Windows下Perforce Command Line处理中文参数的两种方法

Windows下Perforce Command Line处理中文参数的两种方法

    Windows下Command Line中的中文字符是采用ANSI编码来处理的，而Perforce command line要求传入的中文参数（文件（夹）名）为UTF-8编码，所以需要将中文参数转换为UTF-8后再做处理，下面介绍两种处理方法。

一、系统调用

    通过使用WinExec、ShellExecute或system，如：

_snprintf(cmdbuf , 1024 , "p4 -c %s add \"%s\"" , argv[1] , ANSIToUTF8(path.c_str()));
system(cmdbuf);

    此方法的虽然简单，但存在效率问题，因为有创建进程的巨大开销。此外不同版本的Perforce也会出现不同的执行结果，针对特定的中文会出现操作失败的诡异问题。

二、Perforce SDK

    Perforce提供有SDK用以扩展或集成到其他应用中，虽然没有详细的文档，但可以通过学习SDK中的sample文件来学习，此方法最稳定。

    下面代码展示了通过SDK中ClientAPI来递归添加指定文件夹下的所有文件：

# include "clientapi.h"
# include "i18napi.h"
# include "CharSetConvertUtil.h"
# include <string>
# include <vector>
# include <list>
# include <io.h>

using namespace std;

// structure to hold a directory and all its filenames.
struct FILELIST
{
    string path;
    vector<string> theList;
};


void TransverseDirectory(string path, list<FILELIST>& theList)
{
    struct _finddatai64_t data;
    string fname = path + "\\*.*";
    long h = _findfirsti64(fname.c_str(),&data);
    if(h >= 0)
    {
        FILELIST thisList;
        theList.push_back(thisList);
        list<FILELIST>::iterator it = theList.end();
        it--;

        (*it).path = path;
        do {
            if( (data.attrib & _A_SUBDIR) )
            {
                // make sure we skip "." and "..".  Have to use strcmp here because
                // some file names can start with a dot, so just testing for the
                // first dot is not suffient.
                if( strcmp(data.name,".") != 0 &&strcmp(data.name,"..") != 0)
                {
                    // We found a sub-directory, so get the files in it too
                    fname = path + "\\" + data.name;
                    // recurrsion here!
                    TransverseDirectory(fname,theList);
                }

            }
            else
            {
                // this is just a normal filename.  So just add it to our vector
                (*it).theList.push_back(data.name);

            }
        }while( _findnexti64(h,&data) == 0);
        _findclose(h);

    }
}

int main( int argc, char **argv )
{
    list<FILELIST> MyList;
    string path;
    ClientUser ui;
    ClientApi client;
    StrBuf msg;
    Error e;

    if(argc < 4)
    {
        fprintf( stderr , "P4 Transverse Add: Arguments Error!\n");
        return -1;
    }

    client.SetPort(argv[1]);
    client.SetClient(argv[2]);
    client.SetTrans(CharSetApi::UTF_8 , CharSetApi::UTF_8 , 
        CharSetApi::UTF_8,
        CharSetApi::UTF_8);

    TransverseDirectory(argv[3],MyList);

    // Connect to server
    client.Init( &e );

    if( e.Test() )
    {
        e.Fmt( &msg );
        fprintf( stderr, "%s\n", msg.Text() );
        return -1;
    }

    list<FILELIST>::iterator it;
    for(it = MyList.begin(); it != MyList.end(); it++)
    {
        vector<string>::iterator its;
        for(its = (*it).theList.begin(); its != (*it).theList.end(); its++)
        {
            path = (*it).path + "\\" + (*its);
            char* pText = ANSIToUTF8(path.c_str());
            client.SetArgv( 1 , &pText);
            client.Run( "add" , &ui );
        }
    }

    // Close connection
    client.Final( &e );

    if( e.Test() )
    {
        e.Fmt( &msg );
        fprintf( stderr, "%s\n", msg.Text() );
        return -1;
    }
    
    return 0;
}

    此方法省去了创建p4进程的开销，任务执行效率会提高不少，而且也不会出现执行结果不稳定的问题。   

 

附一：SDK下载地址

ftp://ftp.perforce.com/perforce/

附二：附上ANSI转UTF-8代码

wchar_t* ANSIToUnicode( const char* str )
{
    int    textlen ;
    wchar_t * result;
    textlen = MultiByteToWideChar( CP_ACP, 0, str,-1,    NULL,0 );  
    result = (wchar_t *)malloc((textlen+1)*sizeof(wchar_t));  
    memset(result,0,(textlen+1)*sizeof(wchar_t));  
    MultiByteToWideChar(CP_ACP, 0,str,-1,(LPWSTR)result,textlen );  
    return    result;  
}

char* UnicodeToANSI( const wchar_t *str )
{
    char * result;
    int textlen;
    // wide char to multi char
    textlen = WideCharToMultiByte( CP_ACP,    0,    str,    -1,    NULL, 0, NULL, NULL );
    result =(char *)malloc((textlen+1)*sizeof(char));
    memset( result, 0, sizeof(char) * ( textlen + 1 ) );
    WideCharToMultiByte( CP_ACP, 0, str, -1, result, textlen, NULL, NULL );
    return result;
}

wchar_t* UTF8ToUnicode( const char* str )
{
    int    textlen ;
    wchar_t * result;
    textlen = MultiByteToWideChar( CP_UTF8, 0, str,-1,    NULL,0 );  
    result = (wchar_t *)malloc((textlen+1)*sizeof(wchar_t));  
    memset(result,0,(textlen+1)*sizeof(wchar_t));  
    MultiByteToWideChar(CP_UTF8, 0,str,-1,(LPWSTR)result,textlen );  
    return    result;  
}

char* UnicodeToUTF8( const wchar_t *str )
{
    char * result;
    int textlen;
    // wide char to multi char
    textlen = WideCharToMultiByte( CP_UTF8,    0,    str,    -1,    NULL, 0, NULL, NULL );
    result =(char *)malloc((textlen+1)*sizeof(char));
    memset(result, 0, sizeof(char) * ( textlen + 1 ) );
    WideCharToMultiByte( CP_UTF8, 0, str, -1, result, textlen, NULL, NULL );
    return result;
}

char* ANSIToUTF8(const char* str)
{
    wchar_t* pUnicodeBuff = ANSIToUnicode(str);
    char* pUtf8Buff = UnicodeToUTF8(pUnicodeBuff);
    free(pUnicodeBuff);

    return pUtf8Buff;
}