Struct OrientVerdict 

pub struct OrientVerdict {
    pub flipped: bool,
    pub all_closed: bool,
    pub all_orientable: bool,
    pub components: u32,
}

What orient_mesh_outward_verdict concluded about a mesh’s SURFACE TOPOLOGY, on top of whether it re-wound anything.

The orienter has to decide, per connected component, whether that component is CLOSED (every welded edge shared by exactly two triangles) and ORIENTABLE (no winding contradiction closes a cycle), because only such a component has a meaningful “outward” to flip toward. It then discarded the answer. Nothing downstream can recover it: by the time a mesh reaches the hasher or the FFI boundary the adjacency has not been rebuilt, and rebuilding it is this pass’s whole cost.

The consumer that needs it is a divergence-theorem volume. Over an OPEN surface that sum is not approximate, it is ARBITRARY — the boundary-loop flux grows with the distance to the reference point, so the “volume” of a sheet is whatever you referenced it to. There is no way to spot that from the number itself; it looks like an ordinary positive volume. This verdict is what lets a consumer refuse instead of guessing.

Fields

flipped: bool

At least one triangle’s winding was flipped, so any baked normals are stale. Exactly the bool orient_mesh_outward returns.

all_closed: bool

EVERY connected component is closed (no boundary edge, no non-manifold edge). False if there are no components at all — see Self::INDETERMINATE.

all_orientable: bool

EVERY connected component is orientable (consistent winding propagated without contradiction). Independent of all_closed: a Klein-bottle-like or self-touching component can be edge-manifold yet non-orientable.

components: u32

Connected components found (BFS over welded-edge adjacency). A single closed body is 1; a faceted brep holding two disjoint solids, or a solid plus the shell of its own cavity, is 2.

Implementations

impl OrientVerdict

pub const INDETERMINATE: Self

The verdict for a mesh the orienter refused to analyse: fewer than two triangles, or a malformed position/index buffer. Nothing is known, so nothing is claimed — all_closed is FALSE rather than vacuously true, so a consumer gating on it emits nothing. That is also the correct geometric answer: a mesh of under two triangles cannot enclose anything.

pub fn is_single_closed_solid(&self) -> bool

The one shape for which a divergence-theorem volume over this mesh is trustworthy: EXACTLY ONE closed, orientable component.

Why not the weaker “all components closed”. This pass flips each closed component so its OWN signed volume is positive. For two disjoint solids that is right and the sum is their combined volume. But for a solid whose cavity is modelled as a second, inner shell it is wrong in the worst way: the cavity shell is also made positive, so the sum reports outer + cavity where the true volume is outer − cavity. Telling those two arrangements apart needs a containment test this pass does not do — and the orientation it already applied has destroyed the sign that would have distinguished them. One component has no such ambiguity.

Trait Implementations

impl Clone for OrientVerdict

fn clone(&self) -> OrientVerdict

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Copy for OrientVerdict

impl Debug for OrientVerdict

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for OrientVerdict

fn default() -> OrientVerdict

Returns the “default value” for a type.

impl Eq for OrientVerdict

impl PartialEq for OrientVerdict

fn eq(&self, other: &OrientVerdict) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for OrientVerdict