C / C++ / MFC

I'm stuck, not sure which direction to go

2-Dec-11 12:19

I saw that too. I think there are 2 of them, 1 in string, and 1 in istream, and that the sample I found on c++ website was in mistake of it's use. I'm using fstream, and it does not exist in that header.

hmmmmmmmmmm

jkirkerx2-Dec-11 19:16

Re: I'm stuck, not sure which direction to go

2-Dec-11 19:16

I got the getline to work, but I can't figure out how to advance to the next line.

And even if I can get to the next line, I can't figure out how to build the buffer with sql command text until I reach GO. Sure like that stringbuilder in asp.net.

The wifstream was suppose to return the unicode chars from the file, but I get ansi chars.

I'm close here, just need lots of refinement.

Any ideas are welcomed.

// Load the SQL Server Commands
    std::wifstream *sqlDBFile = new std::wifstream;
    sqlDBFile->open(szFilePath);

    // if the file stream opened
    if (sqlDBFile->is_open() ) {

        // While file stream is in memory
        while (sqlDBFile->good() ) {
            // Make a temp sql command buffer
            WCHAR *pzSQLCommand = new WCHAR[1024];

            while (sqlDBFile->good() ) {
                // read the line of text
                WCHAR szAnsiBuffer[1024];
                sqlDBFile->getline(szAnsiBuffer, 1024, L'\n');

                int jdx = 0;
                WCHAR *szBufferLine = new WCHAR[1024];
                for (int i = 2; i < 1024; ++i) {
                    szBufferLine[jdx] = szAnsiBuffer[i];
                    i++;
                    jdx++;
                }
                int iLength = wcslen(szBufferLine);
                szBufferLine[iLength] = L'\0';

                // if text != GO, add to buffer
                if ((wcsncmp( szBufferLine, L"GO", wcslen(L"GO") )) != 0) {
                    wcsncat_s(pzSQLCommand, wcslen(pzSQLCommand), szBufferLine, wcslen(szBufferLine) );

                    pzSQLCommand[wcslen(pzSQLCommand) + 1] = L'\0';
                    break;
                }
                else {
                    // if text = GO, terminate the buffer and execute the buffered command
                    hr = pICommandText->SetCommandText( DBGUID_DBSQL, pzSQLCommand );
                    hr = pICommandText->QueryInterface( IID_ICommandProperties, ( void ** ) &pICommandProperties );
                    hr = pICommandText->Execute( NULL, IID_IRowset, NULL, &cRowsAffected, ( IUnknown ** ) &pIRowset );

                }
            }
        }

Richard MacCutchan3-Dec-11 0:03

Richard MacCutchan

3-Dec-11 0:03

C++

// read the line of text
WCHAR szAnsiBuffer[1024];
sqlDBFile->getline(szAnsiBuffer, 1024, L'\n');

int jdx = 0;
WCHAR *szBufferLine = new WCHAR[1024];
for (int i = 2; i < 1024; ++i) {
    szBufferLine[jdx] = szAnsiBuffer[i];
    i++;
    jdx++;
}
int iLength = wcslen(szBufferLine);
szBufferLine[iLength] = L'\0';

I'm not sure what you are trying to do here but most of this code is redundant. You have defined your input buffer as WCHAR so the resulting line will be Unicode, There is no need (in fact it's quite dangerous) for the copying character by character. All you need is:

C++

// read the line of text
WCHAR szBufferLine[1024];
sqlDBFile->getline(szBufferLine, 1024, L'\n');

This will leave you with a null terminated array of Unicode characters in szBufferLine, which you can then test, for the L"GO" string, and append to your command.

A couple of other things that you should also do:

1. When you create the pzSQLCommand buffer, is to add a null at the beginning so your wcsncat_s() calls do not trash your memory, thus:

C++

WCHAR *pzSQLCommand = new WCHAR[1024];
*pzSQLCommand = L'\0';

2. Use the proper length setting of your destination buffer (see here[^]) in your wcsncat_s() call thus:

C++

wcsncat_s(pzSQLCommand, 1024, szBufferLine, wcslen(szBufferLine) );

Unrequited desire is character building. OriginalGriff
I'm sitting here giving you a standing ovation - Len Goodman

Re: I'm stuck, not sure which direction to go

jkirkerx3-Dec-11 8:00

3-Dec-11 8:00

The code is for reading a sql text file, which is exported from sql management. By running the file, you can create a database and populate the data.

It's weird, or I'm reading the results wrong, but on get line, I get the every other character just like the socket program I wrote. I guess I am misinterpreting the data, and it is real unicode, with the 2nd character not programmed.

So one of the original code modules I had was like yours except for the small changes, but I was using ofstream, that kept erasing my file.

The cat works now since placing the null terminator in the creation of the pzSQLCommand, but only cats the first character "y", for "p" must be a special character that halts the copy. I need to walk through it and see.

The getline command returns this. LC 0 LR 0 LE 0 LA 0 LT 0 LE 0 and so on

+		szCommandLine	0x0012f0e0 "ÿþC"	wchar_t [1024]

The first file line, plus rest of command for reference

SQL

CREATE TABLE [dbo].[BatchSettlementIndex] (
  [BatchID] int IDENTITY(1, 1) NOT NULL,
  [BatchName] varchar(80) COLLATE SQL_Latin1_General_CP1_CI_AS NOT NULL,
  [BatchDate] datetime NULL,
  [BatchAmount] decimal(10, 2) NULL
)
ON [PRIMARY]
GO

I guess I should move forward 1 move and get the next line working first, since the mechanics are better now. Your wisdom has solidified the design, in which I thank you for. If you think of something let me know.

FYI:
I'm almost done with all the little programs I need for my wizard. This is the next to last piece I need, then I can take all the parts, and compile them into some dll's within the project, and make a new gui to support them.

It's to the point where I have gone far beyond the need of InstallShield 2012, and will try the stock deployment to deploy this program, and let it download the asp.net stuff I wrote.

Nextline

jkirkerx3-Dec-11 8:26

i can't load dll builded in debug unicode,what's difference between 'debug' and 'debug unicode'?

3-Dec-11 8:26

Well that was fast. The nextline works now. One more to go here, I'll work on the first 2 characters when I get back in 3 hours.

// Load the SQL Server Commands
    std::wifstream *sqlDBFile = new std::wifstream;
    sqlDBFile->open(szFilePath, std::wifstream::in);

    // if the file stream opened
    if (sqlDBFile->is_open() ) {

        // While file stream is in memory
        while (sqlDBFile->good() ) {
            // Make a temp sql command buffer
            WCHAR *pzSQLCommand = new WCHAR[1024];
            *pzSQLCommand = L'\0';

            while ( sqlDBFile->get() ) {
                // read the line of text
                WCHAR szCommandLine[1024];
                sqlDBFile->getline(szCommandLine, 1024, L'\n');


                // if text != GO, add to buffer
                if (szCommandLine != L"GO") {
                    wcsncat_s(pzSQLCommand, 1024, szCommandLine, wcslen(szCommandLine) );
                    break;
                }
                else {
                    // if text = GO, terminate the buffer and execute the buffered command
                    hr = pICommandText->SetCommandText( DBGUID_DBSQL, pzSQLCommand );
                    hr = pICommandText->QueryInterface( IID_ICommandProperties, ( void ** ) &pICommandProperties );
                    hr = pICommandText->Execute( NULL, IID_IRowset, NULL, &cRowsAffected, ( IUnknown ** ) &pIRowset );

                }
            }
        }

        // close the sql command file
        sqlDBFile->close();
        sqlDBFile = NULL;
    }

Re: Nextline

Richard MacCutchan3-Dec-11 9:13

Richard MacCutchan

3-Dec-11 9:13

You got it. Remember WCHAR is a 16 bit character set, many of which map 1:1 with the ASCII set padded with a zero byte, which is why it looks like an ASCII character followed by a zero.

Unrequited desire is character building. OriginalGriff
I'm sitting here giving you a standing ovation - Len Goodman

Member 77984732-Dec-11 5:06

Member 7798473

2-Dec-11 5:06

i've some code written as a dll in C++ in Visual Studio 2008. I've to be able to debug it. But i can't load dll in test exe. There isn't any problem on my test exe because when i try to load another dll there isn't any problem. Dll that given to me builded in debug unicode. Is it may be the cause of the problem? And how can i convert it to just debug mode? i really need your help., Thanks in advance.

Re: i can't load dll builded in debug unicode,what's difference between 'debug' and 'debug unicode'?

JackDingler2-Dec-11 5:19

JackDingler

2-Dec-11 5:19

The debug and debug unicode project settings load different versions of the Microsoft MFC libraries and DLLs.

All of your MFC based projects, need to be compiled and built with the same project settings.

Re: i can't load dll builded in debug unicode,what's difference between 'debug' and 'debug unicode'?

Erudite_Eric2-Dec-11 5:44

Erudite_Eric

2-Dec-11 5:44

So, as the other poster said, rebuld your test exe with unicode debug set. (Hope you used the _T before all strings to enable the project to be built both ways.)

==============================

Nothing to say.

Re: i can't load dll builded in debug unicode,what's difference between 'debug' and 'debug unicode'?

Richard MacCutchan2-Dec-11 6:20

Richard MacCutchan

2-Dec-11 6:20

It makes no difference how the DLL is built as to whether you can use it or not, you just have to use the correct calling sequences. It would be better if you explained exactly what you are doing or trying to do, show some code if it will help, and what results you see.

Unrequited desire is character building. OriginalGriff
I'm sitting here giving you a standing ovation - Len Goodman

class object as parameter of Template Function

002comp2-Dec-11 0:29

002comp

2-Dec-11 0:29

Hello Friends

I am creating a Function template Whose declaration is like this

C++

template <class T>
void Test(string name,T value,bool exist)

Defination 
template<class T> void sample::Test(string name,T value,bool Exist)

Now,i am passing Color class object to fn Test,like this

Color c;
Test<Color>("name",c,true);

And Its Giving Linking error that
error LNK2019: unresolved external symbol "public: void __thiscall sample::Test<class color="">

Any Ideas?
Regards
Yogesh

CPallini2-Dec-11 0:43

2-Dec-11 0:43

Where did you put such statements?

If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler.
-- Alfonso the Wise, 13th Century King of Castile.

This is going on my arrogant assumptions. You may have a superb reason why I'm completely wrong.
-- Iain Clarke

[My articles]

002comp2-Dec-11 0:50

002comp

2-Dec-11 0:50

I put these statements in my program.
I provide the overview of what i am Doing.

Regards
Yogesh

CPallini2-Dec-11 0:53

2-Dec-11 0:53

'my program' is a bit vague. Do you know, for instance, that template method implementations should go into header files (not into source files)?

image processing toolkits | batch image processing

C^hris L^osinger2-Dec-11 1:29

C^hris L^osinger

2-Dec-11 1:29

(unless you are only going to use that template inside that one source file)

CPallini2-Dec-11 1:32

2-Dec-11 1:32

Good point. Thumbs Up | :thumbsup:

002comp2-Dec-11 1:30

002comp

2-Dec-11 1:30

I was putting separate in header and cpp files.

Now,i put its definition in header itself.No Errors.

Will u please Explain Why it is so ?

Thanks & Regards
Yogesh

CPallini2-Dec-11 1:44

2-Dec-11 1:44

See, for instance: "[35.12] Why can't I separate the definition of my templates class from its declaration and put it inside a .cpp file?"
[^].