Add compute pipeline builder

src/lib.rs

@@ -8,7 +8,7 @@ pub use self::{
     bind_group::{BindGroupBuilder, BindGroupLayoutBuilder},
     buffer::{BufferExt, BulkBufferBuilder},
     context::{Context, ContextBuilder},
-    pipeline::RenderPipelineBuilder,
+    pipeline::{ComputePipelineBuilder, RenderPipelineBuilder},
     texture::{Texture, TextureBuilder},
 };
 

src/pipeline.rs

@@ -106,3 +106,54 @@ impl<'a> RenderPipelineBuilder<'a> {
             })
     }
 }
+
+pub struct ComputePipelineBuilder<'a> {
+    label: &'a str,
+    shader: &'a wgpu::ShaderModule,
+    layout_descriptor: Option<wgpu::PipelineLayoutDescriptor<'a>>,
+}
+
+impl<'a> ComputePipelineBuilder<'a> {
+    pub fn new(label: &'a str, shader: &'a wgpu::ShaderModule) -> Self {
+        Self {
+            label,
+            shader,
+            layout_descriptor: None,
+        }
+    }
+
+    pub fn with_layout(
+        mut self,
+        label: &'a str,
+        bind_group_layouts: &'a [&wgpu::BindGroupLayout],
+        push_constant_ranges: &'a [wgpu::PushConstantRange],
+    ) -> Self {
+        let layout_descriptor = wgpu::PipelineLayoutDescriptor {
+            label: Some(label),
+            bind_group_layouts,
+            push_constant_ranges,
+        };
+
+        self.layout_descriptor = Some(layout_descriptor);
+        self
+    }
+
+    pub fn build(self, context: &Context) -> wgpu::ComputePipeline {
+        let raw_layout;
+        let layout = if let Some(descriptor) = self.layout_descriptor {
+            raw_layout = context.device.create_pipeline_layout(&descriptor);
+            Some(&raw_layout)
+        } else {
+            None
+        };
+
+        context
+            .device
+            .create_compute_pipeline(&wgpu::ComputePipelineDescriptor {
+                label: Some(self.label),
+                layout,
+                module: self.shader,
+                entry_point: "compute",
+            })
+    }
+}