Comprehensive Guide to Python File Types: From .py to .pyd, How Many Have You Encountered?

Hello everyone, I am Programmer Wan Feng.

Recently, with the rise of AI programming, the popularity of Python has decreased, and I finally have time to focus on the Python topics I have always been interested in.

My exploration of Python is mainly based on two books:

、

The deeper I understand Python’s advanced syntax, the more I can appreciate the intricacies of its design.

Today, let’s talk about the various file types in the Python world.

“

As a Python developer, you must often deal with .py files. However, there are many other important file types in the Python ecosystem, each with its unique purpose.

📁 Core Python File Types

File Type	Main Purpose	Readable	Generation Method
.py	Python Source Code	✅ Yes	Manually Created
.pyc	Compiled Bytecode	❌ No	Automatically Generated by Python
.pyo	Optimized Bytecode	❌ No	Generated by Python with -O flag
.pyd	Windows Dynamic Link Library	❌ No	Compiled from Cython/C Extensions
.so	Linux/Mac Dynamic Library	❌ No	Compiled from Cython/C Extensions
.pyw	Console-less Python Script	✅ Yes	Manually Created
.pyx	Cython Source Code	✅ Yes	Manually Created

There are some file types you may not have encountered for two reasons:

You may not have such complex work scenarios, such as the use of .so files.
Some types are automatically hidden by editors, such as .pyc files.

🔧 Detailed Analysis

1. .py – Python Source Code File

This is the most common Python file, containing human-readable Python code.

# hello.py
def greet(name):
    return f"Hello, {name}!"

if __name__ == "__main__":
    print(greet("Python Developer: Programmer Wan Feng"))

2. .pyc – Compiled Bytecode File

The Python interpreter compiles .py files into bytecode to speed up subsequent execution.

Generation Method：

# Python automatically generates .pyc files in the __pycache__ directory
python -m py_compile hello.py

File Structure：

Located in __pycache__ directory
Naming format: hello.cpython-39.pyc
Contains Python bytecode, not machine code

3. .pyd – Windows Dynamic Link Library

.pyd files are essentially DLL files but can be directly imported by Python.

“

This type can be understood as the jar package format in Java, which can be packaged and shared with others.

Creation Example (using Cython):

# Install Cython
pip install cython

# Create Cython file
# hello.pyx
def cython_greet(name):
    return f"Hello from Cython, {name}!"

# setup.py
from setuptools import setup
from Cython.Build import cythonize

setup(ext_modules=cythonize("hello.pyx"))

# Compile to generate .pyd
python setup.py build_ext --inplace

Using .pyd File：

# Import like a regular module
import hello
print(hello.cython_greet("Programmer Wan Feng"))

4. .pyx – Cython Source Code File

Cython is a superset of Python that allows writing C extensions.

# fastmath.pyx
def fibonacci(int n):
    cdef int i
    cdef double a = 0.0, b = 1.0
    for i in range(n):
        a, b = b, a + b
    return a

5. .pyw – Console-less Script

On Windows, .pyw files do not display a console window when run, making them suitable for GUI applications.

# my_app.pyw
import tkinter as tk

root = tk.Tk()
root.title("Console-less Application")
root.mainloop()

📦 Package and Distribution Files

Package Related Files

File Type	Purpose
__init__.py	Package initialization file (optional in Python 3.3+)
main.py	Entry point when the package is executed as a script

Distribution and Installation

File Type	Purpose
.whl	Distribution format for Python packages (wheel)
.egg	Legacy package distribution format
setup.py	Package installation script
setup.cfg	Package configuration
pyproject.toml	Modern package configuration
requirements.txt	Dependency list

🛠️ Configuration File Types

Project Configuration

# pyproject.toml (modern Python project)
[build-system]
requires = ["setuptools", "wheel"]

[project]
name = "my-project"
version = "0.1.0"

# setup.cfg (traditional configuration)
[metadata]
name = my-project
version = 0.1.0

# requirements.txt
requests>=2.25.0
pandas>=1.3.0

Environment and Tool Configuration

File Type	Purpose
.python-version	pyenv version file
Pipfile	pipenv dependency management
Pipfile.lock	Dependency lock file
environment.yml	conda environment configuration

🔍 Special Purpose Files

1. .pyi – Stub Files

Used for type hints, does not contain implementation code.

# math.pyi
def sqrt(x: float) -> float: ...
def pow(x: float, y: float) -> float: ...

2. .pth – Path Configuration Files

Add custom paths to the Python path.

# my_paths.pth
/home/user/my_python_libs
../relative/path/to/modules

3. .pyz – Self-contained Applications

Zip applications containing all dependencies.

# Create .pyz file
python -m zipapp my_app -o app.pyz

# Run
python app.pyz

💻 Development Tool Files

Testing Related

# test_example.py
import pytest

def test_addition():
    assert 1 + 1 == 2

# conftest.py (pytest configuration)
import pytest

@pytest.fixture
def sample_data():
    return {"key": "value"}

Code Quality

File Type	Purpose
.pylintrc	Pylint configuration
.flake8	Flake8 configuration
.coveragerc	Test coverage configuration
.pre-commit-config.yaml	Git hook configuration

🎯 Real Project Example

A typical Python project structure:

my_project/
├── src/
│   ├── __init__.py
│   ├── main.py
│   └── utils.py
├── tests/
│   ├── __init__.py
│   └── test_main.py
├── docs/
│   └── conf.py
├── .python-version
├── pyproject.toml
├── requirements.txt
├── setup.py
└── README.md

💡 Practical Tips

1. View .pyc File Contents

python -m dis hello.pyc

2. Compile Python Package to .pyd

# Use Cython to compile in bulk
from Cython.Build import cythonize
from setuptools import setup, Extension

extensions = [
    Extension("my_module", ["my_module.pyx"])
]

setup(ext_modules=cythonize(extensions))

3. Create Professional Distribution Package

# setup.py
from setuptools import setup, find_packages

setup(
    name="my-package",
    version="1.0.0",
    packages=find_packages(),
    install_requires=[
        "requests>=2.25.0",
    ],
    entry_points={
        'console_scripts': [
            'my-command=my_package.cli:main',
        ],
    },
)

🚀 Performance Comparison: .py vs .pyd

In certain scenarios, compiling to .pyd can significantly improve performance:

# Performance test example
import timeit

# Python version
def python_fib(n):
    if n <= 1:
        return n
    return python_fib(n-1) + python_fib(n-2)

# Cython compiled version (assuming compiled to .pyd)
from cython_fib import cython_fib

# Test performance
n = 35
python_time = timeit.timeit(lambda: python_fib(n), number=1)
cython_time = timeit.timeit(lambda: cython_fib(n), number=1)

print(f"Python: {python_time:.2f}s")
print(f"Cython: {cython_time:.2f}s")
print(f"Speedup: {python_time/cython_time:.1f}x")

📊 Summary

The Python file ecosystem is very rich, from source code to compiled files, from configuration to distribution, each file type has its specific purpose:

Development Stage: Mainly using .py and .pyx
Runtime Stage: Involves .pyc, .pyd, and .so
Distribution Stage: Uses .whl and .egg
Configuration Management: Various configuration files

Mastering the characteristics and uses of these file types can help you better organize projects, optimize performance, and manage dependencies.

Interactive Topic: What other special Python file types have you encountered in your projects? Feel free to share your experiences in the comments!