What is the easiest way to use a DLL
file from within Python
?
Specifically, how can this be done without writing any additional wrapper C++
code to expose the functionality to Python
?
Native Python
functionality is strongly preferred over using a third-party library.
I present a fully worked example on how building a shared library
and using it under Python
by means of ctypes
. I consider the Windows
case and deal with DLLs
. Two steps are needed:
The shared library
I consider is the following and is contained in the testDLL.cpp
file. The only function testDLL
just receives an int
and prints it.
#include <stdio.h>
?
extern "C" {
?
__declspec(dllexport)
?
void testDLL(const int i) {
printf("%d\n", i);
}
?
} // extern "C"
To build a DLL
with Visual Studio
from the command line run
"C:\Program Files (x86)\Microsoft Visual Studio 12.0\Common7\Tools\vsdevcmd"
to set the include path and then run
cl.exe /D_USRDLL /D_WINDLL testDLL.cpp /MT /link /DLL /OUT:testDLL.dll
to build the DLL.
DLL
from the IDEAlternatively, the DLL
can be build using Visual Studio
as follows:
Under Python, do the following
import os
import sys
from ctypes import *
lib = cdll.LoadLibrary('testDLL.dll')
lib.testDLL(3)
ctypes can be used to access dlls, here's a tutorial:
This page has a very simple example of calling functions from a DLL file.
Paraphrasing the details here for completeness:
It's very easy to call a DLL function in Python. I have a self-made DLL file with two functions:
add
andsub
which take two arguments.
add(a, b)
returns addition of two numbers
sub(a, b)
returns substraction of two numbersThe name of the DLL file will be "demo.dll"
Program:
from ctypes import*
# give location of dll
mydll = cdll.LoadLibrary("C:\\demo.dll")
result1= mydll.add(10,1)
result2= mydll.sub(10,1)
print "Addition value:"+result1
print "Substraction:"+result2
Output:
Addition value:11
Substraction:9
ctypes will be the easiest thing to use but (mis)using it makes Python subject to crashing. If you are trying to do something quickly, and you are careful, it's great.
I would encourage you to check out Boost Python. Yes, it requires that you write some C++ code and have a C++ compiler, but you don't actually need to learn C++ to use it, and you can get a free (as in beer) C++ compiler from Microsoft.
Maybe with Dispatch
:
from win32com.client import Dispatch
zk = Dispatch("zkemkeeper.ZKEM")
Where zkemkeeper is a registered DLL file on the system... After that, you can access functions just by calling them:
zk.Connect_Net(IP_address, port)
I present a fully worked example on how building a shared library
and using it under Python
by means of ctypes
. I consider the Windows
case and deal with DLLs
. Two steps are needed:
The shared library
I consider is the following and is contained in the testDLL.cpp
file. The only function testDLL
just receives an int
and prints it.
#include <stdio.h>
?
extern "C" {
?
__declspec(dllexport)
?
void testDLL(const int i) {
printf("%d\n", i);
}
?
} // extern "C"
To build a DLL
with Visual Studio
from the command line run
"C:\Program Files (x86)\Microsoft Visual Studio 12.0\Common7\Tools\vsdevcmd"
to set the include path and then run
cl.exe /D_USRDLL /D_WINDLL testDLL.cpp /MT /link /DLL /OUT:testDLL.dll
to build the DLL.
DLL
from the IDEAlternatively, the DLL
can be build using Visual Studio
as follows:
Under Python, do the following
import os
import sys
from ctypes import *
lib = cdll.LoadLibrary('testDLL.dll')
lib.testDLL(3)
If the DLL is of type COM library, then you can use pythonnet.
pip install pythonnet
Then in your python code, try the following
import clr
clr.AddReference('path_to_your_dll')
then instantiate an object as per the class in the DLL, and access the methods within it.
ctypes can be used to access dlls, here's a tutorial:
ctypes will be the easiest thing to use but (mis)using it makes Python subject to crashing. If you are trying to do something quickly, and you are careful, it's great.
I would encourage you to check out Boost Python. Yes, it requires that you write some C++ code and have a C++ compiler, but you don't actually need to learn C++ to use it, and you can get a free (as in beer) C++ compiler from Microsoft.
This page has a very simple example of calling functions from a DLL file.
Paraphrasing the details here for completeness:
It's very easy to call a DLL function in Python. I have a self-made DLL file with two functions:
add
andsub
which take two arguments.
add(a, b)
returns addition of two numbers
sub(a, b)
returns substraction of two numbersThe name of the DLL file will be "demo.dll"
Program:
from ctypes import*
# give location of dll
mydll = cdll.LoadLibrary("C:\\demo.dll")
result1= mydll.add(10,1)
result2= mydll.sub(10,1)
print "Addition value:"+result1
print "Substraction:"+result2
Output:
Addition value:11
Substraction:9
Maybe with Dispatch
:
from win32com.client import Dispatch
zk = Dispatch("zkemkeeper.ZKEM")
Where zkemkeeper is a registered DLL file on the system... After that, you can access functions just by calling them:
zk.Connect_Net(IP_address, port)
ctypes will be the easiest thing to use but (mis)using it makes Python subject to crashing. If you are trying to do something quickly, and you are careful, it's great.
I would encourage you to check out Boost Python. Yes, it requires that you write some C++ code and have a C++ compiler, but you don't actually need to learn C++ to use it, and you can get a free (as in beer) C++ compiler from Microsoft.
ctypes can be used to access dlls, here's a tutorial:
ctypes will be the easiest thing to use but (mis)using it makes Python subject to crashing. If you are trying to do something quickly, and you are careful, it's great.
I would encourage you to check out Boost Python. Yes, it requires that you write some C++ code and have a C++ compiler, but you don't actually need to learn C++ to use it, and you can get a free (as in beer) C++ compiler from Microsoft.
Source: Stackoverflow.com