qjpunicode.cpp

奇趣公司比较新的qt/emd版本

/****************************************************************************
**
** Copyright (C) 1992-2007 Trolltech ASA. All rights reserved.
**
** This file is part of the plugins of the Qt Toolkit.
**
** This file may be used under the terms of the GNU General Public
** License version 2.0 as published by the Free Software Foundation
** and appearing in the file LICENSE.GPL included in the packaging of
** this file.  Please review the following information to ensure GNU
** General Public Licensing requirements will be met:
** http://trolltech.com/products/qt/licenses/licensing/opensource/
**
** If you are unsure which license is appropriate for your use, please
** review the following information:
** http://trolltech.com/products/qt/licenses/licensing/licensingoverview
** or contact the sales department at [e-mail].
**
** In addition, as a special exception, Trolltech gives you certain
** additional rights. These rights are described in the Trolltech GPL
** Exception version 1.0, which can be found at
** http://www.trolltech.com/products/qt/gplexception/ and in the file
** GPL_EXCEPTION.txt in this package.
**
** In addition, as a special exception, Trolltech, as the sole copyright
** holder for Qt Designer, grants users of the Qt/Eclipse Integration
** plug-in the right for the Qt/Eclipse Integration to link to
** functionality provided by Qt Designer and its related libraries.
**
** Trolltech reserves all rights not expressly granted herein.
**
** This file is provided AS IS with NO WARRANTY OF ANY KIND, INCLUDING THE
** WARRANTY OF DESIGN, MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
**
****************************************************************************/

/*! \class QJpUnicodeConv
    \reentrant
    \internal
*/

#include "qjpunicode.h"

#include "qbytearray.h"
#ifdef Q_OS_TEMP
#include "qwinfunctions_wce.h"
#endif
#include <stdlib.h>

#define        USE_JISX0212

#define        Q_STRICT

#define        IsLatin(c)        (/*((c) >= 0x00) &&*/ ((c) <= 0x7f))
#define        IsKana(c)        (((c) >= 0xa1) && ((c) <= 0xdf))
#define        IsJisChar(c)        (((c) >= 0x21) && ((c) <= 0x7e))
#define        IsSjisChar1(c)        ((((c) >= 0x81) && ((c) <= 0x9f)) ||        \
                         (((c) >= 0xe0) && ((c) <= 0xfc)))
#define        IsSjisUDC1(c)        (((c) >= 0xf0) && ((c) <= 0xfc))
#define        IsSjisChar2(c)        (((c) >= 0x40) && ((c) != 0x7f) && ((c) <= 0xfc))

#define        IsSjisIBMVDCChar1(c)        (((c) >= 0xfa) && ((c) <= 0xfc))

static uint jisx0208ToSjis(uint h, uint l);
static
inline uint jisx0208ToSjis(uint jis)
{
    return jisx0208ToSjis((jis & 0xff00) >> 8, (jis & 0x00ff));
}

static uint sjisToJisx0208(uint h, uint l);
#if 0
static
inline uint sjisToJisx0208(uint sjis)
{
    return sjisToJisx0208((sjis & 0xff00) >> 8, (sjis & 0x00ff));
}
#endif

static uint jisx0201ToUnicode11(uint h, uint l);
static uint jisx0208ToUnicode11(uint h, uint l);
static uint jisx0212ToUnicode11(uint h, uint l);

static uint unicode11ToJisx0201(uint h, uint l);
static uint unicode11ToJisx0208(uint h, uint l);
static uint unicode11ToJisx0212(uint h, uint l);

/*
 * Unicode 1.1 conversion.
 */

/*!
  \fn QJpUnicodeConv::QJpUnicodeConv (int r)
  \internal
*/


/*! \internal */
uint QJpUnicodeConv::asciiToUnicode(uint h, uint l) const
{
    if ((h == 0) && (l < 0x80)) {
        return l;
    }
    return 0x0000;
}

/*! \internal */
uint QJpUnicodeConv::jisx0201ToUnicode(uint h, uint l) const
{
    if (h == 0) {
        if (IsLatin(l)) {
            return jisx0201LatinToUnicode(h, l);
        } else if (IsKana(l)) {
            return jisx0201KanaToUnicode(h, l);
        }
    }
    return 0x0000;
}

/*! \internal */
uint QJpUnicodeConv::jisx0201LatinToUnicode(uint h, uint l) const
{
    if ((h == 0) && IsLatin(l)) {
        return jisx0201ToUnicode11(h, l);
    }
    return 0x0000;
}

/*! \internal */
uint QJpUnicodeConv::jisx0201KanaToUnicode(uint h, uint l) const
{
    if ((h == 0) && IsKana(l)) {
        return jisx0201ToUnicode11(h, l);
    }
    return 0x0000;
}

/*! \internal */
uint QJpUnicodeConv::jisx0208ToUnicode(uint h, uint l) const
{
    if (rule & UDC){
        if ((0x75 <= h) && (h <= 0x7e) && IsJisChar(l)/*0x21 - 0x7e*/) {
            // User Defined Char (mapped to Private Use Area)
            return 0xe000 + (h - 0x75) * 0x5e + (l - 0x21);
        }
    }
    if ((rule & NEC_VDC) == 0) {
        if ((h == 0x2d) && (IsJisChar(l)/*0x21 - 0x7c*/)) {
            // NEC Vendor Defined Char
            return 0x0000;
        }
    }
    return jisx0208ToUnicode11(h, l);
}

/*! \internal */
uint QJpUnicodeConv::jisx0212ToUnicode(uint h, uint l) const
{
    if (rule & UDC){
        if ((0x75 <= h) && (h <= 0x7e) && IsJisChar(l)/*0x21 - 0x7e*/) {
            // User Defined Char (mapped to Private Use Area)
            return 0xe3ac + (h - 0x75) * 0x5e + (l - 0x21);
        }
    }
    if ((rule & IBM_VDC) == 0){
        if (((h == 0x73) && (0x73 <= l) && (l <= 0x7e)) ||
            ((h == 0x74) && (IsJisChar(l)/*0x21 - 0x7e*/))) {
            // IBM Vendor Defined Char
            return 0x0000;
        }
    }
    return jisx0212ToUnicode11(h, l);
}

/*! \internal */
uint QJpUnicodeConv::unicodeToAscii(uint h, uint l) const
{
    if ((h == 0) && (l < 0x80)) {
        return l;
    }
    return 0x0000;
}

/*! \internal */
uint QJpUnicodeConv::unicodeToJisx0201(uint h, uint l) const
{
    uint jis;
    if ((jis = unicodeToJisx0201Latin(h, l)) != 0) {
        return jis;
    } else if ((jis = unicodeToJisx0201Kana(h, l)) != 0) {
        return jis;
    }
    return 0x0000;
}

/*! \internal */
uint QJpUnicodeConv::unicodeToJisx0201Latin(uint h, uint l) const
{
    uint jis = unicode11ToJisx0201(h, l);
    if (IsLatin(jis)) {
        return jis;
    }
    return 0x0000;
}

/*! \internal */
uint QJpUnicodeConv::unicodeToJisx0201Kana(uint h, uint l) const
{
    uint jis = unicode11ToJisx0201(h, l);
    if (IsKana(jis)) {
        return jis;
    }
    return 0x0000;
}

/*! \internal */
uint QJpUnicodeConv::unicodeToJisx0208(uint h, uint l) const
{
    if (rule & UDC){
        uint unicode = (h << 8) | l;
        if ((0xe000 <= unicode) && (unicode <= 0xe3ab)) {
            // User Defined Char (mapped to Private Use Area)
            unicode -= 0xe000;
            return ((0x75 + unicode / 0x5e) << 8) | (0x21 + unicode % 0x5e);
        }
    }
    uint jis = unicode11ToJisx0208(h, l);
    if ((rule & NEC_VDC) == 0) {
        if ((0x2d21 <= jis) && (jis <= 0x2d7c)) {
            // NEC Vendor Defined Char
            return 0x0000;
        }
    }
    return jis;
}

/*! \internal */
uint QJpUnicodeConv::unicodeToJisx0212(uint h, uint l) const
{
    if (rule & UDC){
        uint unicode = (h << 8) | l;
        if ((0xe3ac <= unicode) && (unicode <= 0xe757)) {
            // User Defined Char (mapped to Private Use Area)
            unicode -= 0xe3ac;
            return ((0x75 + unicode / 0x5e) << 8) | (0x21 + unicode % 0x5e);
        }
    }
    uint jis = unicode11ToJisx0212(h, l);
    if ((rule & IBM_VDC) == 0){
        if (((0x7373 <= jis) && (jis <= 0x737e)) ||
            ((0x7421 <= jis) && (jis <= 0x747e))) {
            // IBM Vendor Defined Char
            return 0x0000;
        }
    }
    return jis;
}

/*! \internal */
uint QJpUnicodeConv::sjisToUnicode(uint h, uint l) const
{
    if (h == 0) {
        return jisx0201ToUnicode(h, l);
    } else if (IsSjisChar1(h) && IsSjisChar2(l)) {
        return jisx0208ToUnicode(sjisToJisx0208(h, l));
    }
    return 0x0000;
}

/*! \internal */
uint QJpUnicodeConv::unicodeToSjis(uint h, uint l) const
{
    uint jis;
    if ((jis = unicodeToJisx0201(h, l)) != 0x0000) {
        return jis;
    } else if ((jis = unicodeToJisx0208(h, l)) != 0x0000) {
        return jisx0208ToSjis(jis);
    } else if ((jis = unicodeToJisx0212(h, l)) != 0x0000) {
        return 0x0000;
    }
    return 0x0000;
}

/*
 * Unicode 1.1 with JISX0201 conversion.
 */

class QJpUnicodeConv_Unicode_JISX0201 : public QJpUnicodeConv {
public:
    QJpUnicodeConv_Unicode_JISX0201(int r) : QJpUnicodeConv(r) {}

//    uint AsciiToUnicode(uint h, uint l) const;
//    uint Jisx0201ToUnicode(uint h, uint l) const;