Thread: A strange question, I know.

I don't think C++ supports converting named values to their string representation (name as in enum, define, or what ever). However, why is the convert to string step necessary? Why can't you convert or map WM_CREATE to (for example) WM_DESTROY directly?

Edit:
Unless you are willing to do something like

Code:

switch (message) {
    case WM_CREATE: s = "WM_CREATE"; break;
    //etc.
}

But that would still be strange.

These messages are defined as just that - defines! So it's impossible to convert them back to strings. You just have to create a string and do a switch-case and fill the string with the appropriate string.

I don't know exactly what the question is, but maybe you could use this to convert an indentifier to a string?

Code:

#include <stdio.h>
#include <string.h>
#define STRINGIFY(X) (( X ), #X )

int main()
{
	char arr[10];
	int ident;

	strcpy(arr, STRINGIFY(ident));
	puts(arr);

	return 0;
}

Code:

#include <iostream>
#include <string>
#define STRINGIFY(X) (( X ), #X )

int main()
{
	std::string arr;
	int ident;

	arr = STRINGIFY(ident);
	std::cout << arr << std::endl;

	return 0;
}

That macro doesn't compile, and anyway the OP is not interested in the name of the variable (message), but the name of the value it contains.

It might be best to post some more context of the question. Why do you need something like that?

Are we playing C again? We could easily use std::string if that sample would work. I'm unsure how it works.
arr = STRINGIFY(ident);

I think there is going to be no easy way to complete this project, at least in the method that I wanted to.

The purpose of the conversion from UINT to string, then back to UINT was simply for this.

Code:

Func* ptrArray[207]; // set to various values as required by the developer
.
.
.
WndProc() {
           char* string=ConvertDef2String (say WM_CREATE to "WM_CREATE");
           string[1]="W";
           msg=MAKEINTRESOURCE(string); // I believe this is what would do it
           if( *ptrArray[msg] != NULL ) { *ptrArray[msg](params); }
           else { DefWindowProc(params); }
}

But now that I have realized that any number of definitions could technically hold the same values as each other, there is no way that there could be a single method of simply identifying which string could possibly match that particular value. So I have resolved to simply creating int getArrayPos(msg), or something of the like that will reference array positions to the possible msg values and return the appropriate array position. Reference List of some kind basically. Unless someone else can think of something? I can't get the STRINGIFY example to compile on my system either.


PS - Elysia, yes I most likely am ... a hard habit to break. I'm still having issues abandoning #include <stdio.h> and all the functions that come with it.

Code:

WndProc() {

Doesn't work in C++. All functions must explicitly return something.
Anyway, what you might want is a map.

Actually, this sounds a lot like you want to do function pointers for specific messages, does it not?
Perhaps this thread might help: http://cboard.cprogramming.com/showthread.php?t=98360

You're still writing C code in the C++ forum.

...Except it's not good to post C code in the C++ forum and vice versa, yes?

Elysia,
I realize that WndProc() doesn't work in C++, it doesn't work in C either. Its a psuedo/code mix just to display what I was attempting to do, in actuality it should look like LRESULT CALLBACK WndProc(HWND hwnd, UINT msg, WPARAM wParam, LPARAM lParam), but I just placed WndProc() for easier typing. A case of laziness, nothing more.


Sorry, just realized the last portion of what you typed. Yes I am trying to create an array of function pointers that will allow me to call each repectively. I will have to look in to that, it might just be what I'm looking for. Thanks again!

I would be all for using a std::map to convert form one Windows message type ID to another, but not to and from strings.
That to me seems like an exercize in Softcoding, for which you'd need to read this:
http://thedailywtf.com/Articles/Soft_Coding.aspx
http://en.wikipedia.org/wiki/Softcoding