|
Alexander <su wrote:<="" div="">This is very weird..
Yes, it certainly has nothing to do with C++.txtspeak is the realm of 9 year old children, not developers. Christian Graus
|
|
|
|
|
Hi, I am working on a project involving some 2D coordinates. When it comes to allocating a number of 2D coordinates, I code like this:
float (*coords)[2] = new float[num][2]; // num: Number of 2D coordinates, variable of type int
However, I do not know how to allocate an array of 2D coordinates? That is,
float (**coords)[2]; // I want to have it as [row][col][2], row & col: variable of type int
Am I declaring it correct? I know I can use float*** with the last layer's dimension as 2, but I just want to know if anyone is using this method. Thanks.modified on Sunday, March 14, 2010 1:57 AM
|
|
|
|
|
It could be implemented as following too  :
class CFPoint
{
float m_fX;
float m_fY;
public:
...
};
{
CFPoints* pcFPoints = new CFPoint[num];
}virtual void BeHappy() = 0;
|
|
|
|
|
there are many ways to do something like this.
the cleanest is to use a struct to hold your coordinate pairs:
struct floatPoint
{
float x;
float y;
};
floatPair * coords = new floatPair[num];
...
delete [] floatPair;
the safest is to use an STL container to hold the points:
#include < vector >
struct floatPoint
{
float x;
float y;
};
std::vector < floatPair > coords(num);
allocating a single buffer:
float *pBuffer = new [num * 2];
int x = 3;
int y = 2;
float fxy = pBuffer[y * 2 + x];
...
delete [] pBuffer;
etc.
|
|
|
|
|
Thanks for your reply, Eugen and Chris.
The situation is that I need to put the data inside the functions provided by a library. The prototypes are like this:
<br />
int Output1(float out1[][2]);<br />
int Output2(float out1[][2], float out2[][2]);<br />
int Output3(float out1[][2], float out2[][2], float out3[][2]);<br />
According to different cases, I need to call one of these functions in each process. I think it implies that the row and column are varying. How I want to call them is like this:
<br />
float (**coords)[2];<br />
<br />
Output1(coords[0]);
Output2(coords[0], coords[1]);
I have searched on the web and found something like this (what I planned to do if (**)[2] does not work):
<br />
float ***coords;<br />
<br />
Output1((float(*)[2])coords[0]);
Output2((float(*)[2])coords[0], (float(*)[2])coords[1]);
|
|
|
|
|
In all of the above you do not seem to be allocating any space for your arrays. The first requirement is to know how many rows are required by the functions OutputX. For example you have the prototype:
int Output1(float out1[][2]);
which means that Output1 takes the address of a two-dimensional array, with some number of rows each of two columns. Assuming it requires 20 rows you could allocate the array thus:
float out1[20][2];
int i = Output1(out1);
However, without knowing more details of the above functions it is difficult to be more specific.txtspeak is the realm of 9 year old children, not developers. Christian Graus
|
|
|
|
|
Maybe my previous description is not specific enough. Hope the code below explain more clearly.
int row0 = 0, col0 = 0;
float ***coords;
int Output1(out1[10][2]);
int Output2(out1[20][2], out2[20][2]);
int Output3(out1[30][2], out2[30][2], out3[30][2]);
int Output4(out1[40][2], out2[40][2], out3[40][2], out4[40][2]);
void Function(const int nMode)
{
int row, col;
switch (nMode)
{
case 0: row = 1; col = 10; break;
case 1: row = 2; col = 20; break;
case 2: row = 3; col = 30; break;
case 3: row = 4; col = 40; break;
}
if (row0 != row || col0 != col)
{
coords = new float**[row];
for (int i = 0; i < row; i++) {
coords[i] = new float*[col];
for (int j = 0; j < col; j++)
coords[i][j] = new float[2]; }
row0 = row; col0 = col;
}
switch (nMode)
{
case 0: Output1(coords[0]); break;
case 1: Output2(coords[0], coords[1]); break;
case 2: Output3(coords[0], coords[1], coords[2]); break;
case 3: Output4(coords[0], coords[1], coords[2], coords[3]); break;
}
}
So in this way, I do not need to declare a bunch of static arrays like:
float out1[10][2];
float out2[2][20][2];
float out3[3][30][2];
float out4[4][40][2];
And some of them may even not to be used throughout the process (if nMode sticks to a certain choice by user). Thus I think dynamic allocation will be more efficient. (Although the allocation and deallocation may cost some effort.)
Going back to original subject, my question is "Is it possible that I can use float (**)[2] instead of float***, since the dimension of the last layer is always 2?"
Thanks.
|
|
|
|
|
| I think my previous suggestion was maybe closer to what you want. You seem to be introducing complexity where it is not needed, and I think your allocation of the array contents is incorrect. I must confess that I am having difficulty following the above logic, and still do not quite understand what you are trying to achieve. txtspeak is the realm of 9 year old children, not developers. Christian Graus
|
|
|
|
|
Nevermind. I actually implemented the code above in my project and everything seems working, without memory leak, crash or data problem. I will go deeper to inspect too. Thanks for your reply.
|
|
|
|
|
|
|
Sorry, correct post should be here[^]
|
|
|
|
|
Hi,
I get this error after I imported an old VC6 explorer-extension dll project to VC9. I needed to do it because I have to compile the extension for 64-bit. I didn't get this error in VC6. Anyone knows what that means and how to fix it?There is sufficient light for those who desire to see, and there is sufficient darkness for those of a contrary disposition.
Blaise Pascal
|
|
|
|
|
Please check your DEF file :
; KOYOCE.def : Declares the module parameters.
LIBRARY "KOYO32.DLL"
EXPORTS
; DllCanUnloadNow @1 PRIVATE
; DllGetClassObject @2 PRIVATE
; DllRegisterServer @3 PRIVATE
; DllUnregisterServer @4 PRIVATE
; DllzenOnGetClassCLSID @5 PRIVATE
DllCanUnloadNow PRIVATE
DllGetClassObject PRIVATE
DllRegisterServer PRIVATE
DllUnregisterServer PRIVATE
DllzenOnGetClassCLSID PRIVATE virtual void BeHappy() = 0;
|
|
|
|
|
| Thanks, that fixed it. There is sufficient light for those who desire to see, and there is sufficient darkness for those of a contrary disposition.
Blaise Pascal
|
|
|
|
|
Hi,
I made an installer program fiew years ago, that use ShellExecute and spawn functions. It worked for years.
Today, thoses functions don't work anymore on Windows 7, returning an access denied code (-5).
I tried with ShellExecute, ShellExecuteEx, spawn, ShellExecute 'runas' ... nothing.
What is strange is that the same code works when launching NotePad.exe or Regedit.exe, but not my executables !
What should I do ? Do I have to do something whith my launched executables ? Something whith the manifest or something ?
Thanks for your Help,
Nicolas.
|
|
|
|
|
| Do your executables require administrative privileges? The access denied error suggests that there is some problem with that. There is sufficient light for those who desire to see, and there is sufficient darkness for those of a contrary disposition.
Blaise Pascal
|
|
|
|
|
Well, I never set any privileges settings to my executables : they are just standard MFC apps (orinaly made using VC6 and now with VC 2008).
How can I know if my exe requires administrative privileges ? How do I set these (from Windows or in the code) ???
Even if I give all access to the files from windows shell, it doesn't work.modified on Friday, March 12, 2010 10:51 AM
|
|
|
|
|
| see if this thread[^] helps you. There is sufficient light for those who desire to see, and there is sufficient darkness for those of a contrary disposition.
Blaise Pascal
|
|
|
|
|
Hello folks!
I was just wondering, and thought i ask you, maybe some of you have nice ideas. So here's the deal: there is a static function i have to implement, but its declaration/definition has to match a certian predefined "form", i can write the implementation as i see fit but i can't change the number and types of parameters passed. Let's call this function: TheStaticFunction(). This is actually a callback function that gets called regularry while a 3rd party system is doing its processing. Now, if i want to pass some parameters to this method, i have to use some global variable(s), set them before letting the 3rd party processing begin, eg:
int global_parameter_variable;
void TheStaticFunction()
{
... use value of global_parameter_variable as needed ...
}
...
void Do3rdPartProcessing()
{
global_parameter_variable = 123;
Perform3rdPArtyProcessing();
}This aproach works ok as long as i don't want to use multiple threads to do the processing but with different parameters per thread, since there's only one global global_parameter_variable.
So my question is: what could be done to achieve this?
One solution that comes to my mind is having some global object that stores the parameter values associated with thread handles. So when a thread is trying to use the 3rd party method, it stores its own parameter value associated with its own handle in this global object and inside the static method this global object is used to query the parameter value for the current thread. So something like this:
CThreadedParameters globa_parameter_retriever;
void TheStaticFunction()
{
...
int the_parameter = globa_parameter_retriever.GetParameterValueForThread(GetCurrentThreadHandle());
...
}
void Do3rdPartProcessing(int parameter)
{
globa_parameter_retriever.SetParameterValueForThread(GetCurrentThreadHandle(), parameter);
Perform3rdPArtyProcessing();
globa_parameter_retriever.RemoveParameterValueFor(GetCurrentThreadHandle());
}Of course access to the globa_parameter_retriever would be properly synchronized. This method would probably work but you would have to perform a lookup based on the current thread's handle every time TheStaticFunction is called, althorough this is most likely now that much of an overhaul. So does anyone have any other idea or knows any other, simpler method that exists and i just don't know about it? This all is for now only hypothetical so i can't give you any test results. Thanks for your oppinions.> The problem with computers is that they do what you tell them to do and not what you want them to do. <
> Sometimes you just have to hate coding to do it well. <
|
|
|
|
|
You can use the follow code:
__declspec( thread ) int tls_i = 1;
in visual c++.
|
|
|
|
|
| Thank you, it looks very promising. > The problem with computers is that they do what you tell them to do and not what you want them to do. <
> Sometimes you just have to hate coding to do it well. <
|
|
|
|
|
|
| That also looks good, thank you. > The problem with computers is that they do what you tell them to do and not what you want them to do. <
> Sometimes you just have to hate coding to do it well. <
|
|
|
|
|
#if !defined (__THREADLOCAL_H__)
#define __THREADLOCAL_H__
#include < list >
class TlsException
{
public:
explicit TlsException (DWORD error) : m_error (error) {}
private:
DWORD m_error;
TlsException ();
};
template < class T > class CISThreadLocal
{
public:
explicit CISThreadLocal () : m_index (::TlsAlloc ())
{
if (m_index == ~0)
{
#ifdef _DEBUG
throw TlsException (GetLastError ());
#endif
}
InitializeCriticalSection(&g_CritSection);
}
~CISThreadLocal ()
{
if (m_index != ~0)
{
::TlsFree (m_index);
}
m_index = 0;
EnterCriticalSection(&g_CritSection);
for (std::list< T *>::iterator it = m_tsdList.begin(); it!=m_tsdList.end(); it++)
{
delete (*it);
}
m_tsdList.clear();
LeaveCriticalSection(&g_CritSection);
DeleteCriticalSection(&g_CritSection);
}
inline T* operator -> ()
{
return GetThreadLocal ();
}
T* GetThreadLocal ()
{
T* tsd = 0;
try
{
if (m_index != ~0)
{
tsd = reinterpret_cast <T*> (::TlsGetValue (m_index));
if (tsd == 0)
{
tsd = new T;
BOOL success = ::TlsSetValue (m_index, tsd);
if (!success)
throw TlsException (::GetLastError ());
EnterCriticalSection(&g_CritSection);
m_tsdList.push_back(tsd);
LeaveCriticalSection(&g_CritSection);
}
}
}
catch (...)
{
tsd = NULL;
}
return tsd;
}
bool HasTLS()
{
T* tsd = 0;
try
{
if (m_index != ~0)
{
tsd = reinterpret_cast <T*> (::TlsGetValue (m_index));
}
}
catch (...)
{
tsd = NULL;
}
return (tsd != 0);
}
void DeleteThreadTLS()
{
try
{
if (m_index != ~0)
{
T* tsd = reinterpret_cast <T*> (::TlsGetValue (m_index));
if (tsd)
{
BOOL success = ::TlsSetValue (m_index, 0);
if (!success)
throw TlsException (::GetLastError ());
EnterCriticalSection(&g_CritSection);
m_tsdList.remove(tsd);
delete tsd;
LeaveCriticalSection(&g_CritSection);
}
}
}
catch (...)
{
}
}
private:
DWORD m_index;
std::list<T *> m_tsdList;
CRITICAL_SECTION g_CritSection;
CISThreadLocal (const CISThreadLocal&);
void operator = (const CISThreadLocal&);
};
the way it works is:
1. define a class that will hold all your thread-safe data:
class CMyThreadSafeData
{
...
int m_foo;
....
};
2. declare and define a single global variable of type CISThreadLocal:
extern CISThreadLocal < CMyThreadSafeData > threadSafeData;
...
CISThreadLocal < CMyThreadSafeData > threadSafeData;
3. whenever you need to access that thread-safe data, do this:
int threadFoo = threadSafeData->m_foo;
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
