First, the documentation for IcmpSendEcho2() contradicts itself:
It says:
The IcmpSendEcho2 function is called synchronously if the ApcRoutine or Event parameters are NULL
Then it says:
The IcmpSendEcho2 function is called asynchronously when either the ApcRoutine or Event parameters are specified
I presume the first one should be "if the ApcRoutine AND Event paramters are NULL"?
Also, it says under the return value:
When called asynchronously, the IcmpSendEcho2 function returns ERROR_IO_PENDING to indicate the operation is in progress
But I don't see that, I see it return 0 and GetLastError() returns ERROR_IO_PENDING. So, can both cases exist, or is the documentation completely wrong?
Now on to the next issue. I wanted to use IcmpSendEcho2() asynchronously using the ACP callback without events. This way, I didn't have to worry about resources should the number of hosts to process be extremely large. However, it doesn't work because no callback occurs. I found this in the documentation under the AcpRoutine parameter:
The routine that is called when the calling thread is in an alertable thread and an ICMPv4 reply arrives.
So I believe my problem is the main thread is not in an alterable state. Since I don't have an event to wait on, and I don't want to wait beyond the time it takes to complete everything, how do I put the main thread in an alterable state without having to guess using something like SleepEx()? Also, if I did use something like SleepEx(10,TRUE), would all the callbacks occur, or do you have to sit in a loop?
My callback context structure includes a shared global OutstandingCount type variable so I'd know when all requests were completed.
Also the ReplyBuffer is in the context structure. Another little nugget hidden in the documentation regarding the ReplyBuffer when using it asynchronously is:
The application must parse the data pointed to by ReplyBuffer parameter using the IcmpParseReplies function
So, the main question here: How are you supposed to properly use the IcmpSendEcho2() function with a AcpRoutine and no Event in a main thread?
-- Update --
Not sure if I should ask an entirely new question but now a problem where it doesn't call the ApcRoutine for every IcmpSendEcho2Ex() sent. The following code works for my normal network adapters (which are 255.255.255.0) but hangs for a 255.255.0.0 network because the outstandingcount never gets to zero.
The adapter it hangs on is:
VirtualBox Host-Only Ethernet Adapter
DHCP Enable: Yes
Autoconfiguration Enabled: Yes
Autoconfiguration IPv4Address: 169.254.21.120
Subnet Mask: 255.255.0.0
Also wonder how long it would take on networks like 10. with a subnet of 255.0.0.0.
Here's the code that starts with the IPV4Scan() built as x64 on Win10 x64:
#define PIO_APC_ROUTINE_DEFINED
#include <winternl.h>
#include <iphlpapi.h>
#include <IcmpAPI.h>
//--------------
// types
//--------------
typedef DWORD (WINAPI *LPFN_IcmpSendEcho2)(HANDLE, HANDLE , PIO_APC_ROUTINE, PVOID, IPAddr, LPVOID, WORD, PIP_OPTION_INFORMATION, LPVOID, DWORD, DWORD);
typedef DWORD (WINAPI *LPFN_IcmpSendEcho2Ex)(HANDLE, HANDLE , PIO_APC_ROUTINE, PVOID, IPAddr, IPAddr, LPVOID, WORD, PIP_OPTION_INFORMATION, LPVOID, DWORD, DWORD);
typedef HANDLE (WINAPI *LPFN_IcmpCreateFile)();
typedef BOOL (WINAPI *LPFN_IcmpCloseHandle)(HANDLE);
typedef DWORD (WINAPI *LPFN_IcmpParseReplies)(LPVOID, DWORD);
BYTE PingSignature[]={ 'X', 'Y', 'Z', '1', '2', '3', '4', '5', '6', '7', '8' };
typedef struct _sPingContext
{
ULONG *OutstandingCount; // shared number of pings outstanding
CMutex *Mutex; // mutex for ipsfound
CNumericBuffer<uint32_t> *IPsFound; // list of ips found (MSBF format)
LPFN_IcmpParseReplies fnIcmpParseReplies; // function pointer
BYTE ReplyBuffer[sizeof(ICMP_ECHO_REPLY) + sizeof(PingSignature) + sizeof(IO_STATUS_BLOCK) + 8]; // reply buffer (see API docs)
_sPingContext(ULONG *outstandingcount, CMutex *mutex, CNumericBuffer<uint32_t> *ipsfound, LPFN_IcmpParseReplies fnicmpparsereplies)
{
OutstandingCount=outstandingcount;
Mutex=mutex;
IPsFound=ipsfound;
fnIcmpParseReplies=fnicmpparsereplies;
memset(ReplyBuffer, 0, sizeof(ReplyBuffer));
};
} sPingContext, *psPingContext;
//-------------------------------------------------------------------------
// Purpose: Callback for async ping
//
// Input: ioresult - [i] io result of async operation
// pingccontext - [i] context passed on ping
// replysize - [i] reply size of ReplyBuffer
//
// Output: na
//
// Notes:
//
VOID PingCallbackCommon(DWORD ioresult, sPingContext* pingcontext, DWORD replysize)
{
// parse response buffer
if (pingcontext) {
if (ioresult==IP_SUCCESS) {
if (pingcontext->fnIcmpParseReplies(pingcontext->ReplyBuffer, replysize)) {
// point to reply buffer
PICMP_ECHO_REPLY pechoreply=reinterpret_cast<PICMP_ECHO_REPLY>(pingcontext->ReplyBuffer);
if (pechoreply->Status==IP_SUCCESS) {
// check response
if (pechoreply->DataSize==sizeof(PingSignature)) {
if (memcmp(pechoreply->Data, PingSignature, pechoreply->DataSize)==0) {
// successful ping
pingcontext->Mutex->Lock();
pingcontext->IPsFound->AddItem(pechoreply->Address);
pingcontext->Mutex->Unlock();
}
}
}
}
}
// reduce count
InterlockedDecrement(pingcontext->OutstandingCount);
// clean up
delete pingcontext;
}
}
//-------------------------------------------------------------------------
// Purpose: Callback for async ping
//
// Input: apccontext - [i] context passed on ping
//
// Output: na
//
// Notes:
//
VOID PingCallbackOld(PVOID apcontext)
{
sPingContext *pingcontext=reinterpret_cast<sPingContext*>(apcontext);
PingCallbackCommon(IP_SUCCESS, pingcontext, sizeof(pingcontext->ReplyBuffer));
}
//-------------------------------------------------------------------------
// Purpose: Callback for async ping
//
// Input: apccontext - [i] context passed on ping
// iostatusblock - [i] status of request
//
// Output: na
//
// Notes:
//
VOID PingCallback(PVOID apcontext, PIO_STATUS_BLOCK iostatusblock, ULONG reserved)
{
PingCallbackCommon(iostatusblock->Status, reinterpret_cast<sPingContext*>(apcontext), iostatusblock->Information);
}
//-------------------------------------------------------------------------
// Purpose: build list of network hosts using IPv4 Ping
//
// Input: subnet - [i] subnet being scanned (LSB format)
// hoststart - [i] host starting number for scan
// hostend - [i] host ending number for scan
// ips - [io] numeric buffer to update with found addresses
//
// Output: na
//
// Notes:
//
void IPV4Ping(IPAddr sourceip, uint32_t subnet, uint32_t hoststart, uint32_t hostend, CNumericBuffer<uint32_t> &ips)
{
// skip 127. network
if ((sourceip & 0xFF)==127)
return;
bool oldlib=false;
LPFN_IcmpSendEcho2Ex fnIcmpSendEcho2Ex=NULL;
LPFN_IcmpCreateFile fnIcmpCreateFile=NULL;
LPFN_IcmpCloseHandle fnIcmpCloseHandle=NULL;
LPFN_IcmpParseReplies fnIcmpParseReplies=NULL;
// first thing is first - check which set of functions to use
HMODULE hlib=LoadLibrary(_T("iphlpapi.dll"));
if (hlib) {
// load functions
fnIcmpCreateFile=(LPFN_IcmpCreateFile) GetProcAddress(hlib, "IcmpCreateFile");
fnIcmpSendEcho2Ex=(LPFN_IcmpSendEcho2Ex) GetProcAddress(hlib, "IcmpSendEcho2Ex");
fnIcmpCloseHandle=(LPFN_IcmpCloseHandle) GetProcAddress(hlib, "IcmpCloseHandle");
fnIcmpParseReplies=(LPFN_IcmpParseReplies) GetProcAddress(hlib, "IcmpParseReplies");
}
// check if have everything
if (!hlib || fnIcmpCreateFile==NULL || fnIcmpSendEcho2Ex==NULL || fnIcmpCloseHandle==NULL || fnIcmpParseReplies==NULL) {
// no, try old version
oldlib=true;
// clean up
if (hlib) {
FreeLibrary(hlib);
}
// load old lib
hlib=LoadLibrary(_T("icmp.dll"));
// check if loaded
if (hlib) {
// load functions
fnIcmpCreateFile=(LPFN_IcmpCreateFile) GetProcAddress(hlib, "IcmpCreateFile");
fnIcmpSendEcho2Ex=(LPFN_IcmpSendEcho2Ex) GetProcAddress(hlib, "IcmpSendEcho2Ex");
fnIcmpCloseHandle=(LPFN_IcmpCloseHandle) GetProcAddress(hlib, "IcmpCloseHandle");
fnIcmpParseReplies=(LPFN_IcmpParseReplies) GetProcAddress(hlib, "IcmpParseReplies");
}
}
// check if have everything
if (hlib) {
if (fnIcmpCreateFile!=NULL && fnIcmpSendEcho2Ex!=NULL && fnIcmpCloseHandle!=NULL && fnIcmpParseReplies!=NULL) {
// open icmp
HANDLE hicmp=fnIcmpCreateFile();
if (hicmp!=INVALID_HANDLE_VALUE) {
// variables for callback handling
ULONG outstandingcount=0;
CMutex mutex;
// process pings
for (uint32_t host=hoststart; host<=hostend; host++) {
// build full ip
IPAddr ip=subnet | host;
ip=GETMSBFDWORD(&ip);
// create context
sPingContext *pcontext;
if ((pcontext=new sPingContext(&outstandingcount, &mutex, &ips, fnIcmpParseReplies))!=NULL) {
// count request
InterlockedIncrement(&outstandingcount);
// now issue ping
DWORD result=fnIcmpSendEcho2Ex(hicmp,
NULL,
oldlib ? (PIO_APC_ROUTINE) PingCallbackOld : PingCallback,
pcontext,
sourceip,
ip,
PingSignature,
sizeof(PingSignature),
NULL,
pcontext->ReplyBuffer,
sizeof(pcontext->ReplyBuffer),
50);
// check if failed
if (result==0) {
// check if because pending
if (GetLastError()!=ERROR_IO_PENDING) {
// no - use callback to clean up
CDebugPrint::DebugPrint(_T("IcmpSendEcho Error %u\n"), GetLastError());
PingCallbackOld(pcontext);
}
else {
// fire off pending APC callbacks ready
SleepEx(0, TRUE);
}
}
else {
// completed sync - use callback to clean up
PingCallbackOld(pcontext);
}
}
}
// wait for completion
while (outstandingcount) {
// handle callbacks
SleepEx(10, TRUE);
}
// clean up
fnIcmpCloseHandle(hicmp);
}
}
// clean up
FreeLibrary(hlib);
}
}
//-------------------------------------------------------------------------
// Purpose: build list of network hosts by way of IP scan for V4
//
// Input: ipadapteraddress - [i] adapter ip address to build for
//
// Output: na
//
// Notes: ip addresses are MSBF
//
void IPV4Scan(IP_ADAPTER_UNICAST_ADDRESS *ipadapteraddress)
{
// build the subnet mask to use
if (ipadapteraddress->OnLinkPrefixLength<=32 && ipadapteraddress->OnLinkPrefixLength!=0) {
in_addr ia=reinterpret_cast<sockaddr_in*>(ipadapteraddress->Address.lpSockaddr)->sin_addr;
// valid mask length - build mask
uint32_t rangemask=((1U<<(32-ipadapteraddress->OnLinkPrefixLength))-1);
uint32_t mask=~rangemask;
uint32_t subnet=GETMSBFDWORD(&ia.s_addr) & mask;
CDebugPrint::DebugPrint(_T("Subnet %u.%u.%u.%u/%u\n"), (subnet>>24) & 0xFF, (subnet>>16) & 0xFF, (subnet>>8) & 0xFF, (subnet>>0) & 0xFF, ipadapteraddress->OnLinkPrefixLength);
CDebugPrint::DebugPrint(_T("Scanning %u hosts\n"), (UINT32_MAX & rangemask)-1);
CNumericBuffer<uint32_t> ipsfound;
IPV4Ping(ia.s_addr, subnet, 1, (UINT32_MAX & rangemask)-1, ipsfound);
for (UINT i=0; i<(UINT)ipsfound.GetCount(); i++) {
uint32_t ip=ipsfound[i];
CDebugPrint::DebugPrint(_T("Ping found %u.%u.%u.%u\n"), ip & 0xFF, (ip>>8) & 0xFF, (ip>>16) & 0xFF, (ip>>24) & 0xFF);
}
}
else CDebugPrint::DebugPrint(_T("Invalid subnet length %u\n"), ipadapteraddress->OnLinkPrefixLength);
}
I presume the first one should be "if the ApcRoutine AND Event
paramters are NULL"?
yes, you correct.
But I don't see that, I see it return 0 and GetLastError() returns
ERROR_IO_PENDING. So, can both cases exist, or is the documentation
completely wrong?
documentation completely wrong. by fact IcmpSendEcho2[Ex] return BOOL and error code via SetLastError ( more exactly by RtlNtStatusToDosError)
so on asynchronous call it return FALSE (0) and GetLastError() will be ERROR_IO_PENDING if all ok - this mean apc callback will be called, or another error if fail - apc callback will be not called (better call it by self in this case, for common error handling)
how do I put the main thread in an alterable state
this already depend from what your thread doing. in some case possible write event loop with MsgWaitForMultipleObjectsEx function - at once wait on windows events and be alertable. also possible wait on some objects too. if you can not rewrite self message loop with MsgWaitForMultipleObjectsEx - you can do call from worked thread, or periodically call SleepEx(0, TRUE) or undocumented NtTestAlert. without know what your main thread doing - hard say exactly what is better.
demo code can look like:
#include <iphlpapi.h>
#include <IPExport.h>
#include <icmpapi.h>
class EchoRequestContext
{
HANDLE _hFile = 0;
PVOID _ReplyBuffer = 0;
LONG _dwRefCount = 1;
ULONG _dwThreadId = GetCurrentThreadId();
static void WINAPI sOnApc(PVOID This, PIO_STATUS_BLOCK piosb, ULONG )
{
reinterpret_cast<EchoRequestContext*>(This)->OnApc(
RtlNtStatusToDosError(piosb->Status),
(ULONG)piosb->Information);
}
void OnApc(ULONG dwError, ULONG ReplySize)
{
OnReply(dwError, (PICMP_ECHO_REPLY)_ReplyBuffer, ReplySize);
if (_ReplyBuffer) delete [] _ReplyBuffer;
Release();
}
void OnReply(ULONG dwError, PICMP_ECHO_REPLY ReplyBuffer, ULONG ReplySize)
{
if (dwError)
{
DbgPrint("dwError=%u\n", dwError);
return ;
}
if (IcmpParseReplies(ReplyBuffer, ReplySize))
{
__nop();
}
}
~EchoRequestContext()
{
if (_hFile) IcmpCloseHandle(_hFile);
PostThreadMessageW(_dwThreadId, WM_QUIT, 0, 0);
}
public:
void AddRef()
{
InterlockedIncrementNoFence(&_dwRefCount);
}
void Release()
{
if (!InterlockedDecrement(&_dwRefCount))
{
delete this;
}
}
ULONG Create()
{
HANDLE hFile = IcmpCreateFile();
if (hFile == INVALID_HANDLE_VALUE)
{
return GetLastError();
}
_hFile = hFile;
return NOERROR;
}
void SendEcho(
IPAddr DestinationAddress,
const void* RequestData,
WORD RequestSize,
ULONG ReplySize,
ULONG Timeout,
UCHAR Flags,
UCHAR Ttl)
{
if (PVOID ReplyBuffer = new UCHAR[ReplySize])
{
_ReplyBuffer = ReplyBuffer;
IP_OPTION_INFORMATION opt = { Ttl, 0, Flags };
AddRef();
ULONG dwError = IcmpSendEcho2Ex(_hFile, 0, sOnApc, this,
0, DestinationAddress,
const_cast<void*>(RequestData), RequestSize,
&opt, ReplyBuffer, ReplySize, Timeout) ? NOERROR : GetLastError();
switch (dwError)
{
case NOERROR:
case ERROR_IO_PENDING:
break;
default:
OnApc(dwError, 0 );
}
return ;
}
OnApc(ERROR_OUTOFMEMORY, 0);
}
};
#define IP(a, b, c, d) ((ULONG)(a + (b << 8) + (c << 16) + (d << 24)))
void EchoTest()
{
WSADATA wd;
if (NOERROR == WSAStartup(WINSOCK_VERSION, &wd))
{
if (EchoRequestContext* p = new EchoRequestContext)
{
if (p->Create() == NOERROR)
{
p->SendEcho(IP(8,8,8,8), "1234567890ABCDEF", 16, 0x100, 4000, IP_FLAG_DF, 255);
}
p->Release();
}
MSG msg;
__loop:
switch (MsgWaitForMultipleObjectsEx(0, 0, INFINITE,
QS_ALLINPUT, MWMO_ALERTABLE|MWMO_WAITALL))
{
default:
__debugbreak();
break;
case WAIT_FAILED:
break;
case WAIT_OBJECT_0:
while (PeekMessageW(&msg, 0, 0, 0, PM_REMOVE))
{
if (msg.message == WM_QUIT)
{
goto __exit;
}
}
case STATUS_USER_APC: // == WAIT_IO_COMPLETION
goto __loop;
}
__exit:
WSACleanup();
}
}
I am currently writing a simple server using Boost.Asio 1.68 and I am wondering if there is a way for me to add a handler that is executed when the event loop has no other work to do.
Currently I have this:
void completionHandler (boost::asio::io_context* ioCtx){
// poll for some condition
// if (condition) do some work;
ioCtx->post(boost::bind(completionHandler, ioCtx));
}
//elsewhere
ioCtx->post(boost::bind(completionHandler, ioCtx));
However, this doesn't exactly match what I want to do.
This wouldn't do what you expect it to do.
For example, with a single async_accept loop, you would never reach the point of "no other work to do".
Likewise, if only a single party owns outstanding work<> (docs and why) there will never be a situation where there is "no other work to do".
Basically, what you really want to do is to chain the polling:
template <typename Condition, typename Handler, typename Executor>
void run_when(Executor& ex, Condition c, Handler h) {
struct Op {
Executor& _executor;
Condition _cond;
Handler _handler;
Op(Executor& ex, Condition c, Handler h)
: _executor(ex), _cond(std::move(c)), _handler(std::move(h))
{ }
void operator()() const {
if (_cond())
_handler(error_code{});
else
ba::post(_executor, std::move(*this));
}
};
ba::post(ex, Op{ex, std::move(c), std::move(h)});
}
This can be used like:
run_when(io,
[&] { return bool(triggered); },
[](error_code ec) { std::clog << "triggered: " << ec.message() << "\n"; });
Demo
Live On Coliru
#include <boost/asio.hpp>
namespace ba = boost::asio;
using boost::system::error_code;
using namespace std::chrono_literals;
template <typename Condition, typename Handler, typename Executor>
void run_when(Executor& ex, Condition c, Handler h) {
struct RunWhen {
Executor& _executor;
Condition _cond;
Handler _handler;
RunWhen(Executor& ex, Condition c, Handler h)
: _executor(ex), _cond(std::move(c)), _handler(std::move(h))
{ }
void operator()() const {
if (_cond())
_handler(error_code{});
else
ba::post(_executor, std::move(*this));
}
};
ba::post(ex, RunWhen{ex, std::move(c), std::move(h)});
}
#include <iostream>
int main() {
// some state that gets changed in the background
std::atomic_bool triggered { false };
std::thread([&] {
std::this_thread::sleep_for(1.5s);
triggered = true;
}).detach();
ba::io_context io;
// just some background polling that shall not block other work
run_when(io, [&] { return bool(triggered); }, [](error_code ec) { std::clog << "triggered: " << ec.message() << "\n"; });
io.run_for(3s);
}
Prints (after ~1.5s):
triggered: Success
BONUS
Why does our handler take an error_code? Well, in line with other Asio operations, you might want to be able to cancel them. Either you make the caller responsible for extending the lifetime of the the run_when<>(...)::Op instance, complicating life. Or you make it so the Condition calleable can return a code indicating whether the condition was satisfied or the wait was abandoned¹:
Live On Coliru
#include <boost/asio.hpp>
namespace ba = boost::asio;
using boost::system::error_code;
using boost::system::system_error;
using ba::error::operation_aborted;
using namespace std::chrono_literals;
template <typename Condition, typename Handler, typename Executor>
void run_when(Executor& ex, Condition c, Handler h) {
struct Op {
Executor& _executor;
Condition _cond;
Handler _handler;
Op(Executor& ex, Condition c, Handler h)
: _executor(ex), _cond(std::move(c)), _handler(std::move(h))
{ }
void operator()() const {
try {
if (_cond())
_handler(error_code{});
else
ba::post(_executor, std::move(*this));
} catch(system_error const& se) {
_handler(se.code());
}
}
};
ba::post(ex, Op{ex, std::move(c), std::move(h)});
}
#include <random>
auto random_delay() {
static std::mt19937 engine(std::random_device{}());
return (std::uniform_int_distribution<>(1,2)(engine)) * 1s;
}
#include <iostream>
int main() {
// some state that gets changed in the background
std::atomic_bool triggered { false }, canceled { false };
std::thread([&] { std::this_thread::sleep_for(1.5s); triggered = true; }).detach();
// add a randomized cancellation
{
auto cancel_time = random_delay();
std::clog << "hammer time: " << (cancel_time/1.0s) << "s\n";
std::thread([&] { std::this_thread::sleep_for(cancel_time); canceled = true; }).detach();
}
ba::io_context io;
// just some background polling that shall not block other work
auto condition = [&] { return canceled? throw system_error(operation_aborted) : bool(triggered); };
run_when(io, condition, [](error_code ec) { std::clog << "handler: " << ec.message() << "\n"; });
io.run_for(3s);
}
Which prints either
hammer time: 1s
handler: Success
or
hammer time: 2s
handler: Success
depending on the value of random_delay().
¹ (or that the mayor's daughter had a divorce, because error_code is pretty versatile like that)
I want to know when dispatchhas finished with some specific work
service.dispatch(&some_work);
I want to know this because I need to restart some_work if it has finished.
struct work
{
std::shared_ptr<asio::io_service> io_service;
bool ready;
std::mutex m;
template <class F>
void do_some_work(F&& f)
{
if (io_service && ready) {
m.lock();
ready = false;
m.unlock();
io_service->dispatch([&f, this]() {
f();
m.lock();
ready = true;
m.unlock();
});
}
}
work(std::shared_ptr<asio::io_service> io_service)
: io_service(io_service)
, ready(true)
{
}
};
int
main()
{
auto service = std::make_shared<asio::io_service>();
auto w = std::make_shared<asio::io_service::work>(*service);
std::thread t1([&] { service->run(); });
work some_work{ service };
for (;;) {
some_work.do_some_work([] {
std::cout << "Start long draw on thread: " << std::this_thread::get_id()
<< std::endl;
std::this_thread::sleep_for(std::chrono::seconds(5));
std::cout << "End long draw on thread: " << std::this_thread::get_id()
<< std::endl;
});
}
w.reset();
t1.join();
}
There are some problems with the code, for example if some_workgoes out of scope, then the running taskwould still write to ready.
I am wondering if something like this already exists in Asio?
For lifetime issues, the common idiom is indeed to use shared pointers, examples:
Ensure no new wait is accepted by boost::deadline_timer unless previous wait is expired
Boost::Asio Async write failed
Other than that, the completion handler is already that event. So you would do:
void my_async_loop() {
auto This = shared_from_this();
socket_.async_read(buffer(m_buffer, ...,
[=,This](error_code ec, size_t transferred) {
if (!ec) {
// do something
my_async_loop();
}
}
);
}
This will re-schedule an (other?) async operation once the previous has completed.
On the subject of threadsafety, see Why do I need strand per connection when using boost::asio?
It might seem obvious to some, but I was still wondering: is there no way to make the compiler deduce the lambda return type without actually calling it?
Of course you can auto retval = myLambda();, however I want sth like:
T_MY_LAMBDA_RETVAL retval;
...
retval = myLambda();
Why would I want that? Simple:
try {
auto retval = myLambda();
catch (exception e) {
...
}
// retval is undefined but I wonna have it!!! :)
typedef typename std::result_of<decltype(lambda)()>::type return_type;
deduces the return type of invoking lambda with 0 arguments.
As would:
typedef decltype( lambda() ) return_type;
Be careful about possible rvalue references.
To complete Yakk's answer
with:
auto myLambda = [&](int) { return 42; }; // Or any lambda
you may use
using T_MY_LAMBDA_RETVAL = std::result_of<decltype(myLambda)(int)>::type;
or
using T_MY_LAMBDA_RETVAL = decltype(myLambda(std::declval<int>()));
and then use it:
T_MY_LAMBDA_RETVAL retval;
try {
retval = myLambda(42);
catch (exception& e) {
// ...
}
// you may use retval which may be uninitialized.
A lambda expression creates a unique functor type that overloads operator(), this is what gets invoked when you call the lambda. We can make use of this fact to figure out the return type of the lambda expression.
template<typename T>
struct lambda_traits_helper; // undefined
// specialize for pointer to member function
template<typename T, typename Result, typename... Args>
struct lambda_traits_helper<Result(T::*)(Args...) const>
{
using result_type = Result;
};
template<typename Func>
struct lambda_traits
: lambda_traits_helper<decltype(&Func::operator())>
// instantiate helper trait with pointer to operator()
{};
Example usage:
auto lambda = []{ return 42; };
assert((std::is_same<int, lambda_traits<decltype(lambda)>::result_type>::value));
auto lambda = [](int a){ return std::to_string(a); };
assert((std::is_same<std::string, lambda_traits<decltype(lambda)>::result_type>::value));
Live demo