Create wrapper

[akaszynski]

Create a basic wrapper of ftetwild.

Resolves #1.

• Incorporate the same options as in main.cpp
• Incorporate pre-commit
• Add trusted provider to PyPi
• Upload beta to pypi
• Test installing from sdist
• Incorporate cibuildwheel

[akaszynski]

At the moment, works (barely). I can't seem to get the development install working, so you'll have to build the wheel and then install it for local testing. Here's the install command (run from pytetwild repo root directory)

pip install pipx
pipx run build --wheel
pip install dist/*

Right now it's just a raw wrapper. Work with it with:

import pyvista as pv
import numpy as np
from pytetwild import PyfTetWildWrapper

mesh = pv.Icosphere(nsub=0)
print(mesh)
# Extract vertices and faces from the bunny mesh
vertices = np.array(mesh.points, dtype=np.float64)  # Ensure correct data type
faces = np.array(mesh.faces.reshape((-1, 4))[:, 1:4], dtype=np.int32)

# Tetrahedralize the mesh using PyfTetWildWrapper
tetrahedral_mesh_vertices, tetrahedral_mesh_tetrahedra = PyfTetWildWrapper.tetrahedralize_mesh(vertices, faces)

# Convert tetrahedral mesh data back to PyVista mesh for visualization
# Note: PyVista's UnstructuredGrid expects cells to be defined by the number of points in the cell followed by the indices of the points in the cell
cells = np.hstack([np.full((tetrahedral_mesh_tetrahedra.shape[0], 1), 4, dtype=np.int32), tetrahedral_mesh_tetrahedra])

# Define cell types for the UnstructuredGrid, 10 represents tetrahedra
cell_types = np.full(tetrahedral_mesh_tetrahedra.shape[0], 10, dtype=np.uint8)

# Create a PyVista unstructured grid
tetrahedral_mesh = pv.UnstructuredGrid(cells, cell_types, tetrahedral_mesh_vertices)

# Visualize the tetrahedral mesh
tetrahedral_mesh.explode(1).plot(show_edges=True)