modules.models#

class modules.models.Models(*args: Any, **kwargs: Any)#

Bases: Context

class CPUMeshData(combined: ndarray, indices: ndarray, num_indices: int, material: int, path: Path, mesh: Any, aabb: tuple[ndarray, ndarray])#

Bases: object

CPU-side representation of a mesh. (used for loading)

Stores interleaved vertex data, indices, and metadata required for uploading the mesh to the GPU.

Parameters:

  • combined – Interleaved vertex attributes (pos, normal, uv, tangent, bitangent)

  • indices – Triangle index buffer (uint32)

  • num_indices – Total number of indices

  • material – Material ID assigned to the mesh

  • path – Source model path

  • mesh – Original imported mesh reference

__init__(combined: ndarray, indices: ndarray, num_indices: int, material: int, path: Path, mesh: Any, aabb: tuple[ndarray, ndarray]) → None#
aabb: tuple[ndarray, ndarray]#
combined: ndarray#
indices: ndarray#
material: int#
mesh: Any#
num_indices: int#
path: Path#
class Load(path: Path, material: int)#

Bases: object

Model load request descriptor.

Contains the source path and optional material override.

Parameters:

  • path – Path to the model file

  • material – Optional material override (-1 = auto)

__init__(path: Path, material: int) → None#
material: int#
path: Path#
class Mesh#

Bases: TypedDict

“Final representation a mesh (used for rendering)

aabb: tuple[ndarray, ndarray]#
baseVertex: int#
firstIndex: int#
material: int#
num_indices: int#
vao_simple: VAO#
__init__(context)#

Setup model buffers, containg mesh and material information required to render.

Parameters:

context (EmberEngine) – This is the main context of the application

compute_tangents_bitangents(vertices, tex_coords, indices)#

Compute tangent spaces for vertices

Parameters:

  • vertices (ndarray - shape 3.) – The vertices in the mesh

  • tex_coords (ndarray - shape 2.) – The texture coordinates

  • indices (ndarray - c.) – The indices for the mesh

Returns:

tangents, bitanges

Return type:

ndarray; shape 3, ndarray; shape 3

create_matrices(model)#

Collect local model matrices for nodes in this model

draw(model: Model, model_matrix: Matrix44, instant: bool = False, uuid=None) → None#

Begin drawing a model

Parameters:

  • model (Model) – The model object

  • model_matrix (matrix44) – The transformation model matrix, used along with view and projection matrices

  • instant (bool) – either collect the drawcalls and submit them in renderer.end_frame() or render now/nstantly

loadOrFind(path: Path, material: int = -1, lazy_load: bool = True) → int#

Load or find an model, implement find later

Parameters:

  • path (Path) – The path to the model file

  • material (int) – Could contain a material override if not -1

  • lazy (bool) – Lazy load models using threading

model_loader_thread()#

Worker thread for model loading and CPU pre-processing.

Pulls load requests from the queue, assimp loads and prepares models (tangents). Then outputs prepared meshes to the ready queue.

model_loader_thread_flush() → None#

Receives ready CPU prepared meshes from work thread and uploads them to the GPU.

static normalize(vectors)#

Normalize a vector array

prepare_mesh_cpu(mesh, path: Path, material: int) → CPUMeshData#

Convert a mesh into an CPUMeshData CPU-side format.

Generates missing attributes if needed and computes tangents and bitangents for normal mapping.

Parameters:

  • mesh – Imported mesh object

  • path (Path) – Source model path

  • material (int) – Material ID override (-1 = auto)

Returns:

CPU-side mesh data

Return type:

CPUMeshData

prepare_on_CPU(index: int, path: Path, material: int = -1) → None#

Load a model from disk and prepare all meshes on the CPU.

Returns a list of CPU-ready mesh data objects.

Parameters:

  • index (int) – Internal model index

  • path (Path) – Path to the model file

  • material (int) – Material ID override (-1 = auto)

Returns:

List of prepared CPU meshes

Return type:

list[CPUMeshData]

upload_to_GPU(index: int, cpu_meshes: list[CPUMeshData]) → bool#

Upload CPU-prepared meshes to the GPU. Then free the list

Parameters:

  • index (int) – Internal model index

  • cpu_meshes (list[CPUMeshData]) – List of CPU mesh data objects

Returns:

True if upload succeeded

Return type:

bool