Passing unmanaged C++ object into managed C++ code - c++

So as the title states, I am looking to pass an object that is defined in an unmanaged C++ project into a managed C++/CLI project (both projects are in the same solution) while still being able to access the same methods as in the unmanaged C++ project.
This object is a class which imports functions from a separate unmanaged C++ dll (not in the same project or solution) through the means of using typedef's as shown below (excerpt of header):
#pragma unmanaged
typedef BOOL (WINAPI *someBoolFunc) (DWORD Name, DWORD Flags, TCHAR* Text);
class INeedThisClass
{
public:
INeedThisClass();
~INeedThisClass();
someBoolFunc BoolFunc;
};
extern INeedThisClass ReallyNeedThisClass;
I would really like to be able to access the externed object ReallyNeedThisClass from within my managed code by just only passing the object. By having this object I wish to access it's methods through ReallyNeedThisClass->BoolFunc. What is the best way of accessing this object from my managed code?
EDIT: I've been thinking maybe there is a way of using a property to "Get" the object from the unmanaged code and then "Set" the same object in the managed code. Could this be a proper angle to approach this issue from? I've cleaned up the question a bit to get rid of useless details.
UPDATE: I've made progress on this using a wrapper for the unmanaged C++ code, but of course there are still problems. Using the above class as an example here, I will show you what I have managed to do up until now.
Project 1: Unmanaged C++ class (example above)
Project 2: Managed C++/CLR wrapper class (example below)
Wrapper.h:
#pragma once
#include "INeedThisClass.h" //Unmanaged class INeedThisClass
using namespace System;
using namespace System::Runtime::InteropServices;
namespace INeedThisClassWrapper
{
public ref class INTC
{
public:
INTC(void);
virtual ~INTC(void);
delegate bool BoolFunc(UInt32 Name, UInt32 Flags, String^ Text);
private:
INeedThisClass *ReallyNeedThisClass;
};
}
Wrapper.cpp:
#include "Stdafx.h"
#include "Wrapper.h"
using namespace INeedThisClassWrapper
#include <msclr/marshal.h>
using namespace msclr::interop;
INTC::INTC(void)
{
ReallyNeedThisClass = new INeedThisClass();
}
INTC::~INTC(void)
{
if (ReallyNeedThisClass)
{
delete ReallyNeedThisClass;
ReallyNeedThisClass = NULL;
}
}
bool INTC::BoolFunc(UInt32 Name, UInt32 Flags, String^ Text)
{
return ReallyNeedThisClass->BoolFunc;
}
Now I am receiving errors from the compiler stating:
error C2065: 'WINAPI': undeclared identifier
error C2065: 'someBoolFunc': undeclared identifier
error C4430: missing type specifier - int assumed. Note C++ does not support default-int
Any ideas?

First, you have to export INeedThisClass from unmanaged dll and import from managed one.
//unmanaged header
#if UNMANAGED_COMPILE_DEFINITION
#define EXPORT_OR_IMPORT __declspec(dllexport)
#else
#define EXPORT_OR_IMPORT __declspec(dllimport)
#endif
class EXPORT_OR_IMPORT INeedThisClass
{
public:
INeedThisClass();
~INeedThisClass();
someBoolFunc BoolFunc;
};
Also unmanaged dll must be compiled with UNMANAGED_COMPILE_DEFINITION preprocessor definiton.
Second, remove extern INeedThisClass ReallyNeedThisClass; you don't need it. Because ReallyNeedThisClass is a member of wrapper class already.
Third, you need to marshal System::String^ to TCHAR* in order to call BoolFunc
Forth, I do not remember but if you include "Windows.h" or "Winnt.h", the error about WINAPI will be gone.

Related

Use C++/CLI Library in C++ native code and how to link

I want to use some code that executes a http-post, and because I'm not too familiar with c++ and what libraries you can use, and I am probably too dumb to get libcurl and curlpp to work, I found a link explaining how to use the .net version.
Alright so I created a class. Header File:
public ref class Element
{
public:
Element();
virtual ~Element();
void ExecuteCommand();
};
Class file:
#include "Element.h"
Element::Element()
{
}
Element::~Element()
{
Console::WriteLine("deletion");
}
void Element::ExecuteCommand(){
HttpWebRequest^ request = dynamic_cast<HttpWebRequest^>(WebRequest::Create("http://www.google.com"));
request->MaximumAutomaticRedirections = 4;
request->MaximumResponseHeadersLength = 4;
request->Credentials = gcnew NetworkCredential("username", "password", "domain");
HttpWebResponse^ response = dynamic_cast<HttpWebResponse^>(request->GetResponse());
Console::WriteLine("Content length is {0}", response->ContentLength);
Console::WriteLine("Content type is {0}", response->ContentType);
// Get the stream associated with the response.
Stream^ receiveStream = response->GetResponseStream();
// Pipes the stream to a higher level stream reader with the required encoding format.
StreamReader^ readStream = gcnew StreamReader(receiveStream, Encoding::UTF8);
Console::WriteLine("Response stream received.");
Console::WriteLine(readStream->ReadToEnd());
response->Close();
readStream->Close();
}
If I set the configuration type of this project to Application (exe), and create a new .cpp file where I create an Instance of this Element it works fine.
But my question is: Is it possible to create a .dll/.lib Library from this project and use it in a C++ project without CLI? (I don't want to use ^ for pointers :( )
Even if it's not possible, I have another problem.
When I link the library in a C++/CLI project. I get
unresolved token (06000001) Element::.ctor
unresolved token (06000002) Element::~Element
unresolved token (06000003) Element::ExecuteCommand
3 unresolved externals
the code for main.cpp in the second project is just the following:
#include <Element.h>
int main(){
return 0;
}
Thank you
As Hans Passant already stated: you must compile your C++/CLI code as Dynamic Library in order to be able to consume it from an unmanaged application. CLI/Managed code cannot run from/cannot reside in static libraries.
If you change the C++/CLI library target from Static library to Dynamic library you'll be able to compile successfully your unmanaged C++ application.
One thought from my side:
I think you'll be better if you use mixed mode C++/CLI DLLs to consume the managed functionality - you'll be able to free your consumer application completely from referencing the CLR.
The Header of such mixed mode Wrapper for your Element class would look like this:
#pragma once
#pragma unmanaged
#if defined(LIB_EXPORT)
#define DECLSPEC_CLASS __declspec(dllexport)
#else
#define DECLSPEC_CLASS __declspec(dllimport)
#endif
class ElementWrapperPrivate;
class __declspec(dllexport) ElementWrapper
{
private:
ElementWrapperPrivate* helper;
public:
ElementWrapper();
~ElementWrapper();
public:
void ExecuteCommand();
};
And the implementation would look like this:
#include "ElementWrapper.h"
#pragma managed
#include "Element.h"
#include <msclr\auto_gcroot.h>
using namespace System::Runtime::InteropServices;
class ElementWrapperPrivate
{
public:
msclr::auto_gcroot<Element^> elementInst; // For Managed-to-Unmanaged marshalling
};
ElementWrapper::ElementWrapper()
{
helper = new ElementWrapperPrivate();
helper->elementInst = gcnew Element();
}
ElementWrapper::~ElementWrapper()
{
delete helper;
}
void ElementWrapper::ExecuteCommand()
{
helper->elementInst->ExecuteCommand();
}
Then just compile your Element.cpp + ElementWrapper.cpp to a DLL and use the ElementWrapper.h in your unmanaged applications.

How to make a managed (clr) multithreaded c++ .dll?

I am trying to make a managed .dll in c++ that requires the support for multithreading. I am developing in visual Studio 2013, using platform toolset version v120. the reason I need this to be a managed assembly is because it is required to integrate the assembly in LabView.
following the steps in Creating and Using a Managed Assembly in VC++ 2010 gives good results. but I obviously need to implement something more complicated and when I include threading and write the following code:
#pragma once
#include <thread>
using namespace System;
using namespace std;
namespace MultiThread_module {
public ref class multiThreadingTest
{
public:
String^ GetVersion();
int someNumber;
private:
thread testThread;
};
}
I get following errors:
"thread" is not supported when compiling with /clr or /clr:pure.
a member of a managed class cannot be of a non-managed class type
error directive: ERROR: Concurrency Runtime is not supported when
compiling /clr.
error directive: is not supported when compiling with /clr or
/clr:pure.
A friend of mine says it is impossible to write multi-threaded code in Visual Studio without using external packages like boost. It kind of seemed unlikely since Multithreading has already been already there for C# and VB for a long time!
So, I would be happy if you could let me know what I am doing wrong OR if it is really hard to have a managed multithreaded .dll developed in c++?
You can use the managed thread library: System.Threading.Thread.
#pragma once
using namespace System;
using namespace std;
using namespace System::Threading;
namespace MultiThread_module {
public ref class multiThreadingTest
{
public:
String^ GetVersion();
int someNumber;
private:
Thread^ testThread;
};
}
If it's purely CLR then I suggest you use the example provided before. If you want to have the threading completely native and just use CLR to wrap it, I'd like to refer you to my answer at : using clr and std::thread
Might be an old question, but I looked into this same problem before. Since CLR does not allow you to include std::thead at
compile time, you could try to use it only at linking time. Normally
you could resolve this be forward declaring the class in your header
and including them only in your cpp files. However you can forward
declare your own classes in header files, but you can't for
classes in namespace std. According to the C++11 standard, 17.6.4.2.1:
The behavior of a C++ program is undefined if it adds declarations or
definitions to namespace std or to a namespace within namespace std
unless otherwise specified.
A workaround for this problem is to create a threading class that
inherits from std::thread that you can forward declare. The header
file for this class would look like:
#pragma once
#include <thread>
#include <utility>
namespace Threading
{
class Thread : std::thread
{
public:
template<class _Fn, class... _Args> Thread(_Fn fn, _Args... args) : std::thread(fn, std::forward<_Args...>(args...))
{
}
private:
};
}
In the header file that you would like to use the thread you can do
forward declare it like:
#pragma once
// Forward declare the thread class
namespace Threading { class Thread; }
class ExampleClass
{
public:
ExampleClass();
void ThreadMethod();
private:
Threading::Thread * _thread;
};
In your source file you can then use the theading class like:
#include "ExampleClass.h"
#include "Thread.h"
ExampleClass::ExampleClass() :
{
_thread = new Threading::Thread(&ExampleClass::ThreadMethod, this);
}
void ExampleClass::ThreadMethod()
{
}
Hope it might help anyone.

managed c to call an unmanaged c dll , unresolved token

I have 1 native c++ dll, CppApp, another project is managed c++, GatWayLibrary, with /clr
In GatewayLibrary, I called functions from native CppApp dll.
but I got unresolve token errors.
Here are my code snips:
CppApp.h
=========
#ifdef CPPAPP_EXPORTS
#define CPPAPP_API __declspec(dllexport)
#else
#define CPPAPP_API __declspec(dllimport)
#endif
class CPPAPP_API CppApp
{
public:
CppApp();
~CppApp();
ContextManager & contextMgr() { return m_rplContextMng; }
INativeListener* m_listener;
void registerListener(INativeListener* listener)
{
m_listener = listener;
}
...........
}
In separate project, GateWaylibrary, wrap the native dll as
#include "../CppApp/CppApp.h"
#include <vcclr.h>
using namespace System;
using namespace System::Runtime::InteropServices;
#pragma comment(lib, "CppApp.lib")
namespace GatewayLibrary{
//.net equvelant of the argument class
public ref class DotNetEventArg{
internal:
//contructor takes native version of argument to transform
DotNetEventArg(const NativeCPPArgs& args) {
....
...
}
the managed c++ has the linking errors as unresolved tokens for all the function calls from the native c++.
I did include the CppApp.lib as Additional Dependencies
and the directory.
Can anyone please help? many thanks ahead.
Edit:
here is of the place I called the native c++
`GatewayLibrary::EventGateway::EventGateway()
{
nativeCode_ = new CppApp();
//note; using 'this' in ctor is not a good practice
nativeListener_ = new NativeListenerImp(this);
//register native listener
nativeCode_->registerListener(nativeListener_);
}`

How to wrap a C++ class in a C based dll or a CLI based dll?

I am told to import my writen class in C++ into a dll and then use that dll in a c# application. Following this guide I created the dll, but I can't simply use it in a C# application since there are some issues concerning it:
What should I place for the return type of my factory function?
What is the equivalent of const wchar_t* which is my constructors argument type?
How can I retrieve and use my functions return type which is of type vector< wstring>?
These are the problems that prevent me from using my C++ DLL inside my C# applications. I was told that I need to create a wrapper with C++/CLI and then use that inside my C#. But sadly I have no idea about it, I don't know C++.net.
The only thing that currently seems to be a bit more sensational to me is to make it somehow compatible with C and then create a C DLL and use that in my C# application. I have read that in C, class object pointers are accessible through HANDLEs, so I thought that would be good idea to get things going without a lot of changes.
So the question is how can I use Handles to access my class objects in C and use them? And how can I convert a vector<wstring> to its C counterpart?
If I want to use CLI to create a wrapper (DLL?) for my C++ DLL, to be used in other dotnet apps what should I do?
In order to make a C wrapper for a C++ class to be used in for example a C# application you can do the following.
In Visual Studio choose Win32 Console Application and Enter a name, Then click next and on the next pane choose DLL and click finish. When you are done you are represented with a DLL project including 3 files.
testdll.h
testdll.cpp
dllmain
Delete everything that exists inside your testdll.h and testdll.cpp files and copy the following contents to each respectively. Add these lines to your testdll.h
// Our C wrapper for creating a dll to be used in C# apps
// The following ifdef block is the standard way of creating macros which make exporting
// from a DLL simpler. All files within this DLL are compiled with the TESTDLL_EXPORTS
// symbol defined on the command line. This symbol should not be defined on any project
// that uses this DLL. This way any other project whose source files include this file see
// TESTDLL_API functions as being imported from a DLL, whereas this DLL sees symbols
// defined with this macro as being exported.
#ifdef TESTDLL_EXPORTS
#define TESTDLL_API __declspec(dllexport)
#else
#define TESTDLL_API __declspec(dllimport)
#endif
extern "C"
{
TESTDLL_API int OurTestFunction(int x, int y);
}
It is inside this extern "C" block where you define your interface, functions to access your class member functions.Note the TESTDLL before the function prototype. All of your functions must be proceeded by that.
Add these to your testdll.cpp file:
#include "testdll.h"
#include "ourClass.h"
#define DLL_EXPORT
extern "C"
{
OurClass ourObject;
TESTDLL_API int OurTestFunction(int x, int y)
{
return ourObject.Add(x,y);
}
}
You compile this and get a C based dll which can be used in a C# application.
There are couple of things to notice though, The more important ones are:
You need to understand that the code you use as a proxy- i mean
function definition inside your testdll.h, must only use C
compatible types, it is C after all not C++.
is that you would want to be able to allocate new objects of your
class instead of just using one global object to access all methods.
For this, if you need to pass your class objects between member functions, you need to first convert it to a void* which C can understand and then pass it and use it to access your member functions of whatever.
For example I would have something like this inside my testdll.h in order to make user capable of managing the objects indirectly:
#ifdef TESTDLL_EXPORTS
#define TESTDLL_API __declspec(dllexport)
#else
#define TESTDLL_API __declspec(dllimport)
#endif
extern "C"
{
TESTDLL_API int OurTestFunction(int x, int y);
TESTDLL_API void* CreateHandle();
TESTDLL_API void* GetCurrentHandle();
TESTDLL_API void DisposeCurrentHandle();
TESTDLL_API void SetCurrentHandle(void* handle);
TESTDLL_API void* GetHandle();
TESTDLL_API void DisposeHandle(void*);
TESTDLL_API void DisposeArrayBuffers(void);
}
And inside my testdll.cpp I would define them as :
#include "testdll.h"
#include "ourClass.h"
#define DLL_EXPORT
extern "C"
{
OurClass *ourObject;
TESTDLL_API int OurTestFunction(int x, int y)
{
//return ourObject.Add(x,y); -- not any more !!
ourObject = reinterpret_cast<OurClass *>(GetHandle());
}
//Handle operations
TESTDLL_API void* CreateHandle()
{
if (ourObject == nullptr)
{
ourObject = new OurClass ;
}
else
{
delete ourObject ;
ourObject = new OurClass ;
}
return reinterpret_cast<void*>(ourObject);
}
TESTDLL_API void* GetCurrentHandle()
{
return reinterpret_cast<void*>(ourObject );
}
TESTDLL_API void DisposeCurrentHandle()
{
delete ourObject ;
ourObject = nullptr;
}
TESTDLL_API void SetCurrentHandle(void* handle)
{
if (handle != nullptr)
{
ourObject = reinterpret_cast<OurClass *>(handle);
}
else
{
ourObject = new OurClass ;
}
}
//factory utility function
TESTDLL_API void* GetHandle()
{
void* handle = GetCurrentHandle();
if (handle != nullptr)
{
return handle;
}
else
{
ourObject = new OurClass ;
handle = reinterpret_cast <void*>(ourObject );
}
return handle;
}
CDLL_API void DisposeHandle(void* handle)
{
OurClass * tmp = reinterpret_cast<OurClass *>(handle);
delete tmp;
}
TESTDLL_API void DisposeArrayBuffers(void)
{
ourObject = reinterpret_cast<OurClass *>(GetHandle());
return ourObject ->DisposeBuffers();//This is a member function defined solely for this purpose of being used inside this wrapper to delete any allocated resources by our class object.
}
}
And when we compile this Dll, we can easily work with it inside our C# application. Before being able to use our functions defined in this dll we need to use appropriate [ImportDll()]. So for our TestDll we would write:
[DllImport(#"TestDll.dll", CallingConvention = CallingConvention.Cdecl)]
public static extern int OurTestFunction(int firstNumber,int secondNumber);
And finally use it like:
private void btnReadBigram_Click(object sender, EventArgs e)
{
int x = OurTestFunction(10,50);
MessageBox.Show(x.ToString());
}
This is all I did to make my C++ class member functions accessible inside a C# application without any hassle.
Note:
When compiling your C# application make sure you have chosen the x86 Platform for compiling your project not AnyCpu.You can change your platform through properties.
Note 2:
For knowing how to create a C++/CLI wrapper for your native C++ class read this: C++/CLI wrapper for your native C++ class.
Using a native C++ class directly from C# is technically possible, but it's not trivial, and it's rarely even a good idea. For starters, you have to know the names to use to import from the DLL, which will be the names after C++ name-mangling. You also can't directly access things like vector from C#.
There are basically two good options:
The first is to write a DLL with a C interface that uses only types that can be marshalled into CLR types. You may use pointers along with the IntPtr type, but you can't really dereference those pointers. You can pretty much just store them in your C# code and then pass them back to the native DLL when needed. And you can also use simple struct types as long as you don't need deep copy to work on them. This option involves using P/Invoke.
The second option is to write a mixed-mode C++/CLI assembly that implements all the logic that needs to access your native code. This assembly can directly access classes and data from your C# code and also directly access your native code, although you should be forewarned that there are annoying breaks where you can't mix the two. For example, a ref class in C++/CLI can't have a shared_ptr member. However, it can have a raw C++ pointer as a member. A (mixed-mode) native class can also have access to a CLR handle type and make calls into the C# code through this. This option involves using C++ Interop.
It's worth noting that you could also go the other way with C++ Interop. You could have your C# code access a mixed-mode C++/CLI assembly that provides a .NET interface to some native code. However, you will still have to do some translation in this case so it's not hugely better than the first option.
A full tutorial on C++ Interop would be rather lengthy. I suggest you read up here and do some further investigation of C++ Interop on Google.
C++/CLI introduces managed objects, for which the pointer token * should be replaced with a ^, and a 'new' should be replaced with 'gcnew', you don't need to delete these objects when you're done with them, they'll be garbage collected, [edit] managed classes have a ref keyword in their definition [/edit].
Wrapping the C++ MyClass class in a C++/CLI wrapper class WrapperCLass could look something like this:
#include <stdio.h>
class MyClass
{
public:
void ShowStuff(const wchar_t *a)
{
wprintf(a);
}
};
public ref class WrapperClass
{
MyClass *wrapped;
public:
WrapperClass()
{
wrapped = new MyClass;
}
~WrapperClass()
{
delete wrapped;
}
void ShowStuff(IntPtr string)
{
wrapped->ShowStuff((const wchar_t *)string.ToPointer());
}
};
If you generate a dll with this, you'll be able to use it as a reference in your C# project
and you won't have to use the factory function mechanism.
In C++/CLI are available, so const wchar_t * is as wel.
To convert a System::String to a const wchar_t * you could use something like this:
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
namespace Client
{
class Program
{
static void Main(string[] args)
{
WrapperClass w = new WrapperClass();
IntPtr tmp;
w.ShowStuff(tmp = System.Runtime.InteropServices.Marshal.StringToHGlobalUni("Test"));
System.Runtime.InteropServices.Marshal.FreeHGlobal(tmp);
}
}
}
(There could very well be better ways to do this...)
For your return type you'll have to do the conversion in your wrapper class. Make some .net collection, iterate through your vector, convert the wstring to a System::String, and add it to the .net collection, and return that.

Class as library of global not const variables in cli c++

How should defined class where are only global variables?
I did something like that:
public ref class Klient
{
public:
Klient(){}
// zmienne
static array<DWORD,2>^ klienty = gcnew array<DWORD,2>(40,2);
static int i = 0;
static DWORD pid;
static HANDLE handle;
static String^ nick;
//funkcje
};
But if i include it more than 1 time it won't compile and showing redefinition of class error.
Did you guard your header? In Visual Studio, you should place this directive at the top of all header files:
#pragma once
This is equivalent to the classic C++ header guard:
#ifndef HEADER_SYMBOL_X
#define HEADER_SYMBOL_X
// class declarations go here
#endif // HEADER_SYMBOL_X
If you don't guard your header, C++/CLI will indeed try to redefine your class on each include.
You'll have to be a little more clear, and paste the error you get. Also if you have a "ref" class the compiler generates a default constructor for you, so you don't need to write one.
This code worked for me, I was able to fetch the static int value into my WPF application:
#pragma once
#include "windows.h"
using namespace System;
namespace cppcli
{
public ref class Klient
{
public:
static array<DWORD,2>^ klienty = gcnew array<DWORD,2>(40,2);
static int i = 22;
static DWORD pid;
static HANDLE handle;
static String^ nick;
};
}
Update:
Noticed your comment, yes you need #pragma once in there. I assumed it was there since it's generated automatically by Visual Studio, well good to know that it works :-)