Expand description
Source-format loaders and neutral decoded data for applications built on
viewport-lib.
This crate owns file-format decoding and source-agnostic data structures.
It does not depend on viewport-lib runtime or upload types directly.
§Coordinate convention
Decoded scene data is right-handed Z-up. Source formats that store data in a different frame (e.g. glTF, which is right-handed Y-up) are reoriented at load. Vertex positions, normals, tangents, mesh transforms, skeleton inverse-bind matrices, and animation samples are all rotated once during decoding; downstream consumers do not need to re-rotate.
Re-exports§
pub use error::IoError;pub use types::AnimationChannel;pub use types::AnimationClip;pub use types::AnimationInterpolation;pub use types::AnimationSampler;pub use types::AnimationTrack;pub use types::AnimationTrackValues;pub use types::AttributeData;pub use types::AttributeDomain;pub use types::AttributeValues;pub use types::CELL_SENTINEL;pub use types::DecodedDataSet;pub use types::GaussianSplatSet;pub use types::HdrImageData;pub use types::Joint;pub use types::MAX_JOINTS;pub use types::MaterialData;pub use types::PointSet;pub use types::RasterImageData;pub use types::SceneData;pub use types::SceneMesh;pub use types::ShDegree;pub use types::Skeleton;pub use types::SkinWeights;pub use types::SparseGrid;pub use types::StructuredVolume;pub use types::SurfaceMesh;pub use types::TextureSource;pub use types::VolumeGridGeometry;pub use types::VolumeMesh;