| // Copyright 2013 The Flutter Authors. All rights reserved. |
| // Use of this source code is governed by a BSD-style license that can be |
| // found in the LICENSE file. |
| |
| // @dart = 2.6 |
| part of engine; |
| |
| /// Generic callback signature, used by [_futurize]. |
| typedef Callback<T> = void Function(T result); |
| |
| /// Signature for a method that receives a [_Callback]. |
| /// |
| /// Return value should be null on success, and a string error message on |
| /// failure. |
| typedef Callbacker<T> = String Function(Callback<T> callback); |
| |
| /// Converts a method that receives a value-returning callback to a method that |
| /// returns a Future. |
| /// |
| /// Return a [String] to cause an [Exception] to be synchronously thrown with |
| /// that string as a message. |
| /// |
| /// If the callback is called with null, the future completes with an error. |
| /// |
| /// Example usage: |
| /// |
| /// ```dart |
| /// typedef IntCallback = void Function(int result); |
| /// |
| /// String _doSomethingAndCallback(IntCallback callback) { |
| /// new Timer(new Duration(seconds: 1), () { callback(1); }); |
| /// } |
| /// |
| /// Future<int> doSomething() { |
| /// return _futurize(_doSomethingAndCallback); |
| /// } |
| /// ``` |
| Future<T> futurize<T>(Callbacker<T> callbacker) { |
| final Completer<T> completer = Completer<T>.sync(); |
| final String error = callbacker((T t) { |
| if (t == null) { |
| completer.completeError(Exception('operation failed')); |
| } else { |
| completer.complete(t); |
| } |
| }); |
| if (error != null) { |
| throw Exception(error); |
| } |
| return completer.future; |
| } |
| |
| /// Converts [matrix] to CSS transform value. |
| String matrix4ToCssTransform(Matrix4 matrix) { |
| return float64ListToCssTransform(matrix.storage); |
| } |
| |
| /// Applies a transform to the [element]. |
| /// |
| /// See [float64ListToCssTransform] for details on how the CSS value is chosen. |
| void setElementTransform(html.Element element, Float32List matrix4) { |
| element.style |
| ..transformOrigin = '0 0 0' |
| ..transform = float64ListToCssTransform(matrix4); |
| } |
| |
| /// Converts [matrix] to CSS transform value. |
| /// |
| /// To avoid blurry text on some screens this function uses a 2D CSS transform |
| /// if it detects that [matrix] is a 2D transform. Otherwise, it uses a 3D CSS |
| /// transform. |
| /// |
| /// See also: |
| /// * https://github.com/flutter/flutter/issues/32274 |
| /// * https://bugs.chromium.org/p/chromium/issues/detail?id=1040222 |
| String float64ListToCssTransform(Float32List matrix) { |
| assert(matrix.length == 16); |
| final TransformKind transformKind = transformKindOf(matrix); |
| if (transformKind == TransformKind.transform2d) { |
| return float64ListToCssTransform2d(matrix); |
| } else if (transformKind == TransformKind.complex) { |
| return float64ListToCssTransform3d(matrix); |
| } else { |
| assert(transformKind == TransformKind.identity); |
| return 'none'; |
| } |
| } |
| |
| /// The kind of effect a transform matrix performs. |
| enum TransformKind { |
| /// No effect. |
| /// |
| /// We do not want to set any CSS properties in this case. |
| identity, |
| |
| /// A transform that contains only 2d scale, rotation, and translation. |
| /// |
| /// We prefer to use "matrix" instead of "matrix3d" in this case. |
| transform2d, |
| |
| /// All other kinds of transforms. |
| /// |
| /// In this case we will use "matrix3d". |
| complex, |
| } |
| |
| /// Detects the kind of transform the [matrix] performs. |
| TransformKind transformKindOf(Float32List matrix) { |
| assert(matrix.length == 16); |
| final Float32List m = matrix; |
| |
| // If matrix contains scaling, rotation, z translation or |
| // perspective transform, it is not considered simple. |
| final bool isSimple2dTransform = |
| m[15] == 1.0 && // start reading from the last element to eliminate range checks in subsequent reads. |
| m[14] == 0.0 && // z translation is NOT simple |
| // m[13] - y translation is simple |
| // m[12] - x translation is simple |
| m[11] == 0.0 && |
| m[10] == 1.0 && |
| m[9] == 0.0 && |
| m[8] == 0.0 && |
| m[7] == 0.0 && |
| m[6] == 0.0 && |
| // m[5] - scale y is simple |
| // m[4] - 2D rotation is simple |
| m[3] == 0.0 && |
| m[2] == 0.0; |
| // m[1] - 2D rotation is simple |
| // m[0] - scale x is simple |
| |
| if (!isSimple2dTransform) { |
| return TransformKind.complex; |
| } |
| |
| // From this point on we're sure the transform is 2D, but we don't know if |
| // it's identity or not. To check, we need to look at the remaining elements |
| // that were not checked above. |
| final bool isIdentityTransform = |
| m[0] == 1.0 && |
| m[1] == 0.0 && |
| m[4] == 0.0 && |
| m[5] == 1.0 && |
| m[12] == 0.0 && |
| m[13] == 0.0; |
| |
| if (isIdentityTransform) { |
| return TransformKind.identity; |
| } else { |
| return TransformKind.transform2d; |
| } |
| } |
| |
| /// Returns `true` is the [matrix] describes an identity transformation. |
| bool isIdentityFloat32ListTransform(Float32List matrix) { |
| assert(matrix.length == 16); |
| return transformKindOf(matrix) == TransformKind.identity; |
| } |
| |
| /// Converts [matrix] to CSS transform 2D matrix value. |
| /// |
| /// The [matrix] must not be a [TransformKind.complex] transform, because CSS |
| /// `matrix` can only express 2D transforms. [TransformKind.identity] is |
| /// permitted. However, it is inefficient to construct a matrix for an identity |
| /// transform. Consider removing the CSS `transform` property from elements |
| /// that apply identity transform. |
| String float64ListToCssTransform2d(Float32List matrix) { |
| assert (transformKindOf(matrix) != TransformKind.complex); |
| return 'matrix(${matrix[0]},${matrix[1]},${matrix[4]},${matrix[5]},${matrix[12]},${matrix[13]})'; |
| } |
| |
| /// Converts [matrix] to a 3D CSS transform value. |
| String float64ListToCssTransform3d(Float32List matrix) { |
| assert(matrix.length == 16); |
| final Float32List m = matrix; |
| if (m[0] == 1.0 && |
| m[1] == 0.0 && |
| m[2] == 0.0 && |
| m[3] == 0.0 && |
| m[4] == 0.0 && |
| m[5] == 1.0 && |
| m[6] == 0.0 && |
| m[7] == 0.0 && |
| m[8] == 0.0 && |
| m[9] == 0.0 && |
| m[10] == 1.0 && |
| m[11] == 0.0 && |
| // 12 can be anything |
| // 13 can be anything |
| m[14] == 0.0 && |
| m[15] == 1.0) { |
| final double tx = m[12]; |
| final double ty = m[13]; |
| return 'translate3d(${tx}px, ${ty}px, 0px)'; |
| } else { |
| return 'matrix3d(${m[0]},${m[1]},${m[2]},${m[3]},${m[4]},${m[5]},${m[6]},${m[7]},${m[8]},${m[9]},${m[10]},${m[11]},${m[12]},${m[13]},${m[14]},${m[15]})'; |
| } |
| } |
| |
| bool get assertionsEnabled { |
| bool k = false; |
| assert(k = true); |
| return k; |
| } |
| |
| final Float32List _tempRectData = Float32List(4); |
| |
| /// Transforms a [ui.Rect] given the effective [transform]. |
| /// |
| /// The resulting rect is aligned to the pixel grid, i.e. two of |
| /// its sides are vertical and two are horizontal. In the presence of rotations |
| /// the rectangle is inflated such that it fits the rotated rectangle. |
| ui.Rect transformRect(Matrix4 transform, ui.Rect rect) { |
| _tempRectData[0] = rect.left; |
| _tempRectData[1] = rect.top; |
| _tempRectData[2] = rect.right; |
| _tempRectData[3] = rect.bottom; |
| transformLTRB(transform, _tempRectData); |
| return ui.Rect.fromLTRB( |
| _tempRectData[0], |
| _tempRectData[1], |
| _tempRectData[2], |
| _tempRectData[3], |
| ); |
| } |
| |
| /// Temporary storage for intermediate data used by [transformLTRB]. |
| /// |
| /// WARNING: do not use this outside [transformLTRB]. Sharing this variable in |
| /// other contexts will lead to bugs. |
| final Float32List _tempPointData = Float32List(16); |
| final Matrix4 _tempPointMatrix = Matrix4.fromFloat32List(_tempPointData); |
| |
| /// Transforms a rectangle given the effective [transform]. |
| /// |
| /// This is the same as [transformRect], except that the rect is specified |
| /// in terms of left, top, right, and bottom edge offsets. |
| void transformLTRB(Matrix4 transform, Float32List ltrb) { |
| // Construct a matrix where each row represents a vector pointing at |
| // one of the four corners of the (left, top, right, bottom) rectangle. |
| // Using the row-major order allows us to multiply the matrix in-place |
| // by the transposed current transformation matrix. The vector_math |
| // library has a convenience function `multiplyTranspose` that performs |
| // the multiplication without copying. This way we compute the positions |
| // of all four points in a single matrix-by-matrix multiplication at the |
| // cost of one `Matrix4` instance and one `Float32List` instance. |
| // |
| // The rejected alternative was to use `Vector3` for each point and |
| // multiply by the current transform. However, that would cost us four |
| // `Vector3` instances, four `Float32List` instances, and four |
| // matrix-by-vector multiplications. |
| // |
| // `Float32List` initializes the array with zeros, so we do not have to |
| // fill in every single element. |
| |
| // Row 0: top-left |
| _tempPointData[0] = ltrb[0]; |
| _tempPointData[4] = ltrb[1]; |
| _tempPointData[8] = 0; |
| _tempPointData[12] = 1; |
| |
| // Row 1: top-right |
| _tempPointData[1] = ltrb[2]; |
| _tempPointData[5] = ltrb[1]; |
| _tempPointData[9] = 0; |
| _tempPointData[13] = 1; |
| |
| // Row 2: bottom-left |
| _tempPointData[2] = ltrb[0]; |
| _tempPointData[6] = ltrb[3]; |
| _tempPointData[10] = 0; |
| _tempPointData[14] = 1; |
| |
| // Row 3: bottom-right |
| _tempPointData[3] = ltrb[2]; |
| _tempPointData[7] = ltrb[3]; |
| _tempPointData[11] = 0; |
| _tempPointData[15] = 1; |
| |
| _tempPointMatrix.multiplyTranspose(transform); |
| |
| ltrb[0] = math.min(math.min(math.min(_tempPointData[0], _tempPointData[1]), _tempPointData[2]), _tempPointData[3]); |
| ltrb[1] = math.min(math.min(math.min(_tempPointData[4], _tempPointData[5]), _tempPointData[6]), _tempPointData[7]); |
| ltrb[2] = math.max(math.max(math.max(_tempPointData[0], _tempPointData[1]), _tempPointData[2]), _tempPointData[3]); |
| ltrb[3] = math.max(math.max(math.max(_tempPointData[4], _tempPointData[5]), _tempPointData[6]), _tempPointData[7]); |
| } |
| |
| /// Returns true if [rect] contains every point that is also contained by the |
| /// [other] rect. |
| /// |
| /// Points on the edges of both rectangles are also considered. For example, |
| /// this returns true when the two rects are equal to each other. |
| bool rectContainsOther(ui.Rect rect, ui.Rect other) { |
| return rect.left <= other.left && |
| rect.top <= other.top && |
| rect.right >= other.right && |
| rect.bottom >= other.bottom; |
| } |
| |
| /// Counter used for generating clip path id inside an svg <defs> tag. |
| int _clipIdCounter = 0; |
| |
| /// Converts Path to svg element that contains a clip-path definition. |
| /// |
| /// Calling this method updates [_clipIdCounter]. The HTML id of the generated |
| /// clip is set to "svgClip${_clipIdCounter}", e.g. "svgClip123". |
| String _pathToSvgClipPath(ui.Path path, |
| {double offsetX = 0, |
| double offsetY = 0, |
| double scaleX = 1.0, |
| double scaleY = 1.0}) { |
| _clipIdCounter += 1; |
| final StringBuffer sb = StringBuffer(); |
| sb.write('<svg width="0" height="0" ' |
| 'style="position:absolute">'); |
| sb.write('<defs>'); |
| |
| final String clipId = 'svgClip$_clipIdCounter'; |
| sb.write('<clipPath id=$clipId clipPathUnits="objectBoundingBox">'); |
| |
| sb.write('<path transform="scale($scaleX, $scaleY)" fill="#FFFFFF" d="'); |
| pathToSvg(path, sb, offsetX: offsetX, offsetY: offsetY); |
| sb.write('"></path></clipPath></defs></svg'); |
| return sb.toString(); |
| } |
| |
| /// Converts color to a css compatible attribute value. |
| String colorToCssString(ui.Color color) { |
| if (color == null) { |
| return null; |
| } |
| final int value = color.value; |
| if ((0xff000000 & value) == 0xff000000) { |
| final String hexValue = (value & 0xFFFFFF).toRadixString(16); |
| final int hexValueLength = hexValue.length; |
| switch (hexValueLength) { |
| case 1: |
| return '#00000$hexValue'; |
| case 2: |
| return '#0000$hexValue'; |
| case 3: |
| return '#000$hexValue'; |
| case 4: |
| return '#00$hexValue'; |
| case 5: |
| return '#0$hexValue'; |
| default: |
| return '#$hexValue'; |
| } |
| } else { |
| final double alpha = ((value >> 24) & 0xFF) / 255.0; |
| final StringBuffer sb = StringBuffer(); |
| sb.write('rgba('); |
| sb.write(((value >> 16) & 0xFF).toString()); |
| sb.write(','); |
| sb.write(((value >> 8) & 0xFF).toString()); |
| sb.write(','); |
| sb.write((value & 0xFF).toString()); |
| sb.write(','); |
| sb.write(alpha.toString()); |
| sb.write(')'); |
| return sb.toString(); |
| } |
| } |
| |
| /// Converts color components to a CSS compatible attribute value. |
| String colorComponentsToCssString(int r, int g, int b, int a) { |
| if (a == 255) { |
| return 'rgb($r,$g,$b)'; |
| } else { |
| final double alphaRatio = a / 255; |
| return 'rgba($r,$g,$b,${alphaRatio.toStringAsFixed(2)})'; |
| } |
| } |
| |
| /// Determines if the (dynamic) exception passed in is a NS_ERROR_FAILURE |
| /// (from Firefox). |
| /// |
| /// NS_ERROR_FAILURE (0x80004005) is the most general of all the (Firefox) |
| /// errors and occurs for all errors for which a more specific error code does |
| /// not apply. (https://developer.mozilla.org/en-US/docs/Mozilla/Errors) |
| /// |
| /// Other browsers do not throw this exception. |
| /// |
| /// In Flutter, this exception happens when we try to perform some operations on |
| /// a Canvas when the application is rendered in a display:none iframe. |
| /// |
| /// We need this in [BitmapCanvas] and [RecordingCanvas] to swallow this |
| /// Firefox exception without interfering with others (potentially useful |
| /// for the programmer). |
| bool _isNsErrorFailureException(dynamic e) { |
| return js_util.getProperty(e, 'name') == 'NS_ERROR_FAILURE'; |
| } |
| |
| /// From: https://developer.mozilla.org/en-US/docs/Web/CSS/font-family#Syntax |
| /// |
| /// Generic font families are a fallback mechanism, a means of preserving some |
| /// of the style sheet author's intent when none of the specified fonts are |
| /// available. Generic family names are keywords and must not be quoted. A |
| /// generic font family should be the last item in the list of font family |
| /// names. |
| const Set<String> _genericFontFamilies = <String>{ |
| 'serif', |
| 'sans-serif', |
| 'monospace', |
| 'cursive', |
| 'fantasy', |
| 'system-ui', |
| 'math', |
| 'emoji', |
| 'fangsong', |
| }; |
| |
| /// A default fallback font family in case an unloaded font has been requested. |
| /// |
| /// -apple-system targets San Francisco in Safari (on Mac OS X and iOS), |
| /// and it targets Neue Helvetica and Lucida Grande on older versions of |
| /// Mac OS X. It properly selects between San Francisco Text and |
| /// San Francisco Display depending on the text’s size. |
| /// |
| /// For iOS, default to -apple-system, where it should be available, otherwise |
| /// default to Arial. BlinkMacSystemFont is used for Chrome on iOS. |
| final String _fallbackFontFamily = _isMacOrIOS ? |
| '-apple-system, BlinkMacSystemFont' : 'Arial'; |
| |
| bool get _isMacOrIOS => operatingSystem == OperatingSystem.iOs || |
| operatingSystem == OperatingSystem.macOs; |
| |
| /// Create a font-family string appropriate for CSS. |
| /// |
| /// If the given [fontFamily] is a generic font-family, then just return it. |
| /// Otherwise, wrap the family name in quotes and add a fallback font family. |
| String canonicalizeFontFamily(String fontFamily) { |
| if (_genericFontFamilies.contains(fontFamily)) { |
| return fontFamily; |
| } |
| if (_isMacOrIOS) { |
| // Unlike Safari, Chrome on iOS does not correctly fallback to cupertino |
| // on sans-serif. |
| // Map to San Francisco Text/Display fonts, use -apple-system, |
| // BlinkMacSystemFont. |
| if (fontFamily == '.SF Pro Text' || fontFamily == '.SF Pro Display' || |
| fontFamily == '.SF UI Text' || fontFamily == '.SF UI Display') { |
| return _fallbackFontFamily; |
| } |
| } |
| return '"$fontFamily", $_fallbackFontFamily, sans-serif'; |
| } |
| |
| /// Converts a list of [Offset] to a typed array of floats. |
| Float32List offsetListToFloat32List(List<ui.Offset> offsetList) { |
| if (offsetList == null) { |
| return null; |
| } |
| final int length = offsetList.length; |
| final floatList = Float32List(length * 2); |
| for (int i = 0, destIndex = 0; i < length; i++, destIndex += 2) { |
| floatList[destIndex] = offsetList[i].dx; |
| floatList[destIndex + 1] = offsetList[i].dy; |
| } |
| return floatList; |
| } |
| |
| /// Apply this function to container elements in the HTML render tree (this is |
| /// not relevant to semantics tree). |
| /// |
| /// On WebKit browsers this will apply `z-order: 0` to ensure that clips are |
| /// applied correctly. Otherwise, the browser will refuse to clip its contents. |
| /// |
| /// Other possible fixes that were rejected: |
| /// |
| /// * Use 3D transform instead of 2D: this does not work because it causes text |
| /// blurriness: https://github.com/flutter/flutter/issues/32274 |
| void applyWebkitClipFix(html.Element containerElement) { |
| if (browserEngine == BrowserEngine.webkit) { |
| containerElement.style.zIndex = '0'; |
| } |
| } |
| |
| final ByteData _fontChangeMessage = JSONMessageCodec().encodeMessage(<String, dynamic>{'type': 'fontsChange'}); |
| |
| // Font load callbacks will typically arrive in sequence, we want to prevent |
| // sendFontChangeMessage of causing multiple synchronous rebuilds. |
| // This flag ensures we properly schedule a single call to framework. |
| bool _fontChangeScheduled = false; |
| |
| FutureOr<void> sendFontChangeMessage() async { |
| if (window._onPlatformMessage != null) |
| if (!_fontChangeScheduled) { |
| _fontChangeScheduled = true; |
| // Batch updates into next animationframe. |
| html.window.requestAnimationFrame((num _) { |
| _fontChangeScheduled = false; |
| window.invokeOnPlatformMessage( |
| 'flutter/system', |
| _fontChangeMessage, |
| (_) {}, |
| ); |
| }); |
| } |
| } |
| |
| // Stores matrix in a form that allows zero allocation transforms. |
| class _FastMatrix64 { |
| final Float64List matrix; |
| double transformedX = 0, transformedY = 0; |
| _FastMatrix64(this.matrix); |
| |
| void transform(double x, double y) { |
| transformedX = matrix[12] + (matrix[0] * x) + (matrix[4] * y); |
| transformedY = matrix[13] + (matrix[1] * x) + (matrix[5] * y); |
| } |
| } |
| |
| /// Roughly the inverse of [ui.Shadow.convertRadiusToSigma]. |
| /// |
| /// This does not inverse [ui.Shadow.convertRadiusToSigma] exactly, because on |
| /// the Web the difference between sigma and blur radius is different from |
| /// Flutter mobile. |
| double convertSigmaToRadius(double sigma) { |
| return sigma * 2.0; |
| } |