Thread: OPENFILENAME ofn;

hi there, i'm new to the forum, i've been visiting cp for a while and i decided to join the forum =).. plus i have a question lol..

Ok the thing is, i have 3000+ lines of code, and the only thing that doesn't work is the ofn declaration..

...
OPENFILENAME ofn;
...
then when i try to access a member of ofn, it's like it hadn't been declared ever.. at first i thought maybe it has to do something with the program flow so i've put it in different parts of the code, still doesn't works...

i was able to compile another prog that used only this OFN struct and some menus and it did work correctly and it's placed almost exactly in the same place... WTF am i doing wrong? sorry for not giving much detail but i've checked every possible thing, i'm so confused... my prog is ready after a week of writing and this &%$& error is the only thing left to figure out before compiling the final version =(...

yeah.. i could code another browser-type dlgbox but am i guilty of being lazy if there's already one.. it just doesn't work =( =( =(

I first tried compiling with DEV-C++ 5.0 Beta 8 (4.9.8.3) so i thought it was a bug in the compiler, i reinstalled DEV-C++ 4.0 and the same thing happened so.... =( please help me...

Originally posted by knutso
Could you post some code?

you might find this piece of the ocde a little akward but lol after recoding and recoding to see if the problem fixed this is what's left.. but it's acceptable hehe.. as long as ti works ='(....

Code:

...........
case WM_COMMAND:
 switch(wParam)
 {
 case (IDM_ENCRYPT || IDM_DECRYPT):
/* start */
                //OFN struct, used for the browser-type dlgbox
                //======================================
                OPENFILENAME ofn;
                ZeroMemory(&ofn, sizeof(ofn));
                ofn.lStructSize = sizeof(ofn);
                ofn.hwndOwner = hwnd;
                //ofn.lpstrFile
                //File name is stored in "lpstrFile" member......
                ofn.nMaxFile = MAX_PATH;
                ofn.Flags = OFN_EXPLORER | OFN_FILEMUSTEXIST;
                //======================================
                ////////////////////////////////////************ =)
                if(WM_COMMAND == IDM_ENCRYPT)
                {
//=============================================
    ofn.lpstrFilter = "All Files (*.*)\0*.*\0";
    ofn.lpstrDefExt = "exe";
//============================================
                  if(GetOpenFileName(&ofn))
                  {
                  FileW.StartCrypt(ofn.lpstrFile, ACTION_ENCRYPT);
                  }
                 else
                  {
                  MessageBox(hwnd, "Unexpected file error.", "ERROR!", MB_OK);
                  }
                }
                else if(WM_COMMAND == IDM_DECRYPT)
                {
//OFN struct, used for the browser dlgbox type..
//============================================
    ofn.lpstrFilter = "File Encryptor Files (*.fem)\0*.fem\0";
    ofn.lpstrDefExt = "fem";
//============================================
                  if(GetOpenFileName(&ofn))
                  {
                  FileW.StartCrypt(ofn.lpstrFile, ACTION_DECRYPT);
                  }
                  else
                  {
                  MessageBox(hwnd, "Unexpected file error.", "ERROR!", MB_OK);
                  }
                }
/* End */       break;               
                case IDM_EXIT:
.......
..........
.................

at first i used szFileName but since it didn't work =( i replaced it for using directly the struct member.. lpstrFile

oh yeah don't make comments about my code commenting lol i was desperate and the smileys in the source code well lol seemed relaxing.. hehehe

i just spotted an error in the code.. but it won't solve anything regarding this problem.. =(

if(WM_COMMAND == IDM_...)...

***
change to
***
if(wParam == IDM...)...
.......

Hi. Don't you hate these seemingly unexplainable errors that turn out to be simple?

Code:

 case (IDM_ENCRYPT || IDM_DECRYPT):
/* start */
                //OFN struct, used for the browser-type dlgbox
                //======================================
                OPENFILENAME ofn;

A. I've never seen || used in a case label.
B. Variables in C must be declared at the start of a code block. That is after a {

Solution:

Code:

case IDM_ENCRYPT:    // Fall through
case IDM_DECRYPT: {  // Start code block
// Now declare variables...

Code:

struct buffy
{
int yay;
};

buffy yp;

I believe that if you don't use a typedef you must declare a struct as such:

Code:

struct buffy yp;

Welcome to the forums. In the future please post the error messages. This makes it a lot easier to see the problem.

Have Fun!

Edit: In the unlikely event that you are actually compiling as C++ and not C you can ignore most of this post.

AFAIK this is not legal in C/C++ (could be in PASCAL)

case (IDM_ENCRYPT || IDM_DECRYPT)://error but not picked up by the compiler
In MSVC .NET the case declared as you have will never be entered. The default, if present, will be used instead.

as case is defined as

case constant-expression :

ie

case IDM_ENCRYPT:
case IDM_DECRYPT:
// code
break;


As the complier does not think you have entered this section of the code (the bad case statement which can never be entered) it does not declare the variable.

Unless you have a good reason for declaring variables inside blocks of code, and not at the function start, then don't.


Try putting the case as per my example and declare the open file name dlg struct at the very begining of the callback.
This will solve the problem.

edit;

try

Code:

case WM_COMMAND:
	idControl	= GET_WM_COMMAND_ID(wParam,lParam) ;
        switch(idControl)
         {

The case statement is not evauated though. It is not a constant.

So it is simply ignored (on my complier MSVC.NET 2003) and the default is called.

Try it on your complier. See if you get the same result I did.

>.(where A, B were #defined as nonzero) was equal to 'case 1:'.

Was either one #defined as 1?

try

Code:

char     sText[64]={0};

for(int iVar=0;iVar<4;iVar++)
{
      switch (iVar)
       {
              case (0||2):
                   sprintf(sText,"reached 0, 2 case");
              break;
              case 1:
                   sprintf(sText,"reached 1 case");
              break;
              default:
                  sprintf(sText,"reached default case");
              break;
       }
       MessageBox(NULL, sText, "Test case", MB_OK);
}

See what I mean about the behavior being undefined?

Bcc5.5 throws 'duplicate case' error on this; my test was quite the same (#define A 0, #define B 2)

edit: gcc (3.3) does the same - evaluates A || B acc. to logical values of A, B and final case value is 0 or 1 (T or F)..