fastofreader 0.2.0

Fast numerical operations implemented in Cython

fastofreader

Fast readers for OpenFOAM mesh files implemented in Cython. Currently supports:

• faces: reads integer index groups from OpenFOAM faces files
• points: reads triplets of floats from OpenFOAM points files
• boundary: reads boundary metadata (nFaces, startFace) from OpenFOAM boundary files

Designed to be simple, dependency-free, and fast.

Author: Tomas Vicar [e-mail]

Installation

For development (editable):

pip install -e .

This builds the Cython extension once and links it in editable mode.

Usage

from fastofreader import (
	read_faces,
	read_points,
	read_faces_from_text,
	read_points_from_text,
	read_boundary,
)

# Faces from file
faces = read_faces("faces_test_file")
print(len(faces), faces[0][:4])

# Points from file
pts = read_points("points_test_file")
print(len(pts), pts[0])

# Boundary file
bounds = read_boundary("boundary_test_file")
print(list(bounds.items())[:1])

# In-memory variants
faces2 = read_faces_from_text("(1 2 3) (4, 5, 6)")
pts2 = read_points_from_text("(0 0 0) (1e-5 0 0)")

API

• read_faces(path: str) -> list[list[int]]
  • Reads all integer groups enclosed in parentheses from a file. Returns a list of groups.
• read_faces_from_text(text: str) -> list[list[int]]
  • Same as above but takes a string input.
• read_points(path: str) -> list[list[float]]
  • Reads point coordinates as triplets (x y z) from a file. Returns a list of [x, y, z].
• read_points_from_text(text: str) -> list[list[float]]
  • Same as above but takes a string input.
• read_boundary(path: str) -> dict[str, dict[str, int]]
  • Parses OpenFOAM boundary file into {name: {"nFaces": int, "startFace": int}}.

Notes:

• Parsers ignore the OpenFOAM header and the outer count/parenthesis wrapper.
• Faces parser accepts separators as whitespace, commas, or semicolons.

Performance tips

• Prefer file-based functions (read_faces, read_points) to avoid extra encoding/decoding.
• Use CPython 3.10+ for best performance.

Development

Requirements are declared in pyproject.toml and built via setuptools. Common tasks:

# Editable install
python -m pip install -e .

# Run tests
pytest -q

License

MIT

Instalace (vývoj)

pip install -e .