Skip to content

Commit

Permalink
Simplify UI + plus button!
Browse files Browse the repository at this point in the history
  • Loading branch information
@Avarel
Avarel committed Sep 20, 2022
1 parent 466de3c commit 7ed18f6
Show file tree
Hide file tree
Showing 8 changed files with 292 additions and 224 deletions.
2 changes: 1 addition & 1 deletion libs/egui_dock
Submodule egui_dock updated from 01d4fe to a9dbb3
4 changes: 2 additions & 2 deletions src/compositor/bind.rs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -92,7 +92,7 @@ impl<'dev> BindingMapper<'dev> {
chunks,
output: {
let mut views = Vec::new();
views.resize_with(chunks as usize, || textures[0].make_view());
views.resize_with(chunks as usize, || textures[0].create_view());
views.into_boxed_slice()
},
map: HashMap::new(),
Expand All @@ -108,7 +108,7 @@ impl<'dev> BindingMapper<'dev> {
fn map_texture(&mut self, texture_index: usize) -> u32 {
let mlen = self.map.len() as u32;
*self.map.entry(texture_index).or_insert_with(|| {
self.output[mlen as usize] = self.textures[texture_index].make_view();
self.output[mlen as usize] = self.textures[texture_index].create_view();
mlen
})
}
Expand Down
94 changes: 56 additions & 38 deletions src/compositor/mod.rs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -149,7 +149,7 @@ impl<'dev> CompositorStage<'dev> {
Self {
bindings: CpuBuffers::new(target.dev.chunks),
buffers: GpuBuffers::new(&target.dev),
output: target.base_composite_texture(),
output: target.create_texture(),
}
}
}
Expand Down Expand Up @@ -182,6 +182,7 @@ impl<'dev> CompositorTarget<'dev> {
/// Initial indices of the 2 triangle strips
const INDICES: [u16; 4] = [0, 1, 2, 3];

/// Create a new compositor target.
pub fn new(dev: &'dev GpuHandle) -> Self {
let device = &dev.device;

Expand All @@ -190,7 +191,7 @@ impl<'dev> CompositorTarget<'dev> {
let vertex_buffer = device.create_buffer_init(&wgpu::util::BufferInitDescriptor {
label: Some("vertex_buffer"),
contents: bytemuck::cast_slice(&vertices),
usage: wgpu::BufferUsages::VERTEX,
usage: wgpu::BufferUsages::VERTEX | wgpu::BufferUsages::COPY_DST,
});

// Index draw buffer
Expand All @@ -211,28 +212,31 @@ impl<'dev> CompositorTarget<'dev> {
}
}

pub fn base_composite_texture(&self) -> GpuTexture {
/// Create an empty texture for this compositor target.
fn create_texture(&self) -> GpuTexture {
GpuTexture::empty_with_extent(&self.dev, self.dim.extent, GpuTexture::OUTPUT_USAGE)
}

pub fn flip_vertices(&mut self, flip_hv: (bool, bool)) {
/// Flip the vertex data's foreground UV of the compositor target.
pub fn flip_vertices(&mut self, horizontal: bool, vertical: bool) {
for v in &mut self.vertices {
v.fg_coords = [
if flip_hv.0 {
if horizontal {
1.0 - v.fg_coords[0]
} else {
v.fg_coords[0]
},
if flip_hv.1 {
if vertical {
1.0 - v.fg_coords[1]
} else {
v.fg_coords[1]
},
];
}
self.reload_vertices_buffer();
self.load_vertex_buffer();
}

/// Rotate the vertex data's foreground UV of the compositor target.
pub fn rotate_vertices(&mut self, ccw: bool) {
let temp = self.vertices[0].fg_coords;
if ccw {
Expand All @@ -246,9 +250,15 @@ impl<'dev> CompositorTarget<'dev> {
self.vertices[3].fg_coords = self.vertices[1].fg_coords;
self.vertices[1].fg_coords = temp;
}
self.reload_vertices_buffer();
self.load_vertex_buffer();
}

/// Transpose the dimensions of the compositor target's output.
pub fn transpose_dimensions(&mut self) -> bool {
self.set_dimensions(self.dim.height, self.dim.width)
}

/// Set the dimensions of the compositor target's output.
pub fn set_dimensions(&mut self, width: u32, height: u32) -> bool {
let buffer_dimensions = BufferDimensions::new(width, height);
if self.dim == buffer_dimensions {
Expand All @@ -271,21 +281,18 @@ impl<'dev> CompositorTarget<'dev> {
true
}

pub fn reload_vertices_buffer(&mut self) {
self.vertex_buffer =
self.dev
.device
.create_buffer_init(&wgpu::util::BufferInitDescriptor {
label: Some("vertex_buffer"),
contents: bytemuck::cast_slice(&self.vertices),
usage: wgpu::BufferUsages::VERTEX,
});
/// Load the GPU vertex buffer with updated data.
fn load_vertex_buffer(&mut self) {
self.dev
.queue
.write_buffer(&self.vertex_buffer, 0, bytemuck::cast_slice(&self.vertices));
}

/// Render composite layers using the compositor pipeline.
pub fn render(
&mut self,
compositor: &CompositorPipeline,
background: Option<[f32; 4]>,
pipeline: &CompositorPipeline,
bg: Option<[f32; 4]>,
layers: &[CompositeLayer],
textures: &[GpuTexture],
) {
Expand All @@ -306,18 +313,20 @@ impl<'dev> CompositorTarget<'dev> {
self.stages.push(CompositorStage::new(self));
}
count += self.render_stage(
compositor,
pipeline,
&mut encoder,
background,
bg,
stage_idx,
&layers[count..],
textures,
) as usize;
stage_idx += 1;
}

// Truncate and remove stages that might no longer be necessary.
self.stages.truncate(stage_idx);

// Copy the texture to the output of the target.
encoder.copy_texture_to_texture(
self.stages[stage_idx - 1].output.texture.as_image_copy(),
self.output_texture
Expand All @@ -332,19 +341,20 @@ impl<'dev> CompositorTarget<'dev> {
}));
}

/// Singular render pass. Many might be required to complete the compositing step.
fn render_stage(
&mut self,
compositor: &CompositorPipeline,
pipeline: &CompositorPipeline,
encoder: &mut CommandEncoder,
background: Option<[f32; 4]>,
bg: Option<[f32; 4]>,
stage_idx: usize,
composite_layers: &[CompositeLayer],
textures: &[GpuTexture],
) -> u32 {
let composite_view = if stage_idx > 0 {
self.stages[stage_idx - 1].output.make_view()
self.stages[stage_idx - 1].output.create_view()
} else {
self.base_composite_texture().make_view()
self.create_texture().create_view()
};

let stage = &mut self.stages[stage_idx];
Expand All @@ -358,7 +368,7 @@ impl<'dev> CompositorTarget<'dev> {
.dev
.device
.create_bind_group(&wgpu::BindGroupDescriptor {
layout: &compositor.blending_bind_group_layout,
layout: &pipeline.blending_bind_group_layout,
entries: &[
wgpu::BindGroupEntry {
binding: 0,
Expand Down Expand Up @@ -390,27 +400,28 @@ impl<'dev> CompositorTarget<'dev> {
label: Some("mixing_bind_group"),
});

let output_view = stage.output.make_view();
let output_view = stage.output.create_view();
let mut render_pass = encoder.begin_render_pass(&wgpu::RenderPassDescriptor {
label: None,
color_attachments: &[
// background color clear pass
Some(wgpu::RenderPassColorAttachment {
view: &output_view,
resolve_target: None,
ops: wgpu::Operations {
load: wgpu::LoadOp::Clear(
background
.map(|[r, g, b, _]| wgpu::Color {
r: f64::from(r),
g: f64::from(g),
b: f64::from(b),
a: 1.0,
})
.unwrap_or(wgpu::Color::TRANSPARENT),
bg.map(|[r, g, b, _]| wgpu::Color {
r: f64::from(r),
g: f64::from(g),
b: f64::from(b),
a: 1.0,
})
.unwrap_or(wgpu::Color::TRANSPARENT),
),
store: true,
},
}),
// compositing pass
Some(wgpu::RenderPassColorAttachment {
view: &output_view,
resolve_target: None,
Expand All @@ -424,13 +435,14 @@ impl<'dev> CompositorTarget<'dev> {
});

// Finish and set the render pass's binding groups and data
render_pass.set_pipeline(&compositor.render_pipeline);
render_pass.set_pipeline(&pipeline.render_pipeline);
// We use push constants for the binding count.
render_pass.set_push_constants(
wgpu::ShaderStages::FRAGMENT,
0,
&stage.bindings.count.to_ne_bytes(),
);
render_pass.set_bind_group(0, &compositor.constant_bind_group, &[]);
render_pass.set_bind_group(0, &pipeline.constant_bind_group, &[]);
render_pass.set_bind_group(1, &blending_bind_group, &[]);
render_pass.set_vertex_buffer(0, self.vertex_buffer.slice(..));
render_pass.set_index_buffer(self.index_buffer.slice(..), wgpu::IndexFormat::Uint16);
Expand All @@ -449,6 +461,7 @@ pub struct CompositorPipeline {
}

impl CompositorPipeline {
/// Create a new compositor pipeline.
pub fn new(dev: &GpuHandle) -> Self {
let device = &dev.device;

Expand Down Expand Up @@ -593,11 +606,16 @@ impl CompositorPipeline {
}
}

pub fn shader_load() -> wgpu::ShaderModuleDescriptor<'static> {
/// Load the shader.
fn shader_load() -> wgpu::ShaderModuleDescriptor<'static> {
// In release mode, the final binary includes the file directly so that
// the binary does not rely on the shader file being at a specific location.
#[cfg(not(debug_assertions))]
{
wgpu::include_wgsl!("../shader.wgsl")
}
// In debug mode, this reads directly from a file so that recompilation
// will not be necessary in the event that only the shader file changes.
#[cfg(debug_assertions)]
{
wgpu::ShaderModuleDescriptor {
Expand Down
12 changes: 7 additions & 5 deletions src/compositor/tex.rs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -51,7 +51,7 @@ impl GpuTexture {
}

/// Make a texture view of this GPU texture.
pub fn make_view(&self) -> wgpu::TextureView {
pub fn create_view(&self) -> wgpu::TextureView {
self.texture
.create_view(&wgpu::TextureViewDescriptor::default())
}
Expand All @@ -66,7 +66,7 @@ impl GpuTexture {
encoder.begin_render_pass(&wgpu::RenderPassDescriptor {
label: None,
color_attachments: &[Some(wgpu::RenderPassColorAttachment {
view: &self.make_view(),
view: &self.create_view(),
resolve_target: None,
ops: wgpu::Operations {
load: wgpu::LoadOp::Clear(color),
Expand Down Expand Up @@ -116,7 +116,7 @@ impl GpuTexture {
/// `dev` should be the same device that created this texture
/// in the first place.
pub fn clone(&self, dev: &GpuHandle) -> Self {
let c = Self::empty_with_extent(
let clone = Self::empty_with_extent(
dev,
self.size,
Self::OUTPUT_USAGE | wgpu::TextureUsages::COPY_DST,
Expand All @@ -128,12 +128,12 @@ impl GpuTexture {
// Copy the data from the texture to the buffer
encoder.copy_texture_to_texture(
self.texture.as_image_copy(),
c.texture.as_image_copy(),
clone.texture.as_image_copy(),
self.size,
);
encoder.finish()
}));
c
clone
}

/// Export the texture to the given path.
Expand All @@ -151,6 +151,7 @@ impl GpuTexture {
mapped_at_creation: false,
});

// Copy the texture to the output buffer
dev.queue.submit(Some({
let mut encoder = dev
.device
Expand Down Expand Up @@ -195,6 +196,7 @@ impl GpuTexture {

let buffer = image::imageops::crop_imm(&buffer, 0, 0, dim.width, dim.height).to_image();

eprintln!("Saving the file to {}", path.display());
tokio::task::spawn_blocking(move || buffer.save(path))
.await
.unwrap()
Expand Down
6 changes: 2 additions & 4 deletions src/gui/canvas.rs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,6 @@ use egui::*;
/// 2D bounding box of f64 precision.
/// The range of data values we show.
#[derive(Clone, Copy, PartialEq, Debug)]
#[cfg_attr(feature = "serde", derive(serde::Deserialize, serde::Serialize))]
struct CanvasViewBounds {
pub(crate) min: [f32; 2],
pub(crate) max: [f32; 2],
Expand Down Expand Up @@ -284,7 +283,6 @@ impl From<bool> for AutoBounds {
}

/// Information about the plot that has to persist between frames.
#[cfg_attr(feature = "serde", derive(serde::Deserialize, serde::Serialize))]
#[derive(Clone)]
struct ViewMemory {
auto_bounds: AutoBounds,
Expand Down Expand Up @@ -573,7 +571,7 @@ impl CanvasView {
};
memory.store(ui.ctx(), plot_id);

// let response = response.on_hover_cursor(CursorIcon::Crosshair);
let response = response.on_hover_cursor(CursorIcon::Crosshair);

InnerResponse {
inner: (),
Expand Down Expand Up @@ -665,8 +663,8 @@ impl PreparedView {
.paint_at(&mut plot_ui, rect);
}

let painter = plot_ui.painter();
if self.show_extended_crosshair {
let painter = plot_ui.painter();
if let Some(pointer) = response.hover_pos() {
painter.vline(
pointer.x,
Expand Down
Loading

0 comments on commit 7ed18f6

Please sign in to comment.