IncludeCPP 2.0.4

Professional C++ Python bindings with type-generic templates and native threading

IncludeCPP - Use C++ in Python, The Easy Way

Write your performance-critical code in C++, use it in Python like it's native. No boilerplate, no manual binding code, no headaches.

Why This Exists

Python is great, but sometimes you need speed. C++ extensions are fast but painful to set up. This tool bridges that gap - write C++ code, declare what you want to expose in a simple .cp file, and use it directly in Python. That's it.

How To Get Started

Install

pip install IncludeCPP

Create Your First C++ Module. Do Following bash cmd in Your Projects Root Directory

1. Initialize your project:

python -m includecpp init

This creates:

• cpp.proj - Project config
• include/ - Put your C++ files here
• plugins/ - Module definitions go here

2. Write some C++ code (include/math_tools.cpp):

#include <vector>
#include <algorithm>

namespace includecpp {

std::vector<int> fast_sort(std::vector<int> data) {
    std::sort(data.begin(), data.end());
    return data;
}

int add(int a, int b) {
    return a + b;
}

}

3. Tell IncludeCPP what to expose (plugins/math_tools.cp):

SOURCE(include/math_tools.cpp) math_tools

PUBLIC(
    math_tools FUNC(fast_sort)
    math_tools FUNC(add)
)

4. Build it:

python -m includecpp --rebuild

5. Use it in Python:

import IncludeCPP as cpp

CPP = cpp.CppApi()
math = CPP.include("math_tools")

result = math.fast_sort([5, 2, 9, 1])
print(result)  # [1, 2, 5, 9]

print(math.add(10, 20))  # 30

That's it. No CMakeLists.txt, no setup.py, no pybind11 boilerplate.

How It Works

1. You write C++ - Normal C++ code in the includecpp namespace
2. You declare exports - Simple .cp files list what functions/classes to expose
3. IncludeCPP generates bindings - Automatic pybind11 code generation
4. Everything builds to AppData - Your project stays clean, builds are cached
5. Import in Python - Use your C++ code like any Python module

The system:

• Parses your .cp files
• Generates pybind11 bindings automatically
• Compiles everything with your system compiler
• Caches builds (rebuilds only what changed)
• Stores artifacts in %APPDATA%/IncludeCPP (Windows) or ~/.local/share/includecpp (Linux)

About The System

This is v2.0 - a complete rewrite focused on making C++ integration actually easy.

What's New in v2.0

Plain-Old-Data Structs:

// In your .cp file:
geometry STRUCT(Point) TYPES(int, float) {
    FIELD(T, x)
    FIELD(T, y)
}

Auto-generates Python integration with .to_dict() and .from_dict() methods.

Module Dependencies:

DEPENDS(math_utils, geometry)

Modules can depend on other modules. Build system handles the order automatically.

Multi-File Modules:

SOURCES(module.cpp, helpers.cpp, utils.cpp)

Split your code across multiple files, treat them as one module.

Type Support

Works with:

• Basic types: int, float, double, string, bool
• STL containers: vector, map, set, array
• Custom structs (v2.0+)
• Custom classes with methods
• Templates (define which types to instantiate)

Features

• Zero configuration - Works out of the box
• Fast incremental builds - Only rebuilds what changed
• Clean project structure - Build artifacts separate from source
• Template support - Generate functions for multiple types
• Thread-safe - GIL release for parallel execution
• Cross-platform - Windows, Linux, macOS

Advanced: Classes and Templates

Expose a C++ class:

// include/game.cpp
class Player {
public:
    int health = 100;
    void damage(int amount) { health -= amount; }
    int get_health() { return health; }
};

// plugins/game.cp
SOURCE(include/game.cpp) game

PUBLIC(
    game CLASS(Player) {
        FIELD(health)
        METHOD(damage)
        METHOD(get_health)
    }
)

# Use in Python
player = game.Player()
player.damage(25)
print(player.health)  # 75

Template functions:

template<typename T>
T maximum(T a, T b) {
    return a > b ? a : b;
}

algorithms TEMPLATE_FUNC(maximum) TYPES(int, float, string)

algorithms.maximum(10, 20)      # 20
algorithms.maximum(1.5, 2.3)    # 2.3
algorithms.maximum("a", "z")    # "z"

Requirements

• Python 3.8+
• C++ compiler (g++, clang++, or MSVC)
• CMake 3.15+
• pybind11 (auto-installed)

Configuration

Edit cpp.proj to customize:

{
  "project": "MyProject",
  "version": "1.0.0",
  "include": "/include",
  "plugins": "/plugins",
  "compiler": {
    "standard": "c++17",
    "optimization": "O3",
    "flags": ["-Wall", "-pthread"]
  }
}

The Next Generation of Coding

C++ gives you speed. Python gives you productivity. Why choose?

This tool is built for developers who:

• Need Python's ecosystem but C++'s performance
• Don't want to maintain complex build systems
• Value simplicity and fast iteration
• Want to write C++ without the usual pain

No corporate backing, no enterprise bloat. Just a tool that works.

---

License: MIT Version: 2.0.0 Status: Production Ready