I am mixed up with std::string, std::wstring, QString, anscii, utf8, unicode and charset

[szuzsq]

In windows, I have a class to converter charset between anscii, utf8, unicode.
But, I can not transfer it to mac or linux, Is Any one can help me?
@
#pragma once

#include <string>
#ifdef WIN32
#include <wtypes.h>
#endif

class CStringConverter {
public:
static std::wstring Ansi2Unicode(std::string szAnsi) {
#ifdef WIN32
//calc block size to be returned
int len = MultiByteToWideChar(CP_ACP, NULL, szAnsi.c_str(), szAnsi.size(), NULL, 0);
//malloc and fill the returned block
wchar_t* szUnicode = new wchar_t[len+1];
MultiByteToWideChar(CP_ACP, NULL, szAnsi.c_str(), szAnsi.size(), szUnicode, len);
szUnicode[len] = 0;
std::wstring rs = szUnicode;
delete[] szUnicode;

    return rs;

#else
//>>?? //fixme
return L"";
#endif
}
static std::string Unicode2Ansi(std::wstring szUnicode) {
#ifdef WIN32
//calc block size to be returned
int len = WideCharToMultiByte(CP_ACP, NULL, szUnicode.c_str(), szUnicode.size(), NULL, 0, NULL, NULL);
//malloc and fill the returned block
char* szAnsi = new char[len + 1];
WideCharToMultiByte(CP_ACP, NULL, szUnicode.c_str(), szUnicode.size(), szAnsi, len, NULL, NULL);
szAnsi[len] = 0;
std::string rs = szAnsi;
delete[] szAnsi;

    return rs;

#else
//>>?? //fixme
return "";
#endif
}

static std::wstring Utf82Unicode(std::string szUtf8) {

#ifdef WIN32
//calc block size to be returned
int len = MultiByteToWideChar(CP_UTF8, NULL, szUtf8.c_str(), szUtf8.size(), NULL, 0);
//malloc and fill the returned block
wchar_t* szUnicode = new wchar_t[len+1];
MultiByteToWideChar(CP_UTF8, NULL, szUtf8.c_str(), szUtf8.size(), szUnicode, len);
szUnicode[len] = 0;
std::wstring rs = szUnicode;
delete[] szUnicode;

    return rs;

#else
//>>?? //fixme
return L"";
#endif
}
static std::string Unicode2Utf8(std::wstring szUnicode) {
#ifdef WIN32
//calc block size to be returned
int len = WideCharToMultiByte(CP_UTF8, NULL, szUnicode.c_str(), szUnicode.size(), NULL, 0, NULL, NULL);
//malloc and fill the returned block
char* szUtf8 = new char[len + 1];
WideCharToMultiByte(CP_UTF8, NULL, szUnicode.c_str(), szUnicode.size(), szUtf8, len, NULL, NULL);
szUtf8[len] = 0;
std::string rs = szUtf8;
delete[] szUtf8;

    return rs;

#else
//>>?? //fixme
return "";
#endif
}

static std::string Ansi2Utf8(std::string szAnsi) {
    return Unicode2Utf8(Ansi2Unicode(szAnsi));
}
static std::string Utf82Ansi(std::string szUtf8) {
    return Unicode2Ansi(Utf82Unicode(szUtf8));
}

};
@

---

@
//demo
std::string szAnsi = "abc123你我他";

std::wstring szUnicode = CStringConverter::Ansi2Unicode(szAnsi);
szAnsi = CStringConverter::Unicode2Ansi(szUnicode);

std::string szUft8 = CStringConverter::Unicode2Utf8(szUnicode);
szUnicode = CStringConverter::Utf82Unicode(szUft8);

szAnsi = CStringConverter::Utf82Ansi(szUft8);
szUft8 = CStringConverter::Ansi2Utf8(szAnsi);
@

[luisvaldes88]

hi, very cool...

I thing QString support everthing you are trying to do. I'm not 100% sure.
did you take a look at QString documentation?

best regards

[fluca1978]

The only conversion that seems to me not possible is from std::wstring, I don't see any constructor of QString with this signature, but all other conversions seems available to me.
Anyway, what is the problem you are having converting your class?

[cincirin]

[quote author="fluca1978" date="1318329336"]The only conversion that seems to me not possible is from std::wstring[/quote] It is "fromStdWString":http://doc.qt.nokia.com/latest/qstring.html#fromStdWString

[fluca1978]

Ops...I didn't notice it!

[cincirin]

[quote author="fluca1978" date="1318331294"]Ops...I didn't notice it![/quote]
No problem :-)
Qt framework is big enough for someone to know everything ... ( at least I think so :-) )

[szuzsq]

My new implement is like this:
@
#pragma once

#include <string>
#include <QString>

class CStringConverter {
public:
static std::wstring Ansi2Unicode(std::string szAnsi) {
QString s = QString::fromAscii(szAnsi.c_str());
return s.toStdWString();
}
static std::string Unicode2Ansi(std::wstring szUnicode) {
QString s = QString::fromStdWString(szUnicode);
return s.toAscii();
}

static std::wstring Utf82Unicode(std::string szUtf8) {
    QString s = QString::fromUtf8(szUtf8.c_str());
    return s.toStdWString();
}
static std::string Unicode2Utf8(std::wstring szUnicode) {
    QString s = QString::fromStdWString(szUnicode);
    return s.toUtf8();
}

static std::string Ansi2Utf8(std::string szAnsi) {
    return Unicode2Utf8(Ansi2Unicode(szAnsi));
}
static std::string Utf82Ansi(std::string szUtf8) {
    return Unicode2Ansi(Utf82Unicode(szUtf8));
}

};
@

---

But there are still problem.
1).it can be compiled using qt under windows, but can not be compiled under mac?
@
static std::string Unicode2Ansi(std::wstring szUnicode) {
QString s = QString::fromStdWString(szUnicode);
//in this statement, the error is:
//converstion from 'QByteArray' to non-scalar type 'std::string' requested
return s.toAscii();
}
@

[szuzsq]

BTW,I try to create a "Qt Console Application" in Mac
@
#untitiled2.pro
QT += core

QT -= gui

DEFINES += UNICODE

TARGET = untitled2
CONFIG += console
CONFIG -= app_bundle

TEMPLATE = app

SOURCES += main.cpp

@

@
//main.cpp
#include <QtCore/QCoreApplication>

#include<string>
#include<QString>

int main(int argc, char *argv[])
{
QCoreApplication a(argc, argv);

std::string stdStr = "abc123你我他_stdStr";
std::wstring stdWStr = L"abc123你我他_stdWStr";

QString qStr="abc123你我他_qStr";

printf(stdStr.c_str());
printf("\n");

wprintf(stdWStr.c_str());
printf("\n");

wprintf(qStr.toStdWString().c_str());
printf("\n");


return a.exec();

}
@

when I run in debug mode, I add a breakpoints before
@
return a.exec();
@
and open "Locals and Expressions" window,
but in that window, there are not any string value.
the table in "Locals and Expressions" window is:
@
Name Value Type

qStr QString
stdStr string
stdWStr wstring
@

the "value" field are null.
I may want to view at the "value" field: "abc123你我他_stdStr","abc123你我他_stdWStr","abc123你我他_qStr", for the std::string,std::wstring,QString variable value.

Is any way for me to watch the std::string, std::wstring, QString variable value, when I debug?

---

And when I watch X11 console,it print as:

abc123ä½ æä»_stdStr
abc123_stdWStr
abc123ä½ æä»_qStr

maybe, it look like don't support chinese or japanese language.

How can I make X11 console to print exactly?
as:
abc123你我他_stdStr
abc123你我他_stdWStr
abc123你我他_qStr

but this program run under window.
the table in "Locals and Expressions" window is:
@
Name Value Type

qStr "abc123ÄãÎÒËû_qStr" QString
stdStr 0x3f9da0 "abc123???" string
stdWStr class std::basic_string<> wstring
@

---

And the console print as:

abc123你我他stdStr
abc123
abc123你我他_qStr

[cincirin]

[quote author="szuzsq" date="1318393533"]
1).it can be compiled using qt under windows, but can not be compiled under mac?
@
static std::string Unicode2Ansi(std::wstring szUnicode) {
QString s = QString::fromStdWString(szUnicode);
//in this statement, the error is:
//converstion from 'QByteArray' to non-scalar type 'std::string' requested
return s.toAscii();
}
@[/quote]

Your "Unicode2Ansi" function involve calling "QByteArray::operator const char * ()":http://doc.qt.nokia.com/latest/qbytearray.html#operator-const-char--2a Maybe on Mac QT_NO_CAST_FROM_BYTEARRAY is implicity defined.
Why you do not call directly "QString::toStdString":http://doc.qt.nokia.com/latest/qstring.html#toStdString ?

[szuzsq]

Because of I using 2 different sdk, and someone using char*, and someone using wchat_t*.
so I must convert between them.

In the last demo, I use std::string, QString, std::wstring(convert to std::string), to write a simple xml file.
@
#include "tinyxml.h"

#include <string>
#include <QString>

int main(int argc, char* argv[]) {
TiXmlDocument doc;
TiXmlElement elem("rene");

//directly using std::string and QString
std::string s = "_你_";
elem.SetAttribute("s", s.c_str());

QString qs = "_我_";
elem.SetAttribute("qs",qs.toStdString().c_str());

//convert std::wstring to std::string
//using by QString for adapter
std::wstring ws = L"_他_";
QString qs2 = QString::fromStdWString(ws);
std::string s2 = qs2.toStdString();
elem.SetAttribute("ws", s2.c_str());

doc.InsertEndChild(elem);

doc.SaveFile("0.xml");

}
@

but the result xml file have some error:
@
<rene s="你" qs="我" ws="?" />
@

and I look forward to the right xml file like:
@
<rene s="你" qs="我" ws="他" />
@

[szuzsq]

when I invoke
@
char* rs = setlocale(LC_CTYPE, "");
@

in windows, the result is: "Chinese_People's Repulic of China 936";
but in mac, the result is :"C".

Is there any wrong?

[cincirin]

Well, then have a look at "convert a QString to char*":http://developer.qt.nokia.com/faq/answer/how_can_i_convert_a_qstring_to_char_and_vice_versa using locale settings.

[szuzsq]

[quote author="cincirin" date="1318412939"]Well, then have a look at "convert a QString to char*":http://developer.qt.nokia.com/faq/answer/how_can_i_convert_a_qstring_to_char_and_vice_versa using locale settings.[/quote]

thank you, but the wrong is also exist.

in fact, if I use QString::fromStdString, fromStdWString, toStdString, toStdWString to convert between std::string and std::wstring,
with latin or ascii word, it run ok, not wrong;
but with chinese or japanese word, it run wrong....

[Franzk]

Note that your compiler may actually have trouble reading non-ascii characters from your file.

[giesbert]

from/toStdString always converts the text. If you want to be safe, you MUST know, which char* convention you use. Codepage based? Utf-8?

I personally prefer utf-8, which is simple with QString:

@
QByteArray ba = string.toUtf8();
@

The byte array has the correct length and a const char* conversion so you can put it to an std::string.