[Python] How to pass a byte buffer from python to C/C++

I am creating a library in C++ with Python bindings.

The library takes buffer created in Python code and processes it in several ways. On my way to achieve that, I used the example below generated by Google's Gemini as starter for that part of the code. But I am getting an error:

Here is the C++ code:

#include <pybind11/pybind11.h>

struct MyStruct {
int value;
double* data;
int size;

// Constructor
MyStruct(int val, int sz) : value(val), size(sz) {
data = new double[sz];
for (int i = 0; i < sz; ++i) {
data = 0.0; // Initialize data to 0
}
}

// Destructor to free allocated memory
~MyStruct() {
delete[] data;
}
};

PYBIND11_MODULE(example, m) {
pybind11::class_<MyStruct>(m, "MyStruct")
.def(pybind11::init<int, int>())
.def_readwrite("value", &MyStruct::value)
.def_readwrite("data", &MyStruct::data)
.def_readwrite("size", &MyStruct::size);
}


Here is the Python code:

import example

my_struct = example.MyStruct(10, 5)

print(f"Value: {my_struct.value}")
print(f"Size: {my_struct.size}")

my_struct.value = 20
print(f"Modified Value: {my_struct.value}")

# Access and modify data through the pointer (be careful!)
for i in range(my_struct.size):
my_struct.data = i * 2.0

print("Data:", [my_struct.data for i in range(my_struct.size)])


And this is the error I am getting:

Value: 10
Size: 5
Modified Value: 20
Traceback (most recent call last):
File "./test.py", line 13, in <module>
my_struct.data = i * 2.0
~~~~~~~~~~~~~~^^^
TypeError: 'float' object does not support item assignment


So, how to fix this error. And my final goal is really to pass a byte buffer from python to C++ code, if there is any better way to do it, please let me know.

Continue reading...