Alpha 2 Rev 11 Issue

[JR]

I have begin to test Aurora because I'm interested in a compiler that allows to use low-level COM more easily that my current development compiler.

After I got the code that calls CLSIDFromString working, I wanted to convert it in a function (see WINAPI_CLSIDFromString below), and cut and pasted the code. The new function didn't work properly. Finally, I discovered the problem: The function has an string parameter called psz and I also had this same variable declared inside the function.

Code Select Expand


sub WINAPI_CLSIDFromString(string psz), GUID
{
    int hr;
    GUID rclsid;
    int nLen;
    pointer pwcs;
    string psz;

    nLen = (len(psz) + 1) * 2;
    pwcs = AllocHeap(nLen);
    RtlZeroMemory(pwcs, nLen);
    hr = MultiByteToWideChar(0, 0, psz, len(psz), pwcs, nLen);
    hr = CLSIDFromString(pwcs, rclsid);
    FreeHeap(pwcs);
    return rclsid;
}


I think that the compiler must flag it as an error, with a message such "Duplicate variable name".

Code Select Expand


/*
Compile this program to run in a console window.
*/

struct GUID,1
{
  unsigned int data1;
  unsigned word data2;
  unsigned word data3;
  unsigned byte data4[8];
}

import void CoTaskMemFree(void *pv);
import int CLSIDFromString(unsigned word *lpsz, GUID *pclsid);
import int StringFromCLSID(GUID *rclsid, unsigned int *ppsz);
import int StringFromGUID2(GUID *rclsid, unsigned word *lpsz, int cchMax);

import int MultiByteToWideChar(
  unsigned int CodePage,          // code page
  unsigned int dwFlags,           // character-type options
  unsigned byte *lpMultiByteStr,  // string to map
  int cbMultiByte,                // number of bytes in string
  unsigned word *lpWideCharStr,   // wide-character buffer
  int cchWideChar                 // size of buffer
);

import int WideCharToMultiByte(
  unsigned int CodePage,            // code page
  unsigned int dwFlags,             // performance and mapping flags
  unsigned word *lpWideCharStr,     // wide-character string
  int cchWideChar,                  // number of chars in string.
  unsigned byte *lpMultiByteStr,    // buffer for new string
  int cbMultiByte,                  // size of buffer
  unsigned byte *lpDefaultChar,     // default for unmappable chars
  unsigned int *lpUsedDefaultChar   // set when default char used
);

import void RtlZeroMemory(void *Destination, unsigned int Length);

global sub main()
{

    GUID rclsid;

    rclsid = WINAPI_CLSIDFromString("{148BD52A-A2AB-11CE-B11F-00AA00530503}");
    print(WINAPI_StringFromGUID2(rclsid));

    do{}until getkey() <> "";
    return;

}

sub WINAPI_CLSIDFromString(string psz), GUID
{
    int hr;
    GUID rclsid;
    int nLen;
    pointer pwcs;

    nLen = (len(psz) + 1) * 2;
    pwcs = AllocHeap(nLen);
    RtlZeroMemory(pwcs, nLen);
    hr = MultiByteToWideChar(0, 0, psz, len(psz), pwcs, nLen);
    hr = CLSIDFromString(pwcs, rclsid);
    FreeHeap(pwcs);
    return rclsid;
}

sub WINAPI_StringFromGUID2(GUID *rclsid), string
{
    int hr;
    string pwsz;
    string psz;
    hr = StringFromGUID2(rclsid, pwsz, 40);
    if hr = 0 return "";
    hr = WideCharToMultiByte(0, 0, pwsz, -1, psz, 255, 0, 0);
    return psz;
}


[Paul Squires]

JOSE !!

Great to see you over here!

I am a Aurora newbie as well so I'm not much help to you, but I know that your COM experience will certainly benefit Aurora development.

[JR]

Hi Paul,

Glad to see you here! In fact, I'm not asking for help, but pointing to an issue with duplicate variable names.

I like the way in which COM support is being implemented in Aurora, with COM classes and interface pointers. It's exactly the approach that I have suggested to Bob without luck.

[Ionic Wind Support Team]

Fixed for the Alpha 3 update. 

[lviklund]

Just a small note of joy.

Paul S, Paul T and JosÃÆ'Ã,© on the same thread discussing a new programming tool, Aurora.
Woww

/Lasse
(Still stuck in Object-C)