diff options
Diffstat (limited to 'src/corelib/text/qchar.cpp')
| -rw-r--r-- | src/corelib/text/qchar.cpp | 120 |
1 files changed, 37 insertions, 83 deletions
diff --git a/src/corelib/text/qchar.cpp b/src/corelib/text/qchar.cpp index 32e293ee9f..7b114e9723 100644 --- a/src/corelib/text/qchar.cpp +++ b/src/corelib/text/qchar.cpp @@ -1,14 +1,6 @@ // Copyright (C) 2022 The Qt Company Ltd. // SPDX-License-Identifier: LicenseRef-Qt-Commercial OR LGPL-3.0-only OR GPL-2.0-only OR GPL-3.0-only -// Don't define it while compiling this module, or USERS of Qt will -// not be able to link. -#ifdef QT_NO_CAST_FROM_ASCII -# undef QT_NO_CAST_FROM_ASCII -#endif -#ifdef QT_NO_CAST_TO_ASCII -# undef QT_NO_CAST_TO_ASCII -#endif #include "qchar.h" #include "qdatastream.h" @@ -63,6 +55,13 @@ QT_BEGIN_NAMESPACE \ingroup string-processing \reentrant + \compares strong + \compareswith strong char16_t QString QStringView QLatin1StringView QUtf8StringView + \endcompareswith + \compareswith strong {const char *} QByteArray QByteArrayView + The contents of the byte array is interpreted as utf-8. + \endcompareswith + In Qt, Unicode characters are 16-bit entities without any markup or structure. This class represents such an entity. It is lightweight, so it can be used everywhere. Most compilers treat @@ -124,9 +123,7 @@ QT_BEGIN_NAMESPACE Starting with Qt 6.0, most QChar constructors are \c explicit. This is done to avoid dangerous mistakes when accidentally mixing - integral types and strings. You can opt-out (and make these - constructors implicit) by defining the macro \c - QT_IMPLICIT_QCHAR_CONSTRUCTION. + integral types and strings. For more information see \l{https://www.unicode.org/ucd/}{"About the Unicode Character Database"}. @@ -164,6 +161,8 @@ QT_BEGIN_NAMESPACE \value [since 5.15] Unicode_12_1 Version 12.1 \value [since 5.15] Unicode_13_0 Version 13.0 \value [since 6.3] Unicode_14_0 Version 14.0 + \value [since 6.5] Unicode_15_0 Version 15.0 + \value [since 6.8] Unicode_15_1 Version 15.1 \value Unicode_Unassigned The value is not assigned to any character in version 8.0 of Unicode. @@ -323,6 +322,7 @@ QT_BEGIN_NAMESPACE \value Script_Kaithi \value Script_Kannada \value Script_Katakana + \value [since 6.5] Script_Kawi \value Script_KayahLi \value Script_Kharoshthi \value [since 5.15] Script_KhitanSmallScript @@ -357,6 +357,7 @@ QT_BEGIN_NAMESPACE \value [since 5.6] Script_Multani \value Script_Myanmar \value [since 5.5] Script_Nabataean + \value [since 6.3] Script_NagMundari \value [since 5.15] Script_Nandinagari \value [since 5.11] Script_Newa \value Script_NewTaiLue @@ -1365,7 +1366,7 @@ static const QChar * QT_FASTCALL decompositionHelper( { if (ucs4 >= Hangul_SBase && ucs4 < Hangul_SBase + Hangul_SCount) { // compute Hangul syllable decomposition as per UAX #15 - const uint SIndex = ucs4 - Hangul_SBase; + const char32_t SIndex = ucs4 - Hangul_SBase; buffer[0] = QChar(Hangul_LBase + SIndex / Hangul_NCount); // L buffer[1] = QChar(Hangul_VBase + (SIndex % Hangul_NCount) / Hangul_TCount); // V buffer[2] = QChar(Hangul_TBase + SIndex % Hangul_TCount); // T @@ -1744,42 +1745,42 @@ QDataStream &operator>>(QDataStream &in, QChar &chr) *****************************************************************************/ /*! - \fn bool QChar::operator==(QChar c1, QChar c2) + \fn bool QChar::operator==(const QChar &c1, const QChar &c2) Returns \c true if \a c1 and \a c2 are the same Unicode character; otherwise returns \c false. */ /*! - \fn int QChar::operator!=(QChar c1, QChar c2) + \fn bool QChar::operator!=(const QChar &c1, const QChar &c2) Returns \c true if \a c1 and \a c2 are not the same Unicode character; otherwise returns \c false. */ /*! - \fn int QChar::operator<=(QChar c1, QChar c2) + \fn bool QChar::operator<=(const QChar &c1, const QChar &c2) Returns \c true if the numeric Unicode value of \a c1 is less than or equal to that of \a c2; otherwise returns \c false. */ /*! - \fn int QChar::operator>=(QChar c1, QChar c2) + \fn bool QChar::operator>=(const QChar &c1, const QChar &c2) Returns \c true if the numeric Unicode value of \a c1 is greater than or equal to that of \a c2; otherwise returns \c false. */ /*! - \fn int QChar::operator<(QChar c1, QChar c2) + \fn bool QChar::operator<(const QChar &c1, const QChar &c2) Returns \c true if the numeric Unicode value of \a c1 is less than that of \a c2; otherwise returns \c false. */ /*! - \fn int QChar::operator>(QChar c1, QChar c2) + \fn bool QChar::operator>(const QChar &c1, const QChar &c2) Returns \c true if the numeric Unicode value of \a c1 is greater than that of \a c2; otherwise returns \c false. @@ -1815,7 +1816,7 @@ static void decomposeHelper(QString *str, bool canonical, QChar::UnicodeVersion QString &s = *str; const unsigned short *utf16 = reinterpret_cast<unsigned short *>(s.data()); - const unsigned short *uc = utf16 + s.length(); + const unsigned short *uc = utf16 + s.size(); while (uc != utf16 + from) { char32_t ucs4 = *(--uc); if (QChar(ucs4).isLowSurrogate() && uc != utf16) { @@ -1861,26 +1862,26 @@ struct UCS2SurrogatePair { inline bool operator<(const UCS2SurrogatePair &ligature1, const UCS2SurrogatePair &ligature2) { return QChar::surrogateToUcs4(ligature1.p1.u1, ligature1.p1.u2) < QChar::surrogateToUcs4(ligature2.p1.u1, ligature2.p1.u2); } -inline bool operator<(uint u1, const UCS2SurrogatePair &ligature) +inline bool operator<(char32_t u1, const UCS2SurrogatePair &ligature) { return u1 < QChar::surrogateToUcs4(ligature.p1.u1, ligature.p1.u2); } -inline bool operator<(const UCS2SurrogatePair &ligature, uint u1) +inline bool operator<(const UCS2SurrogatePair &ligature, char32_t u1) { return QChar::surrogateToUcs4(ligature.p1.u1, ligature.p1.u2) < u1; } -static uint inline ligatureHelper(uint u1, uint u2) +static char32_t inline ligatureHelper(char32_t u1, char32_t u2) { if (u1 >= Hangul_LBase && u1 < Hangul_SBase + Hangul_SCount) { // compute Hangul syllable composition as per UAX #15 // hangul L-V pair - const uint LIndex = u1 - Hangul_LBase; + const char32_t LIndex = u1 - Hangul_LBase; if (LIndex < Hangul_LCount) { - const uint VIndex = u2 - Hangul_VBase; + const char32_t VIndex = u2 - Hangul_VBase; if (VIndex < Hangul_VCount) return Hangul_SBase + (LIndex * Hangul_VCount + VIndex) * Hangul_TCount; } // hangul LV-T pair - const uint SIndex = u1 - Hangul_SBase; + const char32_t SIndex = u1 - Hangul_SBase; if (SIndex < Hangul_SCount && (SIndex % Hangul_TCount) == 0) { - const uint TIndex = u2 - Hangul_TBase; + const char32_t TIndex = u2 - Hangul_TBase; if (TIndex < Hangul_TCount && TIndex) return u1 + TIndex; } @@ -1910,19 +1911,19 @@ static void composeHelper(QString *str, QChar::UnicodeVersion version, qsizetype { QString &s = *str; - if (from < 0 || s.length() - from < 2) + if (from < 0 || s.size() - from < 2) return; - uint stcode = 0; // starter code point + char32_t stcode = 0; // starter code point qsizetype starter = -1; // starter position qsizetype next = -1; // to prevent i == next int lastCombining = 255; // to prevent combining > lastCombining qsizetype pos = from; - while (pos < s.length()) { + while (pos < s.size()) { qsizetype i = pos; char32_t uc = s.at(pos).unicode(); - if (QChar(uc).isHighSurrogate() && pos < s.length()-1) { + if (QChar(uc).isHighSurrogate() && pos < s.size()-1) { ushort low = s.at(pos+1).unicode(); if (QChar(low).isLowSurrogate()) { uc = QChar::surrogateToUcs4(uc, low); @@ -1942,7 +1943,7 @@ static void composeHelper(QString *str, QChar::UnicodeVersion version, qsizetype int combining = p->combiningClass; if ((i == next || combining > lastCombining) && starter >= from) { // allowed to form ligature with S - uint ligature = ligatureHelper(stcode, uc); + char32_t ligature = ligatureHelper(stcode, uc); if (ligature) { stcode = ligature; QChar *d = s.data(); @@ -1969,7 +1970,7 @@ static void composeHelper(QString *str, QChar::UnicodeVersion version, qsizetype static void canonicalOrderHelper(QString *str, QChar::UnicodeVersion version, qsizetype from) { QString &s = *str; - const qsizetype l = s.length()-1; + const qsizetype l = s.size()-1; char32_t u1, u2; char16_t c1, c2; @@ -2057,8 +2058,8 @@ static bool normalizationQuickCheckHelper(QString *str, QString::NormalizationFo enum { NFQC_YES = 0, NFQC_NO = 1, NFQC_MAYBE = 3 }; - const ushort *string = reinterpret_cast<const ushort *>(str->constData()); - qsizetype length = str->length(); + const auto *string = reinterpret_cast<const char16_t *>(str->constData()); + qsizetype length = str->size(); // this avoids one out of bounds check in the loop while (length > from && QChar::isHighSurrogate(string[length - 1])) @@ -2101,57 +2102,10 @@ static bool normalizationQuickCheckHelper(QString *str, QString::NormalizationFo *lastStable = pos; } - if (length != str->length()) // low surrogate parts at the end of text - *lastStable = str->length() - 1; + if (length != str->size()) // low surrogate parts at the end of text + *lastStable = str->size() - 1; return true; } -/*! - \macro QT_IMPLICIT_QCHAR_CONSTRUCTION - \since 6.0 - \relates QChar - - Defining this macro makes certain QChar constructors implicit - rather than explicit. This is done to enforce safe conversions: - - \badcode - - QString str = getString(); - if (str == 123) { - // Oops, meant str == "123". By default does not compile, - // *unless* this macro is defined, in which case, it's interpreted - // as `if (str == QChar(123))`, that is, `if (str == '{')`. - // Likely, not what we meant. - } - - \endcode - - This macro is provided to keep existing code working; it is - recommended to instead use explicit conversions and/or QLatin1Char. - For instance: - - \code - - QChar c1 = 'x'; // OK, unless QT_NO_CAST_FROM_ASCII is defined - QChar c2 = u'x'; // always OK, recommended - QChar c3 = QLatin1Char('x'); // always OK, recommended - - // from int to 1 UTF-16 code unit: must guarantee that the input is <= 0xFFFF - QChar c4 = 120; // compile error, unless QT_IMPLICIT_QCHAR_CONSTRUCTION is defined - QChar c5(120); // OK (direct initialization) - auto c6 = QChar(120); // ditto - - // from int/char32_t to 1/2 UTF-16 code units: - // 𝄞 'MUSICAL SYMBOL G CLEF' (U+1D11E) - auto c7 = QChar(0x1D11E); // compiles, but undefined behavior at runtime - auto c8 = QChar::fromUcs4(0x1D11E); // always OK - auto c9 = QChar::fromUcs4(U'\U0001D11E'); // always OK - // => use c8/c9 as QStringView objects - - \endcode - - \sa QLatin1Char, QChar::fromUcs4, QT_NO_CAST_FROM_ASCII -*/ - QT_END_NAMESPACE |