1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528
//! Procedurally-generated meshes for rendering objects that are
//! specified geometrically, and have nontrivial numbers of vertices each,
//! such as a sphere or cylinder.
use std::sync::Arc;
use glam::{uvec3, vec3, Vec3, Vec3A};
use hexasphere::BaseShape;
use itertools::Itertools as _;
use ordered_float::NotNan;
use smallvec::smallvec;
use re_math::MeshGen;
use re_renderer::{
mesh::{self, GpuMesh, MeshError},
RenderContext,
};
use re_viewer_context::Cache;
// ----------------------------------------------------------------------------
/// Description of a mesh that can be procedurally generated.
///
/// Obtain the actual mesh by passing this to [`WireframeCache`] or [`SolidCache`].
#[derive(Debug, PartialEq, Eq, Hash, Clone, Copy)]
pub enum ProcMeshKey {
/// A unit cube, centered; its bounds are ±0.5.
Cube,
/// A sphere of unit radius.
///
/// The resulting mesh may be scaled to represent spheres and ellipsoids
/// of other sizes.
Sphere {
/// Number of triangle subdivisions to perform to create a finer, rounder mesh.
///
/// If this number is zero, then the “sphere” is an octahedron. Increasing it to N
/// breaks the edges of that octahedron into N segments. Around a great circle of the
/// sphere, there are (N + 1) × 4 segments.
subdivisions: usize,
/// If true, then when a wireframe mesh is generated, it includes only
/// the 3 axis-aligned “equatorial” circles, and not the full triangle mesh.
axes_only: bool,
},
/// A capsule; a cylinder with hemispherical end caps.
///
/// The capsule always has radius 1. It should be scaled to obtain the desired radius.
/// It always extends along the positive direction of the Z axis.
Capsule {
/// The length of the capsule; the distance between the centers of its endpoints.
/// This length must be non-negative.
//
// TODO(#1361): This is a bad approach to rendering capsules of arbitrary
// length, because it fills the cache with many distinct meshes.
// Instead, the renderers should be extended to support “bones” such that a mesh
// can have parts which are independently offset, thus allowing us to stretch a
// single sphere/capsule mesh into an arbitrary length and radius capsule.
// (Tapered capsules will still need distinct meshes.)
length: NotNan<f32>,
/// Number of triangle subdivisions to use to create a finer, rounder mesh.
///
/// The cylinder part of the capsule is approximated as a mesh with (N + 1) × 4
/// flat faces.
subdivisions: usize,
},
}
impl ProcMeshKey {
/// Returns the bounding box which can be computed from the mathematical shape,
/// without regard for its exact approximation as a mesh.
pub fn simple_bounding_box(&self) -> re_math::BoundingBox {
match self {
Self::Sphere {
subdivisions: _,
axes_only: _,
} => {
// sphere’s radius is 1, so its size is 2
re_math::BoundingBox::from_center_size(Vec3::splat(0.0), Vec3::splat(2.0))
}
Self::Cube => {
re_math::BoundingBox::from_center_size(Vec3::splat(0.0), Vec3::splat(1.0))
}
Self::Capsule {
subdivisions: _,
length,
} => re_math::BoundingBox::from_min_max(
Vec3::new(-1.0, -1.0, -1.0),
Vec3::new(1.0, 1.0, 1.0 + length.into_inner()),
),
}
}
}
/// A renderable mesh generated from a [`ProcMeshKey`] by the [`WireframeCache`],
/// which is to be drawn as lines rather than triangles.
#[derive(Debug)]
pub struct WireframeMesh {
#[allow(unused)]
pub bbox: re_math::BoundingBox,
#[allow(unused)]
pub vertex_count: usize,
/// Collection of line strips making up the wireframe.
///
/// TODO(kpreid): This should instead be a GPU buffer, but we don’t yet have a
/// `re_renderer::Renderer` implementation that takes instanced meshes and applies
/// the line shader to them, instead of doing immediate-mode accumulation of line strips.
pub line_strips: Vec<Vec<Vec3>>,
}
/// A renderable mesh generated from a [`ProcMeshKey`] by the [`SolidCache`],
/// which is to be drawn as triangles rather than lines.
///
/// This type is cheap to clone.
#[derive(Clone)]
pub struct SolidMesh {
#[allow(unused)]
pub bbox: re_math::BoundingBox,
/// Mesh to render. Note that its colors are set to black, so that the
/// `MeshInstance::additive_tint` can be used to set the color per instance.
pub gpu_mesh: Arc<GpuMesh>,
}
/// Errors that may arise from attempting to generate a mesh from a [`ProcMeshKey`].
///
/// Currently, this type is private because errors are only logged.
#[derive(thiserror::Error, Debug)]
#[non_exhaustive]
enum GenError {
/// The requested drawing primitive type (solid or wireframe) is not supported
/// for the given [`ProcMeshKey`],
#[error("creating a wireframe mesh is not supported")]
UnimplementedWireframe,
/// Either the GPU mesh could not be allocated,
/// or the generated mesh was not well-formed.
#[error(transparent)]
MeshProcessing(#[from] MeshError),
}
// ----------------------------------------------------------------------------
/// Cache for the computation of wireframe meshes from [`ProcMeshKey`]s.
/// These meshes may then be rendered as instances of the cached
/// mesh.
#[derive(Default)]
pub struct WireframeCache(ahash::HashMap<ProcMeshKey, Option<Arc<WireframeMesh>>>);
impl WireframeCache {
pub fn entry(
&mut self,
key: ProcMeshKey,
render_ctx: &RenderContext,
) -> Option<Arc<WireframeMesh>> {
re_tracing::profile_function!();
self.0
.entry(key)
.or_insert_with(|| {
re_tracing::profile_scope!("proc_mesh::WireframeCache(miss)", format!("{key:?}"));
re_log::trace!("Generating wireframe mesh {key:?}…");
match generate_wireframe(&key, render_ctx) {
Ok(mesh) => Some(Arc::new(mesh)),
Err(err) => {
re_log::warn!(
"Failed to generate mesh {key:?}: {}",
re_error::format_ref(&err)
);
None
}
}
})
.clone()
}
}
impl Cache for WireframeCache {
fn purge_memory(&mut self) {
self.0.clear();
}
fn as_any_mut(&mut self) -> &mut dyn std::any::Any {
self
}
}
/// Generate a wireframe mesh without caching.
///
/// Note: The unstructured error type here is used only for logging.
fn generate_wireframe(
key: &ProcMeshKey,
render_ctx: &RenderContext,
) -> Result<WireframeMesh, GenError> {
re_tracing::profile_function!();
// In the future, render_ctx will be used to allocate GPU memory for the mesh.
_ = render_ctx;
let mesh = match *key {
ProcMeshKey::Cube => {
let corners = [
vec3(-0.5, -0.5, -0.5),
vec3(-0.5, -0.5, 0.5),
vec3(-0.5, 0.5, -0.5),
vec3(-0.5, 0.5, 0.5),
vec3(0.5, -0.5, -0.5),
vec3(0.5, -0.5, 0.5),
vec3(0.5, 0.5, -0.5),
vec3(0.5, 0.5, 0.5),
];
let line_strips: Vec<Vec<Vec3>> = vec![
// bottom:
vec![
// bottom loop
corners[0b000],
corners[0b001],
corners[0b011],
corners[0b010],
corners[0b000],
// joined to top loop
corners[0b100],
corners[0b101],
corners[0b111],
corners[0b110],
corners[0b100],
],
// remaining side edges
vec![corners[0b001], corners[0b101]],
vec![corners[0b010], corners[0b110]],
vec![corners[0b011], corners[0b111]],
];
WireframeMesh {
bbox: key.simple_bounding_box(),
vertex_count: line_strips.iter().map(|v| v.len()).sum(),
line_strips,
}
}
ProcMeshKey::Sphere {
subdivisions,
axes_only,
} => {
let subdiv: hexasphere::Subdivided<(), OctahedronBase> =
hexasphere::Subdivided::new(subdivisions, |_| ());
let sphere_points = subdiv.raw_points();
let line_strips: Vec<Vec<Vec3>> = if axes_only {
let mut buffer: Vec<u32> = Vec::new();
subdiv.get_major_edges_line_indices(&mut buffer, 1, |v| v.push(0));
buffer
.split(|&i| i == 0)
.map(|strip| -> Vec<Vec3> {
strip
.iter()
.map(|&i| sphere_points[i as usize - 1].into())
.collect()
})
.collect()
} else {
subdiv
.get_all_line_indices(1, |v| v.push(0))
.split(|&i| i == 0)
.map(|strip| -> Vec<Vec3> {
strip
.iter()
.map(|&i| sphere_points[i as usize - 1].into())
.collect()
})
.collect()
};
WireframeMesh {
bbox: key.simple_bounding_box(),
vertex_count: line_strips.iter().map(|v| v.len()).sum(),
line_strips,
}
}
ProcMeshKey::Capsule {
length: _,
subdivisions: _,
} => {
// No visualizer asks for these yet, because they are unimplemented.
// Implementing them will require writing a new capsule wireframe algorithm
// that agrees with the solid algorithm.
return Err(GenError::UnimplementedWireframe);
}
};
Ok(mesh)
}
// ----------------------------------------------------------------------------
/// Cache for the computation of triangle meshes from [`ProcMeshKey`]s that depict the
/// shape as a solid object.
#[derive(Default)]
pub struct SolidCache(ahash::HashMap<ProcMeshKey, Option<SolidMesh>>);
impl SolidCache {
pub fn entry(&mut self, key: ProcMeshKey, render_ctx: &RenderContext) -> Option<SolidMesh> {
re_tracing::profile_function!();
self.0
.entry(key)
.or_insert_with(|| {
re_tracing::profile_scope!("proc_mesh::SolidCache(miss)", format!("{key:?}"));
re_log::trace!("Generating solid mesh {key:?}…");
match generate_solid(&key, render_ctx) {
Ok(mesh) => Some(mesh),
Err(err) => {
re_log::warn!(
"Failed to generate mesh {key:?}: {}",
re_error::format_ref(&err)
);
None
}
}
})
.clone()
}
}
impl Cache for SolidCache {
fn purge_memory(&mut self) {
self.0.clear();
}
fn as_any_mut(&mut self) -> &mut dyn std::any::Any {
self
}
}
/// Generate a solid triangle mesh without caching.
fn generate_solid(key: &ProcMeshKey, render_ctx: &RenderContext) -> Result<SolidMesh, GenError> {
re_tracing::profile_function!();
let mesh: mesh::CpuMesh = match *key {
ProcMeshKey::Cube => {
let mut mg = re_math::MeshGen::new();
mg.push_cube(Vec3::splat(0.5), re_math::IsoTransform::IDENTITY);
mesh_from_mesh_gen(format!("{key:?}").into(), mg, render_ctx)
}
ProcMeshKey::Sphere {
subdivisions,
axes_only: _, // no effect on solid mesh
} => {
let subdiv: hexasphere::Subdivided<(), OctahedronBase> =
hexasphere::Subdivided::new(subdivisions, |_| ());
let vertex_positions: Vec<Vec3> =
subdiv.raw_points().iter().map(|&p| p.into()).collect();
// A unit sphere's normals are its positions.
let vertex_normals = vertex_positions.clone();
let num_vertices = vertex_positions.len();
let triangle_indices = subdiv.get_all_indices();
let triangle_indices: Vec<glam::UVec3> = triangle_indices
.into_iter()
.tuples()
.map(|(i1, i2, i3)| glam::uvec3(i1, i2, i3))
.collect();
let materials = materials_for_uncolored_mesh(render_ctx, triangle_indices.len());
mesh::CpuMesh {
label: format!("{key:?}").into(),
// bytemuck is re-grouping the indices into triples without realloc
triangle_indices,
vertex_positions,
vertex_normals,
// Colors are black so that the instance `additive_tint` can set per-instance color.
vertex_colors: vec![re_renderer::Rgba32Unmul::BLACK; num_vertices],
vertex_texcoords: vec![glam::Vec2::ZERO; num_vertices],
materials,
}
}
ProcMeshKey::Capsule {
length,
subdivisions,
} => {
// Design note: there are two reasons why this uses `re_math` instead of `hexasphere`.
//
// First, `re_math` already has a capsule routine, whereas we'd have to postprocess the
// output of `hexasphere`.
//
// Second, one design perspective is that we should in the long run extend `re_math`
// to do *all* our mesh generation, and this is an experiment in that. How exactly that
// will handle wireframes is yet undecided.
let mg_subdivisions = (subdivisions + 1) * 4;
let mut mg = re_math::MeshGen::new();
mg.push_capsule(
1.0,
length.into_inner(),
mg_subdivisions,
mg_subdivisions,
// rotate from the Y axis (baked into MeshGen) onto the Z axis (our choice of
// default orientation, aligned with Rerun’s default of Z-up).
re_math::IsoTransform::from_quat(glam::Quat::from_rotation_x(
std::f32::consts::FRAC_PI_2,
)),
);
mesh_from_mesh_gen(format!("{key:?}").into(), mg, render_ctx)
}
};
mesh.sanity_check()?;
Ok(SolidMesh {
bbox: key.simple_bounding_box(),
gpu_mesh: Arc::new(GpuMesh::new(render_ctx, &mesh)?),
})
}
fn mesh_from_mesh_gen(
label: re_renderer::DebugLabel,
mg: MeshGen,
render_ctx: &RenderContext,
) -> mesh::CpuMesh {
let num_vertices = mg.positions.len();
let triangle_indices: Vec<glam::UVec3> = mg
.indices
.into_iter()
.tuples()
.map(|(i1, i2, i3)| uvec3(i1, i2, i3))
.collect();
let materials = materials_for_uncolored_mesh(render_ctx, triangle_indices.len());
mesh::CpuMesh {
label,
materials,
triangle_indices,
vertex_positions: mg.positions,
vertex_normals: mg.normals,
// Colors are black so that the instance `additive_tint` can set per-instance color.
vertex_colors: vec![re_renderer::Rgba32Unmul::BLACK; num_vertices],
vertex_texcoords: vec![glam::Vec2::ZERO; num_vertices],
}
}
fn materials_for_uncolored_mesh(
render_ctx: &RenderContext,
num_triangles: usize,
) -> smallvec::SmallVec<[mesh::Material; 1]> {
smallvec![mesh::Material {
label: "default material".into(),
index_range: 0..(num_triangles * 3) as u32,
albedo: render_ctx
.texture_manager_2d
.white_texture_unorm_handle()
.clone(),
albedo_factor: re_renderer::Rgba::BLACK,
}]
}
// ----------------------------------------------------------------------------
/// Base shape for [`hexasphere`]'s subdivision algorithm which is an octahedron
/// that is subdivided into a sphere mesh.
/// The value of this shape for us is that it has “equatorial” edges which are
/// perpendicular to the axes of the ellipsoid, which thus align with the quantities
/// the user actually specified (length on each axis), and can be usefully visualized
/// by themselves separately from the subdivision mesh.
///
/// TODO(kpreid): This would also make sense to contribute back to `hexasphere` itself.
#[derive(Clone, Copy, Debug, Default)]
struct OctahedronBase;
impl BaseShape for OctahedronBase {
fn initial_points(&self) -> Vec<Vec3A> {
vec![
Vec3A::NEG_X,
Vec3A::NEG_Y,
Vec3A::NEG_Z,
Vec3A::X,
Vec3A::Y,
Vec3A::Z,
]
}
fn triangles(&self) -> Box<[hexasphere::Triangle]> {
use hexasphere::Triangle;
const TRIANGLES: [Triangle; 8] = [
Triangle::new(0, 2, 1, 1, 4, 0), // -X-Y-Z face
Triangle::new(0, 1, 5, 0, 6, 3), // -X-Y+Z face
Triangle::new(0, 4, 2, 2, 5, 1), // -X+Y-Z face
Triangle::new(0, 5, 4, 3, 7, 2), // -X+Y+Z face
Triangle::new(3, 1, 2, 8, 4, 9), // +X-Y-Z face
Triangle::new(3, 5, 1, 11, 6, 8), // +X-Y+Z face
Triangle::new(3, 2, 4, 9, 5, 10), // +X+Y-Z face
Triangle::new(3, 4, 5, 10, 7, 11), // +X+Y+Z face
];
Box::new(TRIANGLES)
}
/// The octahedron has 12 edges, which we are arbitrarily numbering as follows:
///
/// 0. -X to -Y
/// 1. -X to -Z
/// 2. -X to +Y
/// 3. -X to +Z
/// 4. -Z to -Y
/// 5. -Z to +Y
/// 6. +Z to -Y
/// 7. +Z to +Y
/// 8. +X to -Y
/// 9. +X to -Z
/// 10. +X to +Y
/// 11. +X to +Z
const EDGES: usize = 12;
fn interpolate(&self, a: Vec3A, b: Vec3A, p: f32) -> Vec3A {
hexasphere::interpolation::geometric_slerp(a, b, p)
}
}