C++ and Python Integration Tools Comparison

Overview

This repository compares different tools used for integrating C++ with Python, with a focus on:

1. Inheritance and Polymorphism: The ability to expose C++ class hierarchies to Python, supporting polymorphic behavior.
2. Use of Pointers: Efficient handling of C++ pointers and memory management from Python.

The tools we are testing include:

• Pybind11
• SWIG (Simplified Wrapper and Interface Generator)
• Boost.Python
• cppyy

Each tool is tested with C++ classes that demonstrate inheritance, polymorphism, and pointer usage. The goal is to evaluate which tool provides the best integration between C++ and Python.

---

C++ Code Implementation

The core C++ code defines several classes and their functionalities:

• Shape Class: A base class with a virtual area() method.
• Circle Class: A derived class from Shape that implements area() to compute the area of a circle.
• Rectangle Class: A derived class from Shape that implements area() to compute the area of a rectangle.
• ShapeManager Class: Manages a collection of shapes and computes the total area.
• DataStreamExample Class: Demonstrates pointer usage by managing a dynamic array and exposing it to Python.

Each tool's implementation folder contains the necessary bindings and setup files for building and using the C++ code from Python.

---

Tool Implementations

1. Pybind11

The Pybind11 folder contains the implementation for integrating C++ with Python using the Pybind11 library.

To build and test the project using Pybind11, follow these steps:

Build Instructions

# Navigate to the pybind11/ folder
cd pybind11/

# Create a build directory
mkdir build
cd build/

# Run CMake to configure the build
cmake ..

# Build the project
make

# Run the Python script to test the integration
cd ..
python main.py

Expected Output

After running the above commands, you should see the following output:

Area of shape at index 0 (Circle): 78.54
Area of shape at index 1 (Rectangle): 24.00
Total area of shapes: 102.54
Data Stream Buffer (Initial): [ 0 10 20]
Data Stream Buffer (After Modification): [30 10 20]

Proof of Task Success

• The object-oriented principles of inheritance and polymorphism are evident in how the Circle and Rectangle inherit from the Shape class and override the area() method.

• Pointer usage is demonstrated through the DataStreamExample class, which uses a dynamically allocated buffer (new uint32_t[bufferSize]) and allows its manipulation from Python.

• The successful execution of the Python code, including the correct area calculations and the ability to modify the buffer, proves that the task of integrating C++ with Python has been accomplished effectively.

This demonstrates the integration of C++ functionality into Python, showing that we can handle complex C++ concepts like inheritance, polymorphism, and pointers within Python seamlessly.