qchar.cpp

奇趣公司比较新的qt/emd版本

}

/*!
    Returns the uppercase equivalent if the character is lowercase or titlecase;
    otherwise returns the character itself.
*/
QChar QChar::toUpper() const
{
    const QUnicodeTables::Properties *p = qGetProp(ucs);
    if (!p->upperCaseSpecial)
        return ucs + p->upperCaseDiff;
    return ucs;
}

/*! \overload
Returns the uppercase equivalent of the UCS-4-encoded character specified
by \a ucs4 if the character is lowercase or titlecase; otherwise returns
the character itself.
 */
uint QChar::toUpper(uint ucs4)
{
    if (ucs4 > LAST_UNICODE_CHAR)
        return ucs4;
    const QUnicodeTables::Properties *p = qGetProp(ucs4);
    if (!p->upperCaseSpecial)
        return ucs4 + p->upperCaseDiff;
    return ucs4;
}

/*! \overload
Returns the uppercase equivalent of the UCS-2-encoded character specified
by \a ucs2 if the character is lowercase or titlecase; otherwise returns
the character itself.
 */
ushort QChar::toUpper(ushort ucs2)
{
    const QUnicodeTables::Properties *p = qGetProp(ucs2);
    if (!p->upperCaseSpecial)
        return ucs2 + p->upperCaseDiff;
    return ucs2;
}

/*!
    Returns the title case equivalent if the character is lowercase or uppercase;
    otherwise returns the character itself.
*/
QChar QChar::toTitleCase() const
{
    const QUnicodeTables::Properties *p = qGetProp(ucs);
    if (!p->titleCaseSpecial)
        return ucs + p->titleCaseDiff;
    return ucs;
}

/*!
    \overload
    Returns the title case equivalent of the UCS-4-encoded character specified
    by \a ucs4 if the character is lowercase or uppercase; otherwise returns
    the character itself.
*/
uint QChar::toTitleCase(uint ucs4)
{
    if (ucs4 > LAST_UNICODE_CHAR)
        return ucs4;
    const QUnicodeTables::Properties *p = qGetProp(ucs4);
    if (!p->titleCaseSpecial)
        return ucs4 + p->titleCaseDiff;
    return ucs4;
}

/*!
    \overload
    Returns the title case equivalent of the UCS-2-encoded character specified
    by \a ucs2 if the character is lowercase or uppercase; otherwise returns
    the character itself.
*/
ushort QChar::toTitleCase(ushort ucs2)
{
    const QUnicodeTables::Properties *p = qGetProp(ucs2);
    if (!p->titleCaseSpecial)
        return ucs2 + p->titleCaseDiff;
    return ucs2;
}


static inline uint foldCase(const ushort *ch, const ushort *start)
{
    uint c = *ch;
    if (QChar(c).isLowSurrogate() && ch > start && QChar(*(ch - 1)).isHighSurrogate())
        c = QChar::surrogateToUcs4(*(ch - 1), c);
    return *ch + qGetProp(c)->caseFoldDiff;
}

static inline uint foldCase(uint ch, uint &last)
{
    uint c = ch;
    if (QChar(c).isLowSurrogate() && QChar(last).isHighSurrogate())
        c = QChar::surrogateToUcs4(last, c);
    last = ch;
    return ch + qGetProp(c)->caseFoldDiff;
}

static inline ushort foldCase(ushort ch)
{
    return ch + qGetProp(ch)->caseFoldDiff;
}

/*!
    Returns the case folded equivalent of the character. For most Unicode characters this
    is the same as toLowerCase().
*/
QChar QChar::toCaseFolded() const
{
    return ucs + qGetProp(ucs)->caseFoldDiff;
}

/*!
    \overload
    Returns the case folded equivalent of the UCS-4-encoded character specified
    by \a ucs4. For most Unicode characters this is the same as toLowerCase().
*/
uint QChar::toCaseFolded(uint ucs4)
{
    if (ucs4 > LAST_UNICODE_CHAR)
        return ucs4;
    return ucs4 + qGetProp(ucs4)->caseFoldDiff;
}

/*!
    \overload
    Returns the case folded equivalent of the UCS-2-encoded character specified
    by \a ucs2. For most Unicode characters this is the same as toLowerCase().
*/
ushort QChar::toCaseFolded(ushort ucs2)
{
    return ucs2 + qGetProp(ucs2)->caseFoldDiff;
}


/*!
    \fn char QChar::latin1() const

    Use toLatin1() instead.
*/

/*!
    \fn char QChar::ascii() const

    Use toAscii() instead.
*/

/*!
    \fn char QChar::toLatin1() const

    Returns the Latin-1 character equivalent to the QChar, or 0. This
    is mainly useful for non-internationalized software.

    \sa toAscii(), unicode(), QTextCodec::codecForCStrings()
*/

/*!
    Returns the character value of the QChar obtained using the current
    codec used to read C strings, or 0 if the character is not representable
    using this codec. The default codec handles Latin-1 encoded text,
    but this can be changed to assist developers writing source code using
    other encodings.

    The main purpose of this function is to preserve ASCII characters used
    in C strings. This is mainly useful for developers of non-internationalized
    software.

    \sa toLatin1(), unicode(), QTextCodec::codecForCStrings()
*/
const char QChar::toAscii() const
{
#ifndef QT_NO_CODEC_FOR_C_STRINGS
    if (QTextCodec::codecForCStrings())
        // #####
        return QTextCodec::codecForCStrings()->fromUnicode(QString(*this)).at(0);
#endif
    return ucs > 0xff ? 0 : char(ucs);
}

/*!
    \fn QChar QChar::fromLatin1(char c)

    Converts the Latin-1 character \a c to its equivalent QChar. This
    is mainly useful for non-internationalized software.

    \sa fromAscii(), unicode(), QTextCodec::codecForCStrings()
*/

/*!
    Converts the ASCII character \a c to its equivalent QChar. This
    is mainly useful for non-internationalized software.

    An alternative is to use QLatin1Char.

    \sa fromLatin1(), unicode(), QTextCodec::codecForCStrings()
*/
QChar QChar::fromAscii(char c)
{
#ifndef QT_NO_CODEC_FOR_C_STRINGS
    if (QTextCodec::codecForCStrings())
        // #####
        return QTextCodec::codecForCStrings()->toUnicode(&c, 1).at(0).unicode();
#endif
    return QChar(ushort((uchar)c));
}

#ifndef QT_NO_DATASTREAM
/*!
  \relates QChar

  Writes the char \a chr to the stream \a out.

  \sa {Format of the QDataStream operators}
 */

QDataStream &operator<<(QDataStream &out, const QChar &chr)
{
    out << quint16(chr.unicode());
    return out;
}


/*!
  \relates QChar

  Reads a char from the stream \a in into char \a chr.

  \sa {Format of the QDataStream operators}
 */

QDataStream &operator>>(QDataStream &in, QChar &chr)
{
    quint16 u;
    in >> u;
    chr.unicode() = ushort(u);
    return in;
}
#endif

/*!
    \fn ushort & QChar::unicode()

    Returns a reference to the numeric Unicode value of the QChar.
*/

/*!
    \fn ushort QChar::unicode() const

    \overload
*/

/*****************************************************************************
  Documentation of QChar related functions
 *****************************************************************************/

/*!
    \fn bool operator==(QChar c1, QChar c2)

    \relates QChar

    Returns true if \a c1 and \a c2 are the same Unicode character;
    otherwise returns false.
*/

/*!
    \fn int operator!=(QChar c1, QChar c2)

    \relates QChar

    Returns true if \a c1 and \a c2 are not the same Unicode
    character; otherwise returns false.
*/

/*!
    \fn int operator<=(QChar c1, QChar c2)

    \relates QChar

    Returns true if the numeric Unicode value of \a c1 is less than
    or equal to that of \a c2; otherwise returns false.
*/

/*!
    \fn int operator>=(QChar c1, QChar c2)

    \relates QChar

    Returns true if the numeric Unicode value of \a c1 is greater than
    or equal to that of \a c2; otherwise returns false.
*/

/*!
    \fn int operator<(QChar c1, QChar c2)

    \relates QChar

    Returns true if the numeric Unicode value of \a c1 is less than
    that of \a c2; otherwise returns false.
*/

/*!
    \fn int operator>(QChar c1, QChar c2)

    \relates QChar

    Returns true if the numeric Unicode value of \a c1 is greater than
    that of \a c2; otherwise returns false.
*/

/*!
    \fn bool QChar::mirrored() const

    Use hasMirrored() instead.
*/

/*!
    \fn QChar QChar::lower() const

    Use toLower() instead.
*/

/*!
    \fn QChar QChar::upper() const

    Use toUpper() instead.
*/

/*!
    \fn bool QChar::networkOrdered()

    See if QSysInfo::ByteOrder == QSysInfo::BigEndian instead.
*/


// ---------------------------------------------------------------------------


static QString decompose(const QString &str, bool canonical, QChar::UnicodeVersion version)
{
    unsigned short buffer[3];

    QString s = str;

    const unsigned short *utf16 = s.utf16();
    const unsigned short *uc = utf16 + s.length();
    while (uc != utf16) {
        uint ucs4 = *(--uc);
        if (QChar(ucs4).isLowSurrogate() && uc != utf16) {
            ushort high = *(uc - 1);
            if (QChar(high).isHighSurrogate()) {
                --uc;
                ucs4 = QChar::surrogateToUcs4(high, ucs4);
            }
        }
        if (QChar::unicodeVersion(ucs4) > version)
            continue;
        int length;
        int tag;
        const unsigned short *d = decomposition(ucs4, &length, &tag, buffer);
        if (!d || (canonical && tag != QChar::Canonical))
            continue;

        s.replace(uc - utf16, ucs4 > 0x10000 ? 2 : 1, (const QChar *)d, length);
        // since the insert invalidates the pointers and we do decomposition recursive
        int pos = uc - utf16;
        utf16 = s.utf16();
        uc = utf16 + pos + length;
    }

    return s;
}


static ushort ligature(ushort u1, ushort u2)
{
    // hangul L-V pair
    int LIndex = u1 - Hangul_LBase;
    if (0 <= LIndex && LIndex < Hangul_LCount) {
        int VIndex = u2 - Hangul_VBase;
        if (0 <= VIndex && VIndex < Hangul_VCount)
            return Hangul_SBase + (LIndex * Hangul_VCount + VIndex) * Hangul_TCount;
    }

    // hangul LV-T pair
    int SIndex = u1 - Hangul_SBase;
    if (0 <= SIndex && SIndex < Hangul_SCount && (SIndex % Hangul_TCount) == 0) {
        int TIndex = u2 - Hangul_TBase;
        if (0 <= TIndex && TIndex <= Hangul_TCount)
            return u1 + TIndex;
    }

    const unsigned short index = GET_LIGATURE_INDEX(u2);
    if (index == 0xffff)
        return 0;
    const unsigned short *ligatures = uc_ligature_map+index;
    ushort length = *ligatures;
    ++ligatures;
    // ### use bsearch
    for (uint i = 0; i < length; ++i)
        if (ligatures[2*i] == u1)
            return ligatures[2*i+1];
    return 0;
}

static QString compose(const QString &str)
{
    QString s = str;

    if (s.length() < 2)
        return s;

    // the loop can partly ignore high Unicode as all ligatures are in the BMP
    int starter = 0;
    int lastCombining = 0;
    int pos = 0;
    while (pos < s.length()) {
        uint uc = s.utf16()[pos];
        if (QChar(uc).isHighSurrogate() && pos < s.length()-1) {
            ushort low = s.utf16()[pos+1];
            if (QChar(low).isLowSurrogate()) {
                uc = QChar::surrogateToUcs4(uc, low);
                ++pos;
            }
        }
        int combining = QChar::combiningClass(uc);
        if (starter == pos - 1 || combining > lastCombining) {
            // allowed to form ligature with S
            QChar ligature = ::ligature(s.utf16()[starter], uc);
            if (ligature.unicode()) {
                s[starter] = ligature;
                s.remove(pos, 1);
                continue;
            }
        }
        if (!combining)
            starter = pos;
        lastCombining = combining;
        ++pos;
    }
    return s;
}


static QString canonicalOrder(const QString &str, QChar::UnicodeVersion version)
{
    QString s = str;
    const int l = s.length()-1;
    int pos = 0;
    while (pos < l) {
        int p2 = pos+1;
        uint u1 = s.at(pos).unicode();
        if (QChar(u1).isHighSurrogate()) {
            ushort low = s.at(pos+1).unicode();
            if (QChar(low).isLowSurrogate()) {
                p2++;
                u1 = QChar::surrogateToUcs4(u1, low);
                if (p2 >= l)
                    break;
            }
        }
        uint u2 = s.at(p2).unicode();
        if (QChar(u2).isHighSurrogate() && p2 < l-1) {
            ushort low = s.at(p2+1).unicode();
            if (QChar(low).isLowSurrogate()) {
                p2++;
                u2 = QChar::surrogateToUcs4(u2, low);
            }
        }

        int c2 = QChar::combiningClass(u2);
        if (QChar::unicodeVersion(u2) > version)
            c2 = 0;

        if (c2 == 0) {
            pos = p2+1;
            continue;
        }
        int c1 = QChar::combiningClass(u1);
        if (QChar::unicodeVersion(u1) > version)
            c1 = 0;

        if (c1 > c2) {
            QChar *uc = s.data();
            int p = pos;
            // exchange characters
            if (u2 < 0x10000) {
                uc[p++] = u2;
            } else {
                uc[p++] = QChar::highSurrogate(u2);
                uc[p++] = QChar::lowSurrogate(u2);
            }
            if (u1 < 0x10000) {
                uc[p++] = u1;
            } else {
                uc[p++] = QChar::highSurrogate(u1);
                uc[p++] = QChar::lowSurrogate(u1);
            }
            if (pos > 0)
                --pos;
            if (pos > 0 && s.at(pos).isLowSurrogate())
                --pos;
        } else {
            ++pos;
            if (u1 > 0x10000)
                ++pos;
        }
    }
    return s;
}

int QUnicodeTables::script(unsigned int uc)
{
    if (uc > 0xffff)
        return Common;
    int script = uc_scripts[uc >> 7];
    if (script < ScriptSentinel)
        return script;
    script = (((script - ScriptSentinel) * UnicodeBlockSize) + UnicodeBlockCount);
    script = uc_scripts[script + (uc & 0x7f)];
    return script;
}


Q_CORE_EXPORT QUnicodeTables::LineBreakClass QUnicodeTables::lineBreakClass(uint ucs4)
{
    return (QUnicodeTables::LineBreakClass) qGetProp(ucs4)->line_break_class;
}