ReadFile Unicode Help

hi,
can I suggest an exercise to you ?

Run notepad, save the empty document with name EmptyA.txt
Write inside Hallo, give a return, and save as HalloA.txt, close program.

Open EmptyA.txt, save as EmptyU.txt, but changing Encoding from ANSI to unicode, close
Open HalloA.txt, save it as HalloU.txt, changing Encoding as above.

See Size on Properties of this 4 files:  
EmptyA = 0
EmptyU = 2 (an header)
HalloA  = 7 ( 5 of Hallo +CR+LF)
HalloU = 16 (2hdr, 10 for Hallo +CR(2)+LF (2) )

If you use Visual Studio, open these 4 files using normal file open, no auto but binary:
you'll see what is inside them. In Unicode, each file has 2 bytes header (FF FE)
and each char (also control char) stays on 2 bytes .
For normal char ( ABCde123,.$&()...control too), a byte is always 0.

For literature see:
http://en.wikipedia.org/wiki/Unicode
http://unicode.org/

for support
http://msdn.microsoft.com/en-us/library/2dax2h36.aspx
http://msdn.microsoft.com/en-us/library/dybsewaf.aspx

Returnig to your problem, by notepad, open these html files, and looking in Save As Dialog Box,
see if unicode is pre-selected. About code, decide if it is the case to open the file in binary mode
(I don't suggest) and skip header and read each char in two bytes, or correctly use _TCHAR
for type and appropriate API call

bye
vic

Well, first of all, reading a file that way - byte by byte - is terribly inefficiant. Have you tried the following? Also, the reason for the strange output is that you are outputting the contents as ANSI, use 'wprintf()' instead of 'printf()' as below:

     szFile = CreateFile( szFileName , GENERIC_READ, FILE_SHARE_READ , NULL , OPEN_EXISTING , FILE_ATTRIBUTE_NORMAL , NULL );
        if ( szFile != INVALID_HANDLE_VALUE ){
                szFileSize = GetFileSize( szFile, NULL );
                szBuffer  = new char[szFileSize + 1 ];
                memset(szBuffer, 0 , (szFileSize + 1) * sizeof( char ) ); 
                szReadFile = ReadFile( szFile, szBuffer, 1szFileSize, &szNumberOfBytesRead, NULL ); // read the entire file at once
                        
                if (szReadFile)
                {
                        wprintf(L"File Read Is:%s\n",(wchar_t*)szBuffer);  // output UNICODE buffer 
                }
                else
                {
                        printf("Last Error:%d\n",GetLastError());
                        //handle error here
        
                }
        delete [] szBuffer;
        }
        CloseHandle(szFile);  

Select allOpen in new window

One possible approach you could take is to switch to the "Use Unicode Character Set" option under project properties and change the code as in the code snippet below.

Next, after you have read the file use the IsTextUnicode function http://msdn.microsoft.com/en-us/library/dd318672(VS.85).aspx to determine if the read text is multi-byte or Unicode and print the text. For the test use Notepad to create a file called ANSI.txt save it with the ANSI encoding option and one called Unicode.txt with the encoding option labeled Unicode.

#include "stdafx.h"
#include "Windows.h"
 
void ReadFile(TCHAR* szFileName);
 
int _tmain(int argc, _TCHAR* argv[])
{
	ReadFile(_T("C:\\ANSI.txt"));
	ReadFile(_T("C:\\Unicode.txt"));
 
	system("pause");
	return 0;
}
 
void ReadFile(TCHAR* szFileName){
        HANDLE  szFile =0;
        DWORD   szFileSize = 0;
        TCHAR*   szBuffer;
        BOOL    szReadFile = false;
        int     i=0;
        DWORD   szNumberOfBytesRead =0;
 
        szFile = CreateFile( szFileName , GENERIC_READ, FILE_SHARE_READ , NULL , OPEN_EXISTING , FILE_ATTRIBUTE_NORMAL , NULL );
        if ( szFile != INVALID_HANDLE_VALUE ){
                szFileSize = GetFileSize( szFile, NULL );
                szBuffer  = new TCHAR[szFileSize + 1];
                memset(szBuffer, 0 , (szFileSize + 1) * sizeof( TCHAR ) ); 
                do{
                        szReadFile = ReadFile( szFile, &szBuffer[i++], sizeof(TCHAR) , &szNumberOfBytesRead, NULL );
                } while(szNumberOfBytesRead != 0);
        
                if (szReadFile)
                {
					// If the file is not Unicode
					if(!IsTextUnicode((LPVOID)szBuffer, szFileSize, NULL) == TRUE) {											
						printf("ANSI File Read Is:%s\n", (PCHAR)szBuffer);
					}
					else {
						_tprintf(_T("Unicode File Read Is:%s\n"),szBuffer);
					}
                }
                else
                {
                    _tprintf(_T("Last Error:%d\n"),GetLastError());        
                }
        delete [] szBuffer;
        }
        CloseHandle(szFile);    
}

Select allOpen in new window

UnicodeOption.png

Thank you dolomiti for giving me that piece of info, and thank you jkr for the code fix..

But still my problem is I am trying to read both unicode and non unicode files alternatively, if I use jkr code I can only parse non unicode data and when there is unicode I can only see the unicode limiting bytes 0xFF or whatever they are..
If anyone can direct me how to write a function that supports both unicode and non unicode and how I can be able to parse both..

Regards

>>If anyone can direct me how to write a function that supports both unicode
>>and non unicode and how I can be able to parse both..

No problem ;o)

     szFile = CreateFile( szFileName , GENERIC_READ, FILE_SHARE_READ , NULL , OPEN_EXISTING , FILE_ATTRIBUTE_NORMAL , NULL );
        if ( szFile != INVALID_HANDLE_VALUE ){
                szFileSize = GetFileSize( szFile, NULL );
                szBuffer  = new char[szFileSize + 1 ];
                memset(szBuffer, 0 , (szFileSize + 1) * sizeof( char ) );
                szReadFile = ReadFile( szFile, szBuffer, 1szFileSize, &szNumberOfBytesRead, NULL ); // read the entire file at once
                       
                if (szReadFile)
                {
                        if(IsTextUnicode((LPVOID)szBuffer,szReadFile,NULL) // <------ Test here
                          wprintf(L"File Read Is:%s\n",(wchar_t*)szBuffer);  // output UNICODE buffer
                        else
                          printf("File Read Is:%s\n",szBuffer);  // output ANSIbuffer
                }
                else
                {
                        printf("Last Error:%d\n",GetLastError());
                        //handle error here
       
                }
        delete [] szBuffer;
        }
        CloseHandle(szFile);  

Thank you both, we're almost near.. Is there anyway to convert the detected unicode buffer to multi byte buffer so I can use it in the rest of my functions (using dynamic memory please )?

Regards

Thank you very much :)

Thank everyone that tried to help,  especially LordOfPorts and jkr

Regards