如何将C语言编写的Win32 DLL导入Python？是否可行？

Absolutely you can pull this off—Python has built-in tools that make it straightforward, no prior C experience required. Let’s break this down step by step:

The Go-To Tool: ctypes #

Python’s ctypes module is your best friend here. It’s part of the standard library, so you don’t need to install anything extra. It lets Python call functions from shared libraries (like Win32 DLLs) directly.

Step 1: Figure Out What Functions the DLL Exports #

First, you need to know the names of the functions you want to call, plus their parameter types and return types. If you don’t have documentation for the DLL, you can use Microsoft’s dumpbin tool (included with Visual Studio Build Tools) to list exported functions:

dumpbin /exports your_dll_file.dll

This will spit out a list of all functions the DLL makes available. Note down the exact function names (sometimes they have weird suffixes like @8 from C name mangling—you’ll need those exact names later).

Step 2: Load the DLL in Python #

Open up a Python script and use ctypes.WinDLL to load your DLL. Use WinDLL instead of CDLL because Win32 DLLs typically use the stdcall calling convention (which WinDLL handles correctly):

import ctypes

# Load the DLL—use the full path if it’s not in the same folder as your script
my_dll = ctypes.WinDLL("C:/path/to/your_dll.dll")
# Or if it’s in the same folder:
# my_dll = ctypes.WinDLL("./your_dll.dll")

Step 3: Tell Python About the Function’s Signature #

By default, ctypes assumes functions return an int and takes generic arguments, which can cause crashes if the types don’t match. You need to explicitly define the argument types (argtypes) and return type (restype) for each function you want to call.

For example, if your DLL has a function like this (in C):

int AddTwoNumbers(int num1, int num2);

In Python, you’d set it up like this:

# Define the argument types (two integers)
my_dll.AddTwoNumbers.argtypes = (ctypes.c_int, ctypes.c_int)
# Define the return type (integer)
my_dll.AddTwoNumbers.restype = ctypes.c_int

If the function takes strings, use ctypes.c_char_p (for bytes) or ctypes.c_wchar_p (for Unicode strings), depending on the DLL’s expected input.

Step 4: Call the Function! #

Once you’ve set up the signature, calling the function is just like calling a regular Python function:

result = my_dll.AddTwoNumbers(10, 20)
print(f"Result: {result}")  # Should print 30

Common Pitfalls to Watch For #

• DLL Not Found: Make sure the path to the DLL is correct. If you put the DLL in the same folder as your Python script, using ./your_dll.dll should work. If not, use the full absolute path.
• Mangled Function Names: If dumpbin shows a name like _AddTwoNumbers@8, you need to use that exact name in Python: my_dll._AddTwoNumbers@8.
• Type Mismatches: If you pass the wrong type (e.g., a Python string instead of a c_char_p), your script might crash. Double-check the function’s C signature against your argtypes and restype setup.
• Complex Types: If the DLL uses structs, pointers, or other complex types, you can define them using ctypes (e.g., ctypes.Structure for structs). Start with simple functions first to get the hang of it.

内容的提问来源于stack exchange，提问作者slantalpha