Calling C/C++ Code from Python Using ctypes and Shared Libraries

When Python code runs slowly, integrating C or C++ can improve performance. The process consists of three steps: write the C/C++ implementation, compile it into a shared library, and call the library from Python.

1. Python calling a C function

The C source called_c.c defines a simple function foo . Compile it with:

<code>//编译命令 gcc -o libpycall.so -shared -fPIC called_c.c
#include <stdio.h>
int foo(int a, int b){
    printf("a:%d, b:%d.", a, b);
    return 0;
}
</code>

Then load and invoke it in Python:

<code>import ctypes
dll = ctypes.cdll.LoadLibrary
lib = dll('./libpycall.so')
lib.foo(1, 3)
</code>

Running the script prints a:1, b:3 .

2. Python calling a C++ class

Because C++ name mangling hides symbols, the functions must be exported with extern "C" . The C++ source cpp_called.cpp defines a class TestLib with overloaded display methods and wraps them in an extern block.

<code>//Python调用c++(类）动态链接库
#include <iostream>
using namespace std;
class TestLib {
public:
    void display();
    void display(int a);
};
void TestLib::display() {
    cout << "First display" << endl;
}
void TestLib::display(int a) {
    cout << "Second display:" << a << endl;
}
extern "C" {
    TestLib obj;
    void display() { obj.display(); }
    void display_int(int a) { obj.display(a); }
}
</code>

Compile the shared library:

<code>g++ -o libpycallcpp.so -shared -fPIC cpp_called.cpp</code>

Load and use it from Python:

<code>import ctypes
dll = ctypes.cdll.LoadLibrary
lib = dll('./libpycallcpp.so')
lib.display()
lib.display_int(0)
</code>

The output is:

<code>First display
Second display:0</code>

Both examples show how to bridge Python with native code, providing a foundation for more advanced integrations.