EM_GETLINE And Unicode?

does EM_GETLINE support unicode? if not then is there something simmiliar that supports unicode and if so how come unicode chars are showing as ?'s

>> does EM_GETLINE support unicode? <<

Yes.

>> if so how come unicode chars are showing as ?'s <<

Firstly, EM_GETLINE does not null terminate the buffer so you have to take care of this. It depends where you are outputting the unicode text after you have retrieved it with EM_GETLINE and font support. In my experience, the console and MessageBox, at least with the default font settings, are not able to render many unicode characters. Edit controls and internet browsers seem to do a much better job.

hmmm it should be working its reading from a rtf control and printing to a rtf control and
#define UNICODE
#define _UNICODE is in there :S

heres the code im using

Code:

---

        TCHAR JoinText[128] = {0};
        TCHAR RBN[50] = {0};
        TCHAR pbMsg[50] = {0};
        TCHAR teststring[128] = {0};
        int npt = 0;
        int NumLines;


        NumLines = SendMessage(hwndOut , EM_GETLINECOUNT , (WPARAM)0 , (LPARAM)0);
        *((LPINT)JoinText) = sizeof(JoinText);
        npt = SendMessage(hwndOut, EM_GETLINE, (WPARAM)lines-2, (LPARAM)JoinText);
        JoinText[npt] = 0;
        _stprintf(teststring , _T("%s") , JoinText);
        SendMessage(HH , WM_SETTEXT , (WPARAM)0 , (LPARAM)teststring);

---

and it shows ?s for the unicode

This line:

Code:

---

        *((LPINT)JoinText) = sizeof(JoinText);

---

should be:

Code:

---

        *((LPWORD)JoinText) = sizeof(JoinText) / sizeof(TCHAR); // or use ARRAYSIZE macro

---

but that isn't causing your problem.

Are you sure that UNICODE and _UNICODE are #defined above all #includes?

You can test if unicode is in effect:

Code:

---

if (sizeof(_T("x")) != 4 || sizeof(TEXT("x")) != 4)
      MessageBox(NULL, _T("NO Unicode!"), NULL, 0);
else
      MessageBox(NULL, _T("YES Unicode!"), NULL, 0);

---

its still not working , doing the same thing as before :(

also yeh unicode is defined above the #includes.

Win95, 98 and ME do not support UNICODE in the standard system libraries. NT4 an later is internally based on UNICODE. Maybe you are testing on 9x?

Code:

---

        TCHAR JoinText[128] = {0};
        TCHAR PBN[50] = {0};
        TCHAR pbMsg[50] = {0};
        TCHAR PBL[128] = {0};




        FILE *pbf;

        int npt = 0;

        int NumLines;

        NumLines = SendMessage(hwndOut , EM_GETLINECOUNT , (WPARAM)0 , (LPARAM)0);
        *((LPWORD)JoinText) = sizeof(JoinText) / sizeof(TCHAR); // or use ARRAYSIZE macro
        npt = SendMessage(hwndOut, EM_GETLINE, (WPARAM)lines-2, (LPARAM)JoinText);
        JoinText[npt] = 0;

---

thats on windowsXP the unicode is defined above all includes its at the very top.

it shows unicode chars as question marks , ive tried showing it in a messagebox and tried sending it to a RTF control which works with some other parts of the program(sending unicode to a rtf control).
i think its something with the way its reading it?

any one out there this is really important and i cant work it out

The behaviour seems most unusual. The only thing I can suggest is that you remove the #defines for UNICODE and _UNICODE from any files you've placed them in and pass the defines directly to your compiler itself. This will ensure that they are defined for your whole project. Your compiler documentation should have details on the correct syntax - for example, for gcc:

Code:

---

-DUNICODE -D_UNICODE

---

didnt make any difference its still the same :(

im using lcc-win32 for it and i thought it could be the problem so i tried it in dev-c++ and its still the same

Could you zip up and attach your dev-cpp project and/or your lcc-win32 code + makefile (if they're not too large) so that we can maybe take a closer look at it?

Code:

---

#define UNICODE
#define _UNICODE
#include <windows.h>
#include <tchar.h>
#define ID_TIMER 1
VOID CALLBACK PermBan(HWND, UINT, UINT, DWORD);
LRESULT CALLBACK WindowProcedure(HWND, UINT, WPARAM, LPARAM);

  TCHAR szClassName[] = _T("Class Name");
  HWND hwnd , HackedHandle;
  MSG messages;
  WNDCLASSEX wincl;
  POINT pt;

// Start WinMain() ...

int WINAPI WinMain(HINSTANCE hThisInstance, HINSTANCE hPrevInstance, LPSTR lpszArgument, int nFunsterStil) {


  wincl.hInstance = hThisInstance;
  wincl.lpszClassName = szClassName;
  wincl.lpfnWndProc = WindowProcedure;
  wincl.style = CS_DBLCLKS;
  wincl.cbSize = sizeof(WNDCLASSEX);
  wincl.hIcon = LoadIcon(NULL, IDI_APPLICATION);
  wincl.hIconSm = LoadIcon(NULL, IDI_APPLICATION);
  wincl.hCursor = LoadCursor(NULL, IDC_ARROW);
  wincl.lpszMenuName = NULL;
  wincl.cbClsExtra = 0;
  wincl.cbWndExtra = 0;
  wincl.hbrBackground = (HBRUSH) COLOR_BACKGROUND;

if(!RegisterClassEx(&wincl)) return 0;

hwnd = CreateWindowEx(0, szClassName, _T("Title"), WS_OVERLAPPEDWINDOW, CW_USEDEFAULT, CW_USEDEFAULT, 400, 300, HWND_DESKTOP, NULL, hThisInstance, NULL);

ShowWindow(hwnd, SW_SHOW);
UpdateWindow(hwnd);

while(GetMessage(&messages, NULL, 0, 0))
{
  TranslateMessage(&messages);
  DispatchMessage(&messages);
}

  return messages.wParam;
}

// Start WindowProcedure() ...

LRESULT CALLBACK WindowProcedure(HWND hwnd, UINT message, WPARAM wParam, LPARAM lParam) {

switch (message)
{

  case WM_DESTROY:
    PostQuitMessage(0);
  break;

  case WM_COMMAND:
    // Processing ...
  break;
 
  case WM_CREATE:
 
          MessageBox(0 , _T("") , _T("") , 0);
          GetCursorPos(&pt);
          HackedHandle = WindowFromPoint(pt);
          SetTimer(hwnd , ID_TIMER , 1000 , PermBan);


  default:
    return DefWindowProc(hwnd, message, wParam, lParam);

}

return 0;
}


VOID CALLBACK PermBan(HWND hwnd , UINT message , UINT uTimerID , DWORD dwTime)
{

        TCHAR JoinText[128] = {0};
        TCHAR PBN[50] = {0};
        TCHAR pbMsg[50] = {0};
        TCHAR PBL[128] = {0};
        TCHAR PBC128];

        int npt = 0;

        int NumLines;

        NumLines = SendMessage(HackedHandle , EM_GETLINECOUNT , (WPARAM)0 , (LPARAM)0);
        *((LPWORD)JoinText) = sizeof(JoinText) / sizeof(TCHAR);
        npt = SendMessage(HackedHandle, EM_GETLINE, (WPARAM)NumLines-2, (LPARAM)JoinText);
        JoinText[npt] = 0;

        if (_tcsstr(JoinText,_T("TesT")) && (!_tcsstr(JoinText,_T(">")))){
                MessageBox(0 , JoinText , _T("test") , 0);

        }
}

---

thats the dev c++ code,

any one out there i havnt got it working yet :(

You need to use SendMessageTimeout and not SendMessage if you want to obtain line counts etc from a multi-line edit/richedit that belongs to another application. Also, but not related to your original problem, if you're going to use SetTimer then make sure you kill it off with KillTimer.

thanks for the reply, i tried this and had the same effect as before.

UNICODE is not the problem, the window handle you're using probably is. Try using an explicit handle for a known, open and active richedit/edit control (use spy++ or, if you don't have it, get a free variation from here; use the tool to get the handle of an active edit/richedit control). When you compare that with whatever handle you're trapping in code you'll probably find they're different.

Other test code you might consider implementing include

• Testing the handle you are attempting to use to ensure it actually belongs to an edit/richedit control ( use GetWindowLong/GetWindowLongPtr to test the style bits for ES_MULTILINE (if this style is absent from the window under scrutiny then messages such as EM_GETLINE and EM_GETLINECOUNT will always fail); popup a messagebox if the ES_MULTILINE style is absent).
• Testing the return value from SendMessageTimeOut

If, after taking some or all of these steps, you still have problems then post/attach your amended code.