return new Matrix(raw[0], raw[1], raw[2], raw[3], raw[4], raw[5]);
}
/**
* Concatenates a matrix with the current matrix, effectively combining the
* geometric effects of the two. In mathematical terms, concatenating two
* matrixes is the same as combining them using matrix multiplication.
*
* <p>For example, if matrix <code>m1</code> scales an object by a factor of
* four, and matrix <code>m2</code> rotates an object by 1.5707963267949
* radians(<code>Math.PI/2</code>), then <code>m1.concat(m2)</code>
* transforms <code>m1</code> into a matrix that scales an object by a factor
* of four and rotates the object by <code>Math.PI/2</code> radians. </p>
*
* <p>This method replaces the source matrix with the concatenated matrix. If
* you want to concatenate two matrixes without altering either of the two
* source matrixes, first copy the source matrix by using the
* <code>clone()</code> method, as shown in the Class Examples section.</p>
*
* @param matrix The matrix to be concatenated to the source matrix.
*/
public concat(matrix: Matrix): void {
const m: Float32Array = this.rawData;
const n: Float32Array = matrix.rawData;
let a = m[0] * n[0];
let b = 0.0;
let c = 0.0;
let d = m[3] * n[3];
let tx = m[4] * n[0] + n[4];
let ty = m[5] * n[3] + n[5];
if (m[1] !== 0.0 || m[2] !== 0.0 || n[1] !== 0.0 || n[2] !== 0.0) {
a += m[1] * n[2];
d += m[2] * n[1];
b += m[0] * n[1] + m[1] * n[3];
c += m[2] * n[0] + m[3] * n[2];
tx += m[5] * n[2];
ty += m[4] * n[1];
}
m[0] = a;
m[1] = b;
m[2] = c;
m[3] = d;
m[4] = tx;
m[5] = ty;
}
/**
* Copies a Vector3D object into specific column of the calling Matrix3D
* object.
*
* @param column The column from which to copy the data from.
* @param vector3D The Vector3D object from which to copy the data.
*/
public copyColumnFrom(column: number, vector3D: Vector3D): void {
const raw: Float32Array = this.rawData;
const rawVector3D: Float32Array = vector3D._rawData;
if (column > 2) {
throw 'Column ' + column + ' out of bounds (2)';
} else if (column == 0) {
raw[0] = rawVector3D[0];
raw[1] = rawVector3D[1];
} else if (column == 1) {
raw[2] = rawVector3D[0];
raw[3] = rawVector3D[1];
} else {
raw[4] = rawVector3D[0];
raw[5] = rawVector3D[1];
}
}
/**
* Copies specific column of the calling Matrix object into the Vector3D
* object. The w element of the Vector3D object will not be changed.
*
* @param column The column from which to copy the data from.
* @param vector3D The Vector3D object from which to copy the data.
*/
public copyColumnTo(column: number, vector3D: Vector3D): void {
const raw: Float32Array = this.rawData;
const rawVector3D: Float32Array = vector3D._rawData;
if (column > 2) {
throw new ArgumentError('ArgumentError, Column ' + column + ' out of bounds [0, ..., 2]');
} else if (column == 0) {
rawVector3D[0] = raw[0];
rawVector3D[1] = raw[1];
rawVector3D[2] = 0;
} else if (column == 1) {
rawVector3D[0] = raw[2];
rawVector3D[1] = raw[3];
rawVector3D[2] = 0;
} else {
rawVector3D[0] = raw[4];
rawVector3D[1] = raw[5];
rawVector3D[2] = 1;
}
}
/**
* Copies all of the matrix data from the source Point object into the
* calling Matrix object.
*
* @param sourceMatrix The Matrix object from which to copy the data.
*/
public copyFrom(sourceMatrix: Matrix): void {
const raw: Float32Array = this.rawData;
const sourceRaw: Float32Array = sourceMatrix.rawData;
raw[0] = sourceRaw[0];
raw[1] = sourceRaw[1];
raw[2] = sourceRaw[2];
raw[3] = sourceRaw[3];
raw[4] = sourceRaw[4];
raw[5] = sourceRaw[5];
}
/**
* Copies a Vector3D object into specific row of the calling Matrix object.
*
* @param row The row from which to copy the data from.
* @param vector3D The Vector3D object from which to copy the data.
*/
public copyRowFrom(row: number, vector3D: Vector3D): void {
const raw: Float32Array = this.rawData;
const rawVector3D: Float32Array = vector3D._rawData;
if (row > 2) {
throw new ArgumentError('ArgumentError, Row ' + row + ' out of bounds [0, ..., 2]');
} else if (row == 0) {
raw[0] = rawVector3D[0];
raw[2] = rawVector3D[1];
raw[4] = rawVector3D[2];
} else {
raw[1] = rawVector3D[0];
raw[3] = rawVector3D[1];
raw[5] = rawVector3D[2];
}
}
/**
* Copies specific row of the calling Matrix object into the Vector3D object.
* The w element of the Vector3D object will not be changed.
*
* @param row The row from which to copy the data from.
* @param vector3D The Vector3D object from which to copy the data.
*/
public copyRowTo(row: number, vector3D: Vector3D): void {
const raw: Float32Array = this.rawData;
const rawVector3D: Float32Array = vector3D._rawData;
if (row > 2) {
throw new ArgumentError('ArgumentError, Row ' + row + ' out of bounds [0, ..., 2]');
} else if (row == 0) {
rawVector3D[0] = raw[0];
rawVector3D[1] = raw[2];
rawVector3D[2] = raw[4];
} else if (row == 1) {
rawVector3D[0] = raw[1];
rawVector3D[1] = raw[3];
rawVector3D[2] = raw[5];
} else {
rawVector3D[0] = 0;
rawVector3D[1] = 0;
rawVector3D[2] = 1;
}
}
/**
* Includes parameters for scaling, rotation, and translation. When applied
* to a matrix it sets the matrix's values based on those parameters.
*
* <p>Using the <code>createBox()</code> method lets you obtain the same
* matrix as you would if you applied the <code>identity()</code>,
* <code>rotate()</code>, <code>scale()</code>, and <code>translate()</code>
* methods in succession. For example, <code>mat1.createBox(2,2,Math.PI/4,
* 100, 100)</code> has the same effect as the following:</p>
*
* @param scaleX The factor by which to scale horizontally.
* @param scaleY The factor by which scale vertically.
* @param rotation The amount to rotate, in radians.
* @param tx The number of pixels to translate(move) to the right
* along the <i>x</i> axis.
* @param ty The number of pixels to translate(move) down along the
* <i>y</i> axis.
*/
public createBox(scaleX: number, scaleY: number, rotation: number = 0, tx: number = 0, ty: number = 0): void {
const raw: Float32Array = this.rawData;
if (rotation !== 0) {
const u = Math.cos(rotation);
const v = Math.sin(rotation);
raw[0] = u * scaleX;
raw[1] = v * scaleY;
raw[2] = -v * scaleX;
raw[3] = u * scaleY;
} else {
raw[0] = scaleX;
raw[1] = 0;
raw[2] = 0;
raw[3] = scaleY;
}
raw[4] = tx;
raw[5] = ty;
}
/**
* Creates the specific style of matrix expected by the
* <code>beginGradientFill()</code> and <code>lineGradientStyle()</code>
* methods of the Graphics class. Width and height are scaled to a
* <code>scaleX</code>/<code>scaleY</code> pair and the
* <code>tx</code>/<code>ty</code> values are offset by half the width and
* height.
*
* <p>For example, consider a gradient with the following
* characteristics:</p>
*
* <ul>
* <li><code>GradientType.LINEAR</code></li>
* <li>Two colors, green and blue, with the ratios array set to <code>[0,
* 255]</code></li>
* <li><code>SpreadMethod.PAD</code></li>
* <li><code>InterpolationMethod.LINEAR_RGB</code></li>
* </ul>
*
* <p>The following illustrations show gradients in which the matrix was
* defined using the <code>createGradientBox()</code> method with different
* parameter settings:</p>
*
* @param width The width of the gradient box.
* @param height The height of the gradient box.
* @param rotation The amount to rotate, in radians.
* @param tx The distance, in pixels, to translate to the right along
* the <i>x</i> axis. This value is offset by half of the
* <code>width</code> parameter.
* @param ty The distance, in pixels, to translate down along the
* <i>y</i> axis. This value is offset by half of the
* <code>height</code> parameter.
*/
public createGradientBox(width: number,
height: number,
rotation: number = 0,
tx: number = 0, ty:
number = 0): void {
this.createBox(width / 1638.4, height / 1638.4, rotation, tx + width / 2, ty + height / 2);
}
/**
* Given a point in the pretransform coordinate space, returns the
* coordinates of that point after the transformation occurs. Unlike the
* standard transformation applied using the <code>transformPoint()</code>
* method, the <code>deltaTransformPoint()</code> method's transformation
* does not consider the translation parameters <code>tx</code> and
* <code>ty</code>.
*
* @param point The point for which you want to get the result of the matrix
* transformation.
* @return The point resulting from applying the matrix transformation.
*/
public deltaTransformPoint(point: Point): Point {
const raw: Float32Array = this.rawData;
return new Point(point.x * raw[0] + point.y * raw[2], point.x * raw[1] + point.y * raw[3]);
}
/**
* Sets each matrix property to a value that causes a null transformation. An
* object transformed by applying an identity matrix will be identical to the
* original.
*
* <p>After calling the <code>identity()</code> method, the resulting matrix
* has the following properties: <code>a</code>=1, <code>b</code>=0,
* <code>c</code>=0, <code>d</code>=1, <code>tx</code>=0,
* <code>ty</code>=0.</p>
*
* <p>In matrix notation, the identity matrix looks like this:</p>
*
*/
public identity(): void {
const raw: Float32Array = this.rawData;
raw[0] = 1;
raw[1] = 0;
raw[2] = 0;
raw[3] = 1;
raw[4] = 0;
raw[5] = 0;
}
/**
* Performs the opposite transformation of the original matrix. You can apply
* an inverted matrix to an object to undo the transformation performed when
* applying the original matrix.
*/
public invert(): void {
const raw = this.rawData;
let b = raw[1];
let c = raw[2];
const tx = raw[4];
const ty = raw[5];
if (b === 0 && c === 0) {
const a = raw[0] = 1 / raw[0];
const d = raw[3] = 1 / raw[3];
raw[1] = raw[2] = 0;
raw[4] = -a * tx;
raw[5] = -d * ty;
return;
}
const a = raw[0];
let d = raw[3];
let determinant = a * d - b * c;
if (determinant === 0) {
this.identity();
return;
}
/**
* Multiplying by reciprocal of the |determinant| is only accurate if the reciprocal is
* representable without loss of precision. This is usually only the case for powers of
* two: 1/2, 1/4 ...
*/
determinant = 1 / determinant;
let k = 0;
k = raw[0] = d * determinant;
b = raw[1] = -b * determinant;
c = raw[2] = -c * determinant;
d = raw[3] = a * determinant;
raw[4] = -(k * tx + c * ty);
raw[5] = -(b * tx + d * ty);
}
/**
* Returns a new Matrix object that is a clone of this matrix, with an exact
* copy of the contained object.
*
* @param matrix The matrix for which you want to get the result of the matrix
* transformation.
* @return A Matrix object.
*/
public multiply(matrix: Matrix): Matrix {
const result = new Matrix();
result.a = this.a * matrix.a + this.b * matrix.c;
result.b = this.a * matrix.b + this.b * matrix.d;
result.c = this.c * matrix.a + this.d * matrix.c;
result.d = this.c * matrix.b + this.d * matrix.d;
result.tx = this.tx * matrix.a + this.ty * matrix.c + matrix.tx;
result.ty = this.tx * matrix.b + this.ty * matrix.d + matrix.ty;
return result;
}
/**
* Applies a rotation transformation to the Matrix object.
*
* <p>The <code>rotate()</code> method alters the <code>a</code>,
* <code>b</code>, <code>c</code>, and <code>d</code> properties of the
* Matrix object. In matrix notation, this is the same as concatenating the
* current matrix with the following:</p>
*
* @param angle The rotation angle in radians.
*/
public rotate(angle: number): void {
if (angle !== 0) {
const raw: Float32Array = this.rawData;
const u = Math.cos(angle);
const v = Math.sin(angle);
const ta = raw[0];
const tb = raw[1];
const tc = raw[2];
const td = raw[3];
const ttx = raw[4];
const tty = raw[5];
raw[0] = ta * u - tb * v;
raw[1] = ta * v + tb * u;
raw[2] = tc * u - td * v;
raw[3] = tc * v + td * u;
raw[4] = ttx * u - tty * v;
raw[5] = ttx * v + tty * u;
}
}
/**
* Applies a scaling transformation to the matrix. The <i>x</i> axis is
* multiplied by <code>sx</code>, and the <i>y</i> axis it is multiplied by
* <code>sy</code>.
*
* <p>The <code>scale()</code> method alters the <code>a</code> and
* <code>d</code> properties of the Matrix object. In matrix notation, this
* is the same as concatenating the current matrix with the following
* matrix:</p>
*
* @param sx A multiplier used to scale the object along the <i>x</i> axis.
* @param sy A multiplier used to scale the object along the <i>y</i> axis.
*/
public scale(sx: number, sy: number): void {
const raw: Float32Array = this.rawData;
if (sx !== 1) {
raw[0] *= sx;
raw[2] *= sx;
raw[4] *= sx;
}
if (sy !== 1) {
raw[1] *= sy;
raw[3] *= sy;
raw[5] *= sy;
}
}
/**
* Sets the members of Matrix to the specified values.
*
* @param a The value that affects the positioning of pixels along the
* <i>x</i> axis when scaling or rotating an image.
* @param b The value that affects the positioning of pixels along the
* <i>y</i> axis when rotating or skewing an image.
* @param c The value that affects the positioning of pixels along the
* <i>x</i> axis when rotating or skewing an image.
* @param d The value that affects the positioning of pixels along the
* <i>y</i> axis when scaling or rotating an image..
* @param tx The distance by which to translate each point along the <i>x</i>
* axis.
* @param ty The distance by which to translate each point along the <i>y</i>
* axis.
*/
public setTo(a: number, b: number, c: number, d: number, tx: number, ty: number): void {
const raw: Float32Array = this.rawData;
raw[0] = a;
raw[2] = b;
raw[1] = c;
raw[3] = d;
raw[4] = tx;
raw[5] = ty;
}
/**
* Returns a text value listing the properties of the Matrix object.
*
* @return A string containing the values of the properties of the Matrix
* object: <code>a</code>, <code>b</code>, <code>c</code>,
* <code>d</code>, <code>tx</code>, and <code>ty</code>.
*/
public toString(): string {
return '[Matrix] (a=' + this.a
+ ', b=' + this.b
+ ', c=' + this.c
+ ', d=' + this.d
+ ', tx=' + this.tx
+ ', ty=' + this.ty + ')';
}
/**
* Returns the result of applying the geometric transformation represented by
* the Matrix object to the specified point.
*
* @param point The point for which you want to get the result of the Matrix
* transformation.
* @return The point resulting from applying the Matrix transformation.
*/
public transformPoint(point: Point): Point {
const raw: Float32Array = this.rawData;
return new Point(point.x * raw[0] + point.y * raw[2] + raw[4], point.x * raw[1] + point.y * raw[3] + raw[5]);
}
/**
* Translates the matrix along the <i>x</i> and <i>y</i> axes, as specified
* by the <code>dx</code> and <code>dy</code> parameters.
*
* @param dx The amount of movement along the <i>x</i> axis to the right, in
* pixels.
* @param dy The amount of movement down along the <i>y</i> axis, in pixels.
*/
public translate(dx: number, dy: number): void {
this.rawData[4] += dx;
this.rawData[5] += dy;
}
}