1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328
// Copyright 2006 The Android Open Source Project
// Copyright 2020 Yevhenii Reizner
//
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
//! A [tiny-skia](https://github.com/RazrFalcon/tiny-skia) Bezier path implementation.
//!
//! Provides a memory-efficient Bezier path container, path builder, path stroker and path dasher.
//!
//! Also provides some basic geometry types, but they will be moved to an external crate eventually.
//!
//! Note that all types use single precision floats (`f32`), just like [Skia](https://skia.org/).
#![no_std]
#![warn(missing_docs)]
#![warn(missing_copy_implementations)]
#![warn(missing_debug_implementations)]
#![allow(clippy::approx_constant)]
#![allow(clippy::collapsible_if)]
#![allow(clippy::eq_op)]
#![allow(clippy::excessive_precision)]
#![allow(clippy::identity_op)]
#![allow(clippy::manual_range_contains)]
#![allow(clippy::neg_cmp_op_on_partial_ord)]
#![allow(clippy::too_many_arguments)]
#![allow(clippy::upper_case_acronyms)]
#![allow(clippy::wrong_self_convention)]
#[cfg(not(any(feature = "std", feature = "no-std-float")))]
compile_error!("You have to activate either the `std` or the `no-std-float` feature.");
#[cfg(feature = "std")]
extern crate std;
extern crate alloc;
mod dash;
mod f32x2_t;
mod f32x4_t;
mod floating_point;
mod path;
mod path_builder;
pub mod path_geometry;
mod rect;
mod scalar;
mod size;
mod stroker;
mod transform;
pub use dash::StrokeDash;
pub use f32x2_t::f32x2;
pub use floating_point::*;
pub use path::*;
pub use path_builder::*;
pub use rect::*;
pub use scalar::*;
pub use size::*;
pub use stroker::*;
pub use transform::*;
/// An integer length that is guarantee to be > 0
type LengthU32 = core::num::NonZeroU32;
/// A point.
///
/// Doesn't guarantee to be finite.
#[allow(missing_docs)]
#[repr(C)]
#[derive(Copy, Clone, PartialEq, Default, Debug)]
pub struct Point {
pub x: f32,
pub y: f32,
}
impl From<(f32, f32)> for Point {
#[inline]
fn from(v: (f32, f32)) -> Self {
Point { x: v.0, y: v.1 }
}
}
impl Point {
/// Creates a new `Point`.
pub fn from_xy(x: f32, y: f32) -> Self {
Point { x, y }
}
/// Creates a new `Point` from `f32x2`.
pub fn from_f32x2(r: f32x2) -> Self {
Point::from_xy(r.x(), r.y())
}
/// Converts a `Point` into a `f32x2`.
pub fn to_f32x2(&self) -> f32x2 {
f32x2::new(self.x, self.y)
}
/// Creates a point at 0x0 position.
pub fn zero() -> Self {
Point { x: 0.0, y: 0.0 }
}
/// Returns true if x and y are both zero.
pub fn is_zero(&self) -> bool {
self.x == 0.0 && self.y == 0.0
}
/// Returns true if both x and y are measurable values.
///
/// Both values are other than infinities and NaN.
pub fn is_finite(&self) -> bool {
(self.x * self.y).is_finite()
}
/// Checks that two `Point`s are almost equal.
pub(crate) fn almost_equal(&self, other: Point) -> bool {
!(*self - other).can_normalize()
}
/// Checks that two `Point`s are almost equal using the specified tolerance.
pub(crate) fn equals_within_tolerance(&self, other: Point, tolerance: f32) -> bool {
(self.x - other.x).is_nearly_zero_within_tolerance(tolerance)
&& (self.y - other.y).is_nearly_zero_within_tolerance(tolerance)
}
/// Scales (fX, fY) so that length() returns one, while preserving ratio of fX to fY,
/// if possible.
///
/// If prior length is nearly zero, sets vector to (0, 0) and returns
/// false; otherwise returns true.
pub fn normalize(&mut self) -> bool {
self.set_length_from(self.x, self.y, 1.0)
}
/// Sets vector to (x, y) scaled so length() returns one, and so that (x, y)
/// is proportional to (x, y).
///
/// If (x, y) length is nearly zero, sets vector to (0, 0) and returns false;
/// otherwise returns true.
pub fn set_normalize(&mut self, x: f32, y: f32) -> bool {
self.set_length_from(x, y, 1.0)
}
pub(crate) fn can_normalize(&self) -> bool {
self.x.is_finite() && self.y.is_finite() && (self.x != 0.0 || self.y != 0.0)
}
/// Returns the Euclidean distance from origin.
pub fn length(&self) -> f32 {
let mag2 = self.x * self.x + self.y * self.y;
if mag2.is_finite() {
mag2.sqrt()
} else {
let xx = f64::from(self.x);
let yy = f64::from(self.y);
(xx * xx + yy * yy).sqrt() as f32
}
}
/// Scales vector so that distanceToOrigin() returns length, if possible.
///
/// If former length is nearly zero, sets vector to (0, 0) and return false;
/// otherwise returns true.
pub fn set_length(&mut self, length: f32) -> bool {
self.set_length_from(self.x, self.y, length)
}
/// Sets vector to (x, y) scaled to length, if possible.
///
/// If former length is nearly zero, sets vector to (0, 0) and return false;
/// otherwise returns true.
pub fn set_length_from(&mut self, x: f32, y: f32, length: f32) -> bool {
set_point_length(self, x, y, length, &mut None)
}
/// Returns the Euclidean distance from origin.
pub fn distance(&self, other: Point) -> f32 {
(*self - other).length()
}
/// Returns the dot product of two points.
pub fn dot(&self, other: Point) -> f32 {
self.x * other.x + self.y * other.y
}
/// Returns the cross product of vector and vec.
///
/// Vector and vec form three-dimensional vectors with z-axis value equal to zero.
/// The cross product is a three-dimensional vector with x-axis and y-axis values
/// equal to zero. The cross product z-axis component is returned.
pub fn cross(&self, other: Point) -> f32 {
self.x * other.y - self.y * other.x
}
pub(crate) fn distance_to_sqd(&self, pt: Point) -> f32 {
let dx = self.x - pt.x;
let dy = self.y - pt.y;
dx * dx + dy * dy
}
pub(crate) fn length_sqd(&self) -> f32 {
self.dot(*self)
}
/// Scales Point in-place by scale.
pub fn scale(&mut self, scale: f32) {
self.x *= scale;
self.y *= scale;
}
pub(crate) fn scaled(&self, scale: f32) -> Self {
Point::from_xy(self.x * scale, self.y * scale)
}
pub(crate) fn swap_coords(&mut self) {
core::mem::swap(&mut self.x, &mut self.y);
}
pub(crate) fn rotate_cw(&mut self) {
self.swap_coords();
self.x = -self.x;
}
pub(crate) fn rotate_ccw(&mut self) {
self.swap_coords();
self.y = -self.y;
}
}
// We have to worry about 2 tricky conditions:
// 1. underflow of mag2 (compared against nearlyzero^2)
// 2. overflow of mag2 (compared w/ isfinite)
//
// If we underflow, we return false. If we overflow, we compute again using
// doubles, which is much slower (3x in a desktop test) but will not overflow.
fn set_point_length(
pt: &mut Point,
mut x: f32,
mut y: f32,
length: f32,
orig_length: &mut Option<f32>,
) -> bool {
// our mag2 step overflowed to infinity, so use doubles instead.
// much slower, but needed when x or y are very large, other wise we
// divide by inf. and return (0,0) vector.
let xx = x as f64;
let yy = y as f64;
let dmag = (xx * xx + yy * yy).sqrt();
let dscale = length as f64 / dmag;
x *= dscale as f32;
y *= dscale as f32;
// check if we're not finite, or we're zero-length
if !x.is_finite() || !y.is_finite() || (x == 0.0 && y == 0.0) {
*pt = Point::zero();
return false;
}
let mut mag = 0.0;
if orig_length.is_some() {
mag = dmag as f32;
}
*pt = Point::from_xy(x, y);
if orig_length.is_some() {
*orig_length = Some(mag);
}
true
}
impl core::ops::Neg for Point {
type Output = Point;
fn neg(self) -> Self::Output {
Point {
x: -self.x,
y: -self.y,
}
}
}
impl core::ops::Add for Point {
type Output = Point;
fn add(self, other: Point) -> Self::Output {
Point::from_xy(self.x + other.x, self.y + other.y)
}
}
impl core::ops::AddAssign for Point {
fn add_assign(&mut self, other: Point) {
self.x += other.x;
self.y += other.y;
}
}
impl core::ops::Sub for Point {
type Output = Point;
fn sub(self, other: Point) -> Self::Output {
Point::from_xy(self.x - other.x, self.y - other.y)
}
}
impl core::ops::SubAssign for Point {
fn sub_assign(&mut self, other: Point) {
self.x -= other.x;
self.y -= other.y;
}
}
impl core::ops::Mul for Point {
type Output = Point;
fn mul(self, other: Point) -> Self::Output {
Point::from_xy(self.x * other.x, self.y * other.y)
}
}
impl core::ops::MulAssign for Point {
fn mul_assign(&mut self, other: Point) {
self.x *= other.x;
self.y *= other.y;
}
}