Struct Renderer 

pub struct Renderer { /* private fields */ }

Implementations

impl Renderer

pub fn new( device: Device, queue: Queue, format: TextureFormat, width: u32, height: u32, ) -> Self

pub fn resize(&mut self, width: u32, height: u32)

pub fn camera_mut(&mut self) -> &mut Camera

pub fn camera_ref(&self) -> &Camera

pub fn set_fade(&mut self, fade: f32)

pub fn configure_surface( &self, surface: &Surface<'_>, config: &SurfaceConfiguration, )

Reconfigure a surface owned by the caller (e.g. GraphView) using this renderer’s device. GraphView owns surface/config but not the wgpu Device (that lives inside Renderer), so surface reconfiguration on resize is routed through here to keep device ownership in one place.

pub fn max_surface_dimension(&self) -> u32

The largest surface dimension this device will accept. Configuring a surface beyond it is a validation error, which on wasm aborts the whole instance — see fit_within_max_dimension.

pub fn set_nodes(&mut self, nodes: &[NodeInstanceData])

pub fn set_labels(&mut self, labels: &[LabelDraw<'_>])

Record this frame’s label survivors; shapes only labels not already in the shaped-buffer cache.

pub fn set_pixel_ratio(&mut self, ratio: f32)

Set the host’s device pixel ratio (physical pixels per CSS pixel).

The caller is responsible for sizing the canvas backing store and calling resize with the resulting physical size; this only tells the engine how to interpret the sizes it states in pixels itself. See Camera::pixel_ratio.

pub fn pixel_ratio(&self) -> f32

pub fn set_edges(&mut self, edges: &[EdgeLineData])

pub fn set_halos(&mut self, halos: &[HaloInstanceData])

pub fn tick(&mut self, dt_ms: f32, dur_ms: f32)

Advance the camera glide by dt_ms (paced by dur_ms).

pub fn glide_bounds(&mut self, min: [f32; 2], max: [f32; 2])

Glide the camera to frame world-AABB [min,max].

pub fn fit_view(&mut self)

Fit the camera to the current node set (including radii). No-op if empty.

Fits the bounds exactly, with no lower limit on the framed region — so a one-node set zooms until that node fills the viewport. Callers framing whatever content happens to be loaded almost always want fit_view_at_least instead.

pub fn fit_view_at_least(&mut self, min_extent: f32)

Fit the camera to the current node set, never framing a region tighter than min_extent world units on either axis. No-op if empty.

pub fn render(&mut self, view: &TextureView) -> Result<(), String>

Clear the target, then draw edges beneath nodes, with labels layered on top.