Struct ShapeCache 

pub struct ShapeCache { /* private fields */ }

Implementations

impl ShapeCache

pub fn new(capacity: usize) -> ShapeCache

pub fn growing(capacity: usize, max_capacity: usize) -> ShapeCache

A cache that may grow from capacity up to max_capacity once all of its entries are live, i.e. every one was touched within the last capacity lookups. Evicting an on-screen shape forces the next frame to re-tessellate it, so growing once is the cheaper response to a working set that outgrew the cache.

pub fn capacity(&self) -> usize

pub fn max_capacity(&self) -> usize

pub fn len(&self) -> usize

pub fn is_empty(&self) -> bool

pub fn stats(&self) -> ShapeStats

pub fn reset_stats(&self)

pub fn clear(&self)

pub fn fill( &self, path: &Path, rule: FillRule, opts: RasterOptions, ) -> Rc<Shape>

pub fn fill_styled(&self, path: &Path, style: FillStyle) -> Rc<Shape>

pub fn fill_offset( &self, path: &Path, rule: FillRule, opts: RasterOptions, offset: Vec2, ) -> Rc<Shape>

pub fn stroke( &self, path: &Path, style: &StrokeStyle, opts: RasterOptions, ) -> Rc<Shape>

pub fn stroke_offset( &self, path: &Path, style: &StrokeStyle, opts: RasterOptions, offset: Vec2, ) -> Rc<Shape>

pub fn fill_contours( &self, contours: &[Contour], rule: FillRule, opts: RasterOptions, ) -> Rc<Shape>

pub fn disk(&self, center: Point, radius: f32) -> Rc<Shape>

pub fn rounded_rect(&self, rect: Rect, radius: f32) -> Rc<Shape>

pub fn rect(&self, rect: Rect) -> Rc<Shape>

pub fn begin_prefetch(&self)

Starts a prefetch pass.

While the pass is active, cache misses are recorded instead of tessellated, and each miss returns an empty placeholder Shape. The caller is expected to paint again after ShapeCache::tessellate_prefetch, so the placeholders never reach the screen. Do not retain placeholder shapes across the pass.

pub fn end_prefetch(&self) -> usize

Ends a prefetch pass and returns how many distinct shapes missed the cache. Nothing has been tessellated yet.

pub fn tessellate_prefetch(&self) -> usize

Tessellates the shapes recorded by the last prefetch pass and inserts them into the cache, returning how many were built.

Big batches are split across worker threads; each worker builds its shapes with parallel(false) because the batch already keeps the cores busy. Small batches stay on the calling thread, and a lone shape keeps its own band-parallel rasterizer. Results are inserted in request order so a batch produces the same spans every time.

Trait Implementations

impl Debug for ShapeCache

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl Default for ShapeCache

fn default() -> ShapeCache

Returns the “default value” for a type.