Struct WgpuRenderer 

pub struct WgpuRenderer { /* private fields */ }

Renderer runtime sharing one explicit spatialrust-gpu wgpu device.

Implementations

impl WgpuRenderer

pub fn render_headless( &self, geometry: &GpuGeometry, options: &RenderOptions, ) -> RenderResult<HeadlessRender>

Renders uploaded geometry into device-resident color and depth textures.

The method never reads the target back to the host. Its only host/device crossing is the returned, byte-accounted uniform upload.

pub fn initialized_render_pipeline_count(&self) -> usize

Number of initialized render-pipeline variants.

pub fn readback_rgba( &self, target: &GpuRenderTarget, ) -> RenderResult<ReadbackImage>

Explicitly reads one device-resident target into tightly packed RGBA8.

pub fn pick_point( &self, target: &GpuRenderTarget, x: u32, y: u32, ) -> RenderResult<PickResult>

Explicitly reads the point ID at one target pixel.

impl WgpuRenderer

pub fn new(runtime: Arc<WgpuRuntime>) -> Self

Creates a renderer on a caller-selected wgpu runtime.

pub const fn device_identity(&self) -> &DeviceIdentity

Identity of this exact renderer runtime.

pub fn cached_buffer_count(&self) -> usize

Number of buffers currently retained for reuse.

pub fn upload( &self, primitive: VisualPrimitive<'_>, ) -> RenderResult<(GpuGeometry, TransferReceipt)>

Explicitly uploads one borrowed visual primitive.

Structure-of-arrays positions are packed into a vertex buffer as part of this named operation. The returned receipt records every host/device crossing and its exact logical byte count.