/**************************************************************************** ** ** Copyright (C) 2009 Nokia Corporation and/or its subsidiary(-ies). ** Contact: Nokia Corporation (qt-info@nokia.com) ** ** This file is part of the QtCore module of the Qt Toolkit. ** ** $QT_BEGIN_LICENSE:LGPL$ ** No Commercial Usage ** This file contains pre-release code and may not be distributed. ** You may use this file in accordance with the terms and conditions ** contained in the either Technology Preview License Agreement or the ** Beta Release License Agreement. ** ** GNU Lesser General Public License Usage ** Alternatively, this file may be used under the terms of the GNU Lesser ** General Public License version 2.1 as published by the Free Software ** Foundation and appearing in the file LICENSE.LGPL included in the ** packaging of this file. Please review the following information to ** ensure the GNU Lesser General Public License version 2.1 requirements ** will be met: http://www.gnu.org/licenses/old-licenses/lgpl-2.1.html. ** ** In addition, as a special exception, Nokia gives you certain ** additional rights. These rights are described in the Nokia Qt LGPL ** Exception version 1.0, included in the file LGPL_EXCEPTION.txt in this ** package. ** ** GNU General Public License Usage ** Alternatively, this file may be used under the terms of the GNU ** General Public License version 3.0 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 the GNU General Public License version 3.0 requirements will be ** met: http://www.gnu.org/copyleft/gpl.html. ** ** If you are unsure which license is appropriate for your use, please ** contact the sales department at http://www.qtsoftware.com/contact. ** $QT_END_LICENSE$ ** ****************************************************************************/ #include "qglobal.h" #ifndef QT_NO_SYSTEMLOCALE #define QLOCALE_CPP QT_BEGIN_NAMESPACE class QSystemLocale; static QSystemLocale *QSystemLocale_globalSystemLocale(); QT_END_NAMESPACE #endif #include "qplatformdefs.h" #include "qdatastream.h" #include "qstring.h" #include "qlocale.h" #include "qlocale_p.h" #include "qdatetime_p.h" #include "qnamespace.h" #include "qdatetime.h" #include "qstringlist.h" #include "qvariant.h" #if defined(Q_WS_WIN) # include "qt_windows.h" # include #endif #if !defined(QWS) && defined(Q_OS_MAC) # include "private/qcore_mac_p.h" # include #endif #include "private/qnumeric_p.h" #include #include #include #include #include #include #include #if defined(Q_OS_LINUX) && !defined(__UCLIBC__) # include #endif #if !defined(QT_QLOCALE_NEEDS_VOLATILE) # if defined(Q_CC_GNU) # if __GNUC__ == 4 # define QT_QLOCALE_NEEDS_VOLATILE # elif defined(Q_OS_WIN) # define QT_QLOCALE_NEEDS_VOLATILE # endif # endif #endif #if defined(QT_QLOCALE_NEEDS_VOLATILE) # define NEEDS_VOLATILE volatile #else # define NEEDS_VOLATILE #endif // Sizes as defined by the ISO C99 standard - fallback #ifndef LLONG_MAX # define LLONG_MAX Q_INT64_C(0x7fffffffffffffff) #endif #ifndef LLONG_MIN # define LLONG_MIN (-LLONG_MAX - Q_INT64_C(1)) #endif #ifndef ULLONG_MAX # define ULLONG_MAX Q_UINT64_C(0xffffffffffffffff) #endif #define CONVERSION_BUFF_SIZE 255 QT_BEGIN_NAMESPACE #ifndef QT_QLOCALE_USES_FCVT static char *_qdtoa( NEEDS_VOLATILE double d, int mode, int ndigits, int *decpt, int *sign, char **rve, char **digits_str); #endif Q_CORE_EXPORT char *qdtoa(double d, int mode, int ndigits, int *decpt, int *sign, char **rve, char **digits_str); Q_CORE_EXPORT double qstrtod(const char *s00, char const **se, bool *ok); Q_CORE_EXPORT qlonglong qstrtoll(const char *nptr, const char **endptr, register int base, bool *ok); static qulonglong qstrtoull(const char *nptr, const char **endptr, register int base, bool *ok); /****************************************************************************** ** Helpers for accessing Qt locale database */ QT_BEGIN_INCLUDE_NAMESPACE #include "qlocale_data_p.h" QT_END_INCLUDE_NAMESPACE QLocale::MeasurementSystem QLocalePrivate::measurementSystem() const { for (int i = 0; i < ImperialMeasurementSystemsCount; ++i) { if (ImperialMeasurementSystems[i].languageId == m_language_id && ImperialMeasurementSystems[i].countryId == m_country_id) { return QLocale::ImperialSystem; } } return QLocale::MetricSystem; } // Assumes that code is a // QChar code[3]; // If the code is two-digit the third digit must be 0 static QLocale::Language codeToLanguage(const QChar *code) { ushort uc1 = code[0].unicode(); ushort uc2 = code[1].unicode(); ushort uc3 = code[2].unicode(); if (uc1 == 'n' && uc2 == 'o' && uc3 == 0) uc2 = 'b'; const unsigned char *c = language_code_list; for (; *c != 0; c += 3) { if (uc1 == c[0] && uc2 == c[1] && uc3 == c[2]) return QLocale::Language((c - language_code_list)/3); } return QLocale::C; } // Assumes that code is a // QChar code[2]; static QLocale::Country codeToCountry(const QChar *code) { ushort uc1 = code[0].unicode(); ushort uc2 = code[1].unicode(); const unsigned char *c = country_code_list; for (; *c != 0; c += 2) { if (uc1 == c[0] && uc2 == c[1]) return QLocale::Country((c - country_code_list)/2); } return QLocale::AnyCountry; } static QString languageToCode(QLocale::Language language) { if (language == QLocale::C) return QLatin1String("C"); const unsigned char *c = language_code_list + 3*(uint(language)); QString code; code.resize(c[2] == 0 ? 2 : 3); code[0] = ushort(c[0]); code[1] = ushort(c[1]); if (c[2] != 0) code[2] = ushort(c[2]); return code; } static QString countryToCode(QLocale::Country country) { if (country == QLocale::AnyCountry) return QString(); QString code; code.resize(2); const unsigned char *c = country_code_list + 2*(uint(country)); code[0] = ushort(c[0]); code[1] = ushort(c[1]); return code; } static const QLocalePrivate *findLocale(QLocale::Language language, QLocale::Country country) { unsigned language_id = language; unsigned country_id = country; uint idx = locale_index[language_id]; const QLocalePrivate *d = locale_data + idx; if (idx == 0) // default language has no associated country return d; if (country == QLocale::AnyCountry) return d; Q_ASSERT(d->languageId() == language_id); while (d->languageId() == language_id && d->countryId() != country_id) ++d; if (d->countryId() == country_id && d->languageId() == language_id) return d; return locale_data + idx; } static bool splitLocaleName(const QString &name, QChar *lang_begin, QChar *cntry_begin) { for (int i = 0; i < 3; ++i) lang_begin[i] = 0; for (int i = 0; i < 2; ++i) cntry_begin[i] = 0; int l = name.length(); QChar *lang = lang_begin; QChar *cntry = cntry_begin; int state = 0; const QChar *uc = name.unicode(); for (int i = 0; i < l; ++i) { if (uc->unicode() == '.' || uc->unicode() == '@') break; switch (state) { case 0: // parsing language if (uc->unicode() == '_') { state = 1; break; } if (lang - lang_begin == 3) return false; if (uc->unicode() < 'a' || uc->unicode() > 'z') return false; *lang = *uc; ++lang; break; case 1: // parsing country if (cntry - cntry_begin == 2) { cntry_begin[0] = 0; break; } *cntry = *uc; ++cntry; break; } ++uc; } int lang_len = lang - lang_begin; return lang_len == 2 || lang_len == 3; } static void getLangAndCountry(const QString &name, QLocale::Language &lang, QLocale::Country &cntry) { lang = QLocale::C; cntry = QLocale::AnyCountry; QChar lang_code[3]; QChar cntry_code[2]; if (!splitLocaleName(name, lang_code, cntry_code)) return; lang = codeToLanguage(lang_code); if (lang == QLocale::C) return; if (cntry_code[0].unicode() != 0) cntry = codeToCountry(cntry_code); } static const QLocalePrivate *findLocale(const QString &name) { QLocale::Language lang; QLocale::Country cntry; getLangAndCountry(name, lang, cntry); return findLocale(lang, cntry); } static QString readEscapedFormatString(const QString &format, int *idx) { int &i = *idx; Q_ASSERT(format.at(i).unicode() == '\''); ++i; if (i == format.size()) return QString(); if (format.at(i).unicode() == '\'') { // "''" outside of a quoted stirng ++i; return QLatin1String("'"); } QString result; while (i < format.size()) { if (format.at(i).unicode() == '\'') { if (i + 1 < format.size() && format.at(i + 1).unicode() == '\'') { // "''" inside of a quoted string result.append(QLatin1Char('\'')); i += 2; } else { break; } } else { result.append(format.at(i++)); } } if (i < format.size()) ++i; return result; } static int repeatCount(const QString &s, int i) { QChar c = s.at(i); int j = i + 1; while (j < s.size() && s.at(j) == c) ++j; return j - i; } static const QLocalePrivate *default_lp = 0; static uint default_number_options = 0; #ifndef QT_NO_SYSTEMLOCALE static QByteArray envVarLocale() { static QByteArray lang = 0; #ifdef Q_OS_UNIX lang = qgetenv("LC_ALL"); if (lang.isNull()) lang = qgetenv("LC_NUMERIC"); if (lang.isNull()) #endif lang = qgetenv("LANG"); return lang; } #if defined(Q_OS_WIN) /****************************************************************************** ** Wrappers for Windows locale system functions */ static const char *winLangCodeToIsoName(int code); static QString winIso639LangName(LCID id = LOCALE_USER_DEFAULT); static QString winIso3116CtryName(LCID id = LOCALE_USER_DEFAULT); static QString getWinLocaleInfo(LCTYPE type) { int cnt = 0; LCID id = GetUserDefaultLCID(); QT_WA({ cnt = GetLocaleInfoW(id, type, 0, 0)*2; } , { cnt = GetLocaleInfoA(id, type, 0, 0); }); if (cnt == 0) { qWarning("QLocale: empty windows locale info (%d)", (int)type); return QString(); } QByteArray buff(cnt, 0); QT_WA({ cnt = GetLocaleInfoW(id, type, reinterpret_cast(buff.data()), buff.size()/2); } , { cnt = GetLocaleInfoA(id, type, buff.data(), buff.size()); }); if (cnt == 0) { qWarning("QLocale: empty windows locale info (%d)", (int)type); return QString(); } QString result; QT_WA({ result = QString::fromUtf16(reinterpret_cast(buff.data())); } , { result = QString::fromLocal8Bit(buff.data()); }); return result; } QByteArray getWinLocaleName(LCID id = LOCALE_USER_DEFAULT) { QByteArray result; if (id == LOCALE_USER_DEFAULT) { result = envVarLocale(); QChar lang[3]; QChar cntry[2]; if ( result == "C" || !result.isEmpty() && splitLocaleName(QString::fromLocal8Bit(result), lang, cntry) ) { long id = 0; bool ok = false; id = qstrtoll(result.data(), 0, 0, &ok); if ( !ok || id == 0 || id < INT_MIN || id > INT_MAX ) return result; else return winLangCodeToIsoName( (int)id ); } } if (QSysInfo::WindowsVersion == QSysInfo::WV_95 || (QSysInfo::WindowsVersion & QSysInfo::WV_CE_based)) { result = winLangCodeToIsoName(id != LOCALE_USER_DEFAULT ? id : GetUserDefaultLCID()); } else { if (id == LOCALE_USER_DEFAULT) id = GetUserDefaultLCID(); QString resultuage = winIso639LangName(id); QString country = winIso3116CtryName(id); result = resultuage.toLatin1(); if (!country.isEmpty()) { result += '_'; result += country.toLatin1(); } } return result; } Q_CORE_EXPORT QLocale qt_localeFromLCID(LCID id) { return QLocale(QString::fromLatin1(getWinLocaleName(id))); } static QString winToQtFormat(const QString &sys_fmt) { QString result; int i = 0; while (i < sys_fmt.size()) { if (sys_fmt.at(i).unicode() == QLatin1Char('\'')) { QString text = readEscapedFormatString(sys_fmt, &i); if (text == QLatin1String("'")) result += QLatin1String("''"); else result += QLatin1Char('\'') + text + QLatin1Char('\''); continue; } QChar c = sys_fmt.at(i); int repeat = repeatCount(sys_fmt, i); switch (c.unicode()) { // Date case 'y': if (repeat > 5) repeat = 5; else if (repeat == 3) repeat = 2; switch (repeat) { case 1: result += QLatin1String("yy"); // "y" unsupported by Qt, use "yy" break; case 5: result += QLatin1String("yyyy"); // "yyyyy" same as "yyyy" on Windows break; default: result += QString(repeat, QLatin1Char('y')); break; } break; case 'g': if (repeat > 2) repeat = 2; switch (repeat) { case 2: break; // no equivalent of "gg" in Qt default: result += QLatin1Char('g'); break; } break; case 't': if (repeat > 2) repeat = 2; result += QLatin1String("AP"); // "t" unsupported, use "AP" break; default: result += QString(repeat, c); break; } i += repeat; } return result; } static QString winDateToString(const QDate &date, DWORD flags) { SYSTEMTIME st; memset(&st, 0, sizeof(SYSTEMTIME)); st.wYear = date.year(); st.wMonth = date.month(); st.wDay = date.day(); LCID id = GetUserDefaultLCID(); QT_WA({ TCHAR buf[255]; if (GetDateFormatW(id, flags, &st, 0, buf, 255)) return QString::fromUtf16((ushort*)buf); } , { char buf[255]; if (GetDateFormatA(id, flags, &st, 0, (char*)&buf, 255)) return QString::fromLocal8Bit(buf); }); return QString(); } static QString winTimeToString(const QTime &time) { SYSTEMTIME st; memset(&st, 0, sizeof(SYSTEMTIME)); st.wHour = time.hour(); st.wMinute = time.minute(); st.wSecond = time.second(); st.wMilliseconds = 0; DWORD flags = 0; LCID id = GetUserDefaultLCID(); QT_WA({ TCHAR buf[255]; if (GetTimeFormatW(id, flags, &st, 0, buf, 255)) return QString::fromUtf16((ushort*)buf); } , { char buf[255]; if (GetTimeFormatA(id, flags, &st, 0, (char*)&buf, 255)) return QString::fromLocal8Bit(buf); }); return QString(); } static QString winDayName(int day, bool short_format) { static const LCTYPE short_day_map[] = { LOCALE_SABBREVDAYNAME1, LOCALE_SABBREVDAYNAME2, LOCALE_SABBREVDAYNAME3, LOCALE_SABBREVDAYNAME4, LOCALE_SABBREVDAYNAME5, LOCALE_SABBREVDAYNAME6, LOCALE_SABBREVDAYNAME7 }; static const LCTYPE long_day_map[] = { LOCALE_SDAYNAME1, LOCALE_SDAYNAME2, LOCALE_SDAYNAME3, LOCALE_SDAYNAME4, LOCALE_SDAYNAME5, LOCALE_SDAYNAME6, LOCALE_SDAYNAME7 }; day -= 1; LCTYPE type = short_format ? short_day_map[day] : long_day_map[day]; return getWinLocaleInfo(type); } static QString winMonthName(int month, bool short_format) { static const LCTYPE short_month_map[] = { LOCALE_SABBREVMONTHNAME1, LOCALE_SABBREVMONTHNAME2, LOCALE_SABBREVMONTHNAME3, LOCALE_SABBREVMONTHNAME4, LOCALE_SABBREVMONTHNAME5, LOCALE_SABBREVMONTHNAME6, LOCALE_SABBREVMONTHNAME7, LOCALE_SABBREVMONTHNAME8, LOCALE_SABBREVMONTHNAME9, LOCALE_SABBREVMONTHNAME10, LOCALE_SABBREVMONTHNAME11, LOCALE_SABBREVMONTHNAME12 }; static const LCTYPE long_month_map[] = { LOCALE_SMONTHNAME1, LOCALE_SMONTHNAME2, LOCALE_SMONTHNAME3, LOCALE_SMONTHNAME4, LOCALE_SMONTHNAME5, LOCALE_SMONTHNAME6, LOCALE_SMONTHNAME7, LOCALE_SMONTHNAME8, LOCALE_SMONTHNAME9, LOCALE_SMONTHNAME10, LOCALE_SMONTHNAME11, LOCALE_SMONTHNAME12 }; month -= 1; if (month < 0 || month > 11) return QString(); LCTYPE type = short_format ? short_month_map[month] : long_month_map[month]; return getWinLocaleInfo(type); } static QLocale::MeasurementSystem winSystemMeasurementSystem() { LCID id = GetUserDefaultLCID(); TCHAR output[2]; if (GetLocaleInfo(id, LOCALE_IMEASURE, output, 2)) { QString iMeasure = QT_WA_INLINE( QString::fromUtf16(reinterpret_cast(output)), QString::fromLocal8Bit(reinterpret_cast(output))); if (iMeasure == QString::fromLatin1("1")) { return QLocale::ImperialSystem; } } return QLocale::MetricSystem; } static QString winSystemAMText() { LCID id = GetUserDefaultLCID(); TCHAR output[15]; // maximum length including terminating zero character for Win2003+ if (GetLocaleInfo(id, LOCALE_S1159, output, 15)) { return QT_WA_INLINE( QString::fromUtf16(reinterpret_cast(output)), QString::fromLocal8Bit(reinterpret_cast(output))); } return QString(); } static QString winSystemPMText() { LCID id = GetUserDefaultLCID(); TCHAR output[15]; // maximum length including terminating zero character for Win2003+ if (GetLocaleInfo(id, LOCALE_S2359, output, 15)) { return QT_WA_INLINE( QString::fromUtf16(reinterpret_cast(output)), QString::fromLocal8Bit(reinterpret_cast(output))); } return QString(); } QLocale QSystemLocale::fallbackLocale() const { return QLocale(QString::fromLatin1(getWinLocaleName())); } QVariant QSystemLocale::query(QueryType type, QVariant in = QVariant()) const { LCTYPE locale_info = 0; bool format_string = false; switch(type) { // case Name: // return getWinLocaleName(); case DecimalPoint: locale_info = LOCALE_SDECIMAL; break; case GroupSeparator: locale_info = LOCALE_STHOUSAND; break; case NegativeSign: locale_info = LOCALE_SNEGATIVESIGN; break; case PositiveSign: locale_info = LOCALE_SPOSITIVESIGN; break; case DateFormatLong: locale_info = LOCALE_SLONGDATE; format_string = true; break; case DateFormatShort: locale_info = LOCALE_SSHORTDATE; format_string = true; break; case TimeFormatLong: case TimeFormatShort: locale_info = LOCALE_STIMEFORMAT; format_string = true; break; case DateTimeFormatLong: case DateTimeFormatShort: return query(type == DateTimeFormatLong ? DateFormatLong : DateFormatShort).toString() + QLatin1Char(' ') + query(type == DateTimeFormatLong ? TimeFormatLong : TimeFormatShort).toString(); case DayNameLong: case DayNameShort: return winDayName(in.toInt(), (type == DayNameShort)); case MonthNameLong: case MonthNameShort: return winMonthName(in.toInt(), (type == MonthNameShort)); case DateToStringShort: case DateToStringLong: return winDateToString(in.toDate(), type == DateToStringShort ? DATE_SHORTDATE : DATE_LONGDATE); case TimeToStringShort: case TimeToStringLong: return winTimeToString(in.toTime()); case DateTimeToStringShort: case DateTimeToStringLong: { const QDateTime dt = in.toDateTime(); return winDateToString(dt.date(), type == DateTimeToStringShort ? DATE_SHORTDATE : DATE_LONGDATE) + QLatin1Char(' ') + winTimeToString(dt.time()); } case ZeroDigit: locale_info = LOCALE_SNATIVEDIGITS; break; case LanguageId: case CountryId: { QString locale = QString::fromLatin1(getWinLocaleName()); QLocale::Language lang; QLocale::Country cntry; getLangAndCountry(locale, lang, cntry); if (type == LanguageId) return lang; if (cntry == QLocale::AnyCountry) return fallbackLocale().country(); return cntry; } case MeasurementSystem: return QVariant(static_cast(winSystemMeasurementSystem())); case AMText: return QVariant(winSystemAMText()); case PMText: return QVariant(winSystemPMText()); default: break; } if (locale_info) { QString result = getWinLocaleInfo(locale_info); if (format_string) result = winToQtFormat(result); if (!result.isEmpty()) return result; } return QVariant(); } /* Win95 doesn't have a function to return the ISO lang/country name of the user's locale. Instead it can return a "Windows code". This maps windows codes to ISO country names. */ struct WindowsToISOListElt { int windows_code; char iso_name[6]; }; static const WindowsToISOListElt windows_to_iso_list[] = { { 0x0401, "ar_SA" }, { 0x0402, "bg\0 " }, { 0x0403, "ca\0 " }, { 0x0404, "zh_TW" }, { 0x0405, "cs\0 " }, { 0x0406, "da\0 " }, { 0x0407, "de\0 " }, { 0x0408, "el\0 " }, { 0x0409, "en_US" }, { 0x040a, "es\0 " }, { 0x040b, "fi\0 " }, { 0x040c, "fr\0 " }, { 0x040d, "he\0 " }, { 0x040e, "hu\0 " }, { 0x040f, "is\0 " }, { 0x0410, "it\0 " }, { 0x0411, "ja\0 " }, { 0x0412, "ko\0 " }, { 0x0413, "nl\0 " }, { 0x0414, "no\0 " }, { 0x0415, "pl\0 " }, { 0x0416, "pt_BR" }, { 0x0418, "ro\0 " }, { 0x0419, "ru\0 " }, { 0x041a, "hr\0 " }, { 0x041c, "sq\0 " }, { 0x041d, "sv\0 " }, { 0x041e, "th\0 " }, { 0x041f, "tr\0 " }, { 0x0420, "ur\0 " }, { 0x0421, "in\0 " }, { 0x0422, "uk\0 " }, { 0x0423, "be\0 " }, { 0x0425, "et\0 " }, { 0x0426, "lv\0 " }, { 0x0427, "lt\0 " }, { 0x0429, "fa\0 " }, { 0x042a, "vi\0 " }, { 0x042d, "eu\0 " }, { 0x042f, "mk\0 " }, { 0x0436, "af\0 " }, { 0x0438, "fo\0 " }, { 0x0439, "hi\0 " }, { 0x043e, "ms\0 " }, { 0x0458, "mt\0 " }, { 0x0801, "ar_IQ" }, { 0x0804, "zh_CN" }, { 0x0807, "de_CH" }, { 0x0809, "en_GB" }, { 0x080a, "es_MX" }, { 0x080c, "fr_BE" }, { 0x0810, "it_CH" }, { 0x0812, "ko\0 " }, { 0x0813, "nl_BE" }, { 0x0814, "no\0 " }, { 0x0816, "pt\0 " }, { 0x081a, "sr\0 " }, { 0x081d, "sv_FI" }, { 0x0c01, "ar_EG" }, { 0x0c04, "zh_HK" }, { 0x0c07, "de_AT" }, { 0x0c09, "en_AU" }, { 0x0c0a, "es\0 " }, { 0x0c0c, "fr_CA" }, { 0x0c1a, "sr\0 " }, { 0x1001, "ar_LY" }, { 0x1004, "zh_SG" }, { 0x1007, "de_LU" }, { 0x1009, "en_CA" }, { 0x100a, "es_GT" }, { 0x100c, "fr_CH" }, { 0x1401, "ar_DZ" }, { 0x1407, "de_LI" }, { 0x1409, "en_NZ" }, { 0x140a, "es_CR" }, { 0x140c, "fr_LU" }, { 0x1801, "ar_MA" }, { 0x1809, "en_IE" }, { 0x180a, "es_PA" }, { 0x1c01, "ar_TN" }, { 0x1c09, "en_ZA" }, { 0x1c0a, "es_DO" }, { 0x2001, "ar_OM" }, { 0x2009, "en_JM" }, { 0x200a, "es_VE" }, { 0x2401, "ar_YE" }, { 0x2409, "en\0 " }, { 0x240a, "es_CO" }, { 0x2801, "ar_SY" }, { 0x2809, "en_BZ" }, { 0x280a, "es_PE" }, { 0x2c01, "ar_JO" }, { 0x2c09, "en_TT" }, { 0x2c0a, "es_AR" }, { 0x3001, "ar_LB" }, { 0x300a, "es_EC" }, { 0x3401, "ar_KW" }, { 0x340a, "es_CL" }, { 0x3801, "ar_AE" }, { 0x380a, "es_UY" }, { 0x3c01, "ar_BH" }, { 0x3c0a, "es_PY" }, { 0x4001, "ar_QA" }, { 0x400a, "es_BO" }, { 0x440a, "es_SV" }, { 0x480a, "es_HN" }, { 0x4c0a, "es_NI" }, { 0x500a, "es_PR" } }; static const int windows_to_iso_count = sizeof(windows_to_iso_list)/sizeof(WindowsToISOListElt); static const char *winLangCodeToIsoName(int code) { int cmp = code - windows_to_iso_list[0].windows_code; if (cmp < 0) return 0; if (cmp == 0) return windows_to_iso_list[0].iso_name; int begin = 0; int end = windows_to_iso_count; while (end - begin > 1) { uint mid = (begin + end)/2; const WindowsToISOListElt *elt = windows_to_iso_list + mid; int cmp = code - elt->windows_code; if (cmp < 0) end = mid; else if (cmp > 0) begin = mid; else return elt->iso_name; } return 0; } static QString winIso639LangName(LCID id) { QString result; // Windows returns the wrong ISO639 for some languages, we need to detect them here using // the language code QString lang_code; QT_WA({ TCHAR out[256]; if (GetLocaleInfoW(id, LOCALE_ILANGUAGE, out, 255)) lang_code = QString::fromUtf16((ushort*)out); } , { char out[256]; if (GetLocaleInfoA(id, LOCALE_ILANGUAGE, out, 255)) lang_code = QString::fromLocal8Bit(out); }); if (!lang_code.isEmpty()) { const char *endptr; bool ok; QByteArray latin1_lang_code = lang_code.toLatin1(); int i = qstrtoull(latin1_lang_code, &endptr, 16, &ok); if (ok && *endptr == '\0') { switch (i) { case 0x814: result = QLatin1String("nn"); // Nynorsk break; default: break; } } } if (!result.isEmpty()) return result; // not one of the problematic languages - do the usual lookup QT_WA({ TCHAR out[256]; if (GetLocaleInfoW(id, LOCALE_SISO639LANGNAME , out, 255)) result = QString::fromUtf16((ushort*)out); } , { char out[256]; if (GetLocaleInfoA(id, LOCALE_SISO639LANGNAME, out, 255)) result = QString::fromLocal8Bit(out); }); return result; } static QString winIso3116CtryName(LCID id) { QString result; QT_WA({ TCHAR out[256]; if (GetLocaleInfoW(id, LOCALE_SISO3166CTRYNAME, out, 255)) result = QString::fromUtf16((ushort*)out); } , { char out[256]; if (GetLocaleInfoA(id, LOCALE_SISO3166CTRYNAME, out, 255)) result = QString::fromLocal8Bit(out); }); return result; } #elif defined(Q_OS_MAC) /****************************************************************************** ** Wrappers for Mac locale system functions */ static QByteArray getMacLocaleName() { QByteArray result = envVarLocale(); QChar lang[3]; QChar cntry[2]; if (result.isEmpty() || result != "C" && !splitLocaleName(QString::fromLocal8Bit(result), lang, cntry)) { QCFType l = CFLocaleCopyCurrent(); CFStringRef locale = CFLocaleGetIdentifier(l); result = QCFString::toQString(locale).toUtf8(); } return result; } static QString macMonthName(int month, bool short_format) { month -= 1; if (month < 0 || month > 11) return QString(); #if (MAC_OS_X_VERSION_MAX_ALLOWED >= MAC_OS_X_VERSION_10_4) if (QSysInfo::MacintoshVersion >= QSysInfo::MV_10_4) { QCFType formatter = CFDateFormatterCreate(0, QCFType(CFLocaleCopyCurrent()), kCFDateFormatterNoStyle, kCFDateFormatterNoStyle); QCFType values = static_cast(CFDateFormatterCopyProperty(formatter, short_format ? kCFDateFormatterShortMonthSymbols : kCFDateFormatterMonthSymbols)); if (values != 0) { CFStringRef cfstring = static_cast(CFArrayGetValueAtIndex(values, month)); return QCFString::toQString(cfstring); } } #endif return QString(); } static QString macDayName(int day, bool short_format) { if (day < 1 || day > 7) return QString(); #if (MAC_OS_X_VERSION_MAX_ALLOWED >= MAC_OS_X_VERSION_10_4) if (QSysInfo::MacintoshVersion >= QSysInfo::MV_10_4) { QCFType formatter = CFDateFormatterCreate(0, QCFType(CFLocaleCopyCurrent()), kCFDateFormatterNoStyle, kCFDateFormatterNoStyle); QCFType values = static_cast(CFDateFormatterCopyProperty(formatter, short_format ? kCFDateFormatterShortWeekdaySymbols : kCFDateFormatterWeekdaySymbols)); if (values != 0) { CFStringRef cfstring = static_cast(CFArrayGetValueAtIndex(values, day % 7)); return QCFString::toQString(cfstring); } } #endif return QString(); } static QString macDateToString(const QDate &date, bool short_format) { CFGregorianDate macGDate; macGDate.year = date.year(); macGDate.month = date.month(); macGDate.day = date.day(); macGDate.hour = 0; macGDate.minute = 0; macGDate.second = 0.0; QCFType myDate = CFDateCreate(0, CFGregorianDateGetAbsoluteTime(macGDate, QCFType(CFTimeZoneCopyDefault()))); QCFType mylocale = CFLocaleCopyCurrent(); CFDateFormatterStyle style = short_format ? kCFDateFormatterShortStyle : kCFDateFormatterLongStyle; QCFType myFormatter = CFDateFormatterCreate(kCFAllocatorDefault, mylocale, style, kCFDateFormatterNoStyle); return QCFString(CFDateFormatterCreateStringWithDate(0, myFormatter, myDate)); } static QString macTimeToString(const QTime &time, bool short_format) { CFGregorianDate macGDate; // Assume this is local time and the current date QDate dt = QDate::currentDate(); macGDate.year = dt.year(); macGDate.month = dt.month(); macGDate.day = dt.day(); macGDate.hour = time.hour(); macGDate.minute = time.minute(); macGDate.second = time.second(); QCFType myDate = CFDateCreate(0, CFGregorianDateGetAbsoluteTime(macGDate, QCFType(CFTimeZoneCopyDefault()))); QCFType mylocale = CFLocaleCopyCurrent(); CFDateFormatterStyle style = short_format ? kCFDateFormatterShortStyle : kCFDateFormatterLongStyle; QCFType myFormatter = CFDateFormatterCreate(kCFAllocatorDefault, mylocale, kCFDateFormatterNoStyle, style); return QCFString(CFDateFormatterCreateStringWithDate(0, myFormatter, myDate)); } static QString macToQtFormat(const QString &sys_fmt) { QString result; int i = 0; while (i < sys_fmt.size()) { if (sys_fmt.at(i).unicode() == '\'') { QString text = readEscapedFormatString(sys_fmt, &i); if (text == QLatin1String("'")) result += QLatin1String("''"); else result += QLatin1Char('\'') + text + QLatin1Char('\''); continue; } QChar c = sys_fmt.at(i); int repeat = repeatCount(sys_fmt, i); switch (c.unicode()) { case 'G': // Qt doesn't support these :( case 'Y': case 'D': case 'F': case 'w': case 'W': case 'g': break; case 'u': // extended year - use 'y' if (repeat < 4) result += QLatin1String("yy"); else result += QLatin1String("yyyy"); break; case 'S': // fractional second if (repeat < 3) result += QLatin1String("z"); else result += QLatin1String("zzz"); break; case 'E': if (repeat <= 3) result += QLatin1String("ddd"); else result += QLatin1String("dddd"); break; case 'e': if (repeat >= 2) result += QLatin1String("dd"); else result += QLatin1String("d"); break; case 'a': result += QLatin1String("AP"); break; case 'k': result += QString(repeat, QLatin1Char('H')); break; case 'K': result += QString(repeat, QLatin1Char('h')); break; case 'z': case 'Z': case 'v': result += QLatin1Char('t'); break; default: result += QString(repeat, c); break; } i += repeat; } return result; } QString getMacDateFormat(CFDateFormatterStyle style) { QCFType l = CFLocaleCopyCurrent(); QCFType formatter = CFDateFormatterCreate(kCFAllocatorDefault, l, style, kCFDateFormatterNoStyle); return macToQtFormat(QCFString::toQString(CFDateFormatterGetFormat(formatter))); } static QString getMacTimeFormat(CFDateFormatterStyle style) { QCFType l = CFLocaleCopyCurrent(); QCFType formatter = CFDateFormatterCreate(kCFAllocatorDefault, l, kCFDateFormatterNoStyle, style); return macToQtFormat(QCFString::toQString(CFDateFormatterGetFormat(formatter))); } static QString getCFLocaleValue(CFStringRef key) { QCFType locale = CFLocaleCopyCurrent(); CFTypeRef value = CFLocaleGetValue(locale, key); return QCFString::toQString(CFStringRef(static_cast(value))); } static QLocale::MeasurementSystem macMeasurementSystem() { QCFType locale = CFLocaleCopyCurrent(); CFStringRef system = static_cast(CFLocaleGetValue(locale, kCFLocaleMeasurementSystem)); if (QCFString::toQString(system) == QLatin1String("Metric")) { return QLocale::MetricSystem; } else { return QLocale::ImperialSystem; } } QLocale QSystemLocale::fallbackLocale() const { return QLocale(QString::fromUtf8(getMacLocaleName().constData())); } QVariant QSystemLocale::query(QueryType type, QVariant in = QVariant()) const { switch(type) { // case Name: // return getMacLocaleName(); case DecimalPoint: { QString value = getCFLocaleValue(kCFLocaleDecimalSeparator); return value.isEmpty() ? QVariant() : value; } case GroupSeparator: { QString value = getCFLocaleValue(kCFLocaleGroupingSeparator); return value.isEmpty() ? QVariant() : value; } case DateFormatLong: case DateFormatShort: return macToQtFormat(getMacDateFormat(type == DateFormatShort ? kCFDateFormatterShortStyle : kCFDateFormatterLongStyle)); case TimeFormatLong: case TimeFormatShort: return macToQtFormat(getMacTimeFormat(type == TimeFormatShort ? kCFDateFormatterShortStyle : kCFDateFormatterLongStyle)); case DayNameLong: case DayNameShort: return macDayName(in.toInt(), (type == DayNameShort)); case MonthNameLong: case MonthNameShort: return macMonthName(in.toInt(), (type == MonthNameShort)); case DateToStringShort: case DateToStringLong: return macDateToString(in.toDate(), (type == DateToStringShort)); case TimeToStringShort: case TimeToStringLong: return macTimeToString(in.toTime(), (type == TimeToStringShort)); case NegativeSign: case PositiveSign: case ZeroDigit: case LanguageId: case CountryId: break; case MeasurementSystem: return QVariant(static_cast(macMeasurementSystem())); case AMText: case PMText: break; default: break; } return QVariant(); } #elif defined(Q_OS_UNIX) static uint unixGetSystemMeasurementSystem() { QString meas_locale = QString::fromLocal8Bit(qgetenv("LC_ALL")); if (meas_locale.isEmpty()) { meas_locale = QString::fromLocal8Bit(qgetenv("LC_MEASUREMENT")); } if (meas_locale.isEmpty()) { meas_locale = QString::fromLocal8Bit(qgetenv("LANG")); } if (meas_locale.isEmpty()) { meas_locale = QString::fromLocal8Bit("C"); } if (meas_locale.compare(QString::fromLocal8Bit("Metric"), Qt::CaseInsensitive) == 0) return 0; if (meas_locale.compare(QString::fromLocal8Bit("Other"), Qt::CaseInsensitive) == 0) return 0; const QLocalePrivate* locale = findLocale(meas_locale); return locale->measurementSystem(); } /*! \internal */ QLocale QSystemLocale::fallbackLocale() const { return QLocale(QLatin1String(envVarLocale())); } /*! \internal */ QVariant QSystemLocale::query(QueryType type, QVariant /* in */) const { if (type == MeasurementSystem) { return QVariant(unixGetSystemMeasurementSystem()); } else { return QVariant(); } } #else /*! Returns a fallback locale, that will get used for everything that is not explicitly overridden by the system locale. */ QLocale QSystemLocale::fallbackLocale() const { return QLocale(QLatin1String(envVarLocale())); } /*! Performs a query of the given \a type in the system locale for customized values or conversion. If the method returns a null QVariant, the conversion of the fallbackLocale() will be used. \a in is unused for some of the query types. \sa QSystemLocale::QueryType */ QVariant QSystemLocale::query(QueryType /* type */, QVariant /* in */) const { return QVariant(); } #endif #ifndef QT_NO_SYSTEMLOCALE static QSystemLocale *_systemLocale = 0; Q_GLOBAL_STATIC_WITH_ARGS(QSystemLocale, QSystemLocale_globalSystemLocale, (true)) static QLocalePrivate *system_lp = 0; Q_GLOBAL_STATIC(QLocalePrivate, globalLocalePrivate) #endif /****************************************************************************** ** Default system locale behavior */ /*! \class QSystemLocale \brief The QSystemLocale class can be used to finetune the system locale of the user. \since 4.2 \ingroup i18n \warning This class is only useful in very rare cases. Usually QLocale offers all the functionality required for application development. QSystemLocale allows to override the values provided by the system locale (QLocale::system()). \sa QLocale */ /*! \enum QSystemLocale::QueryType Specifies the type of information queried by query(). For each value the type of information to return from the query() method is listed. \value LanguageId a uint specifying the language. \value CountryId a uint specifying the country. \value DecimalPoint a QString specifying the decimal point. \value GroupSeparator a QString specifying the group separator. \value ZeroDigit a QString specifying the zero digit. \value NegativeSign a QString specifying the minus sign. \value PositiveSign a QString specifying the plus sign. \value DateFormatLong a QString specifying the long date format \value DateFormatShort a QString specifying the short date format \value TimeFormatLong a QString specifying the long time format \value TimeFormatShort a QString specifying the short time format \value DayNameLong a QString specifying the name of a weekday. the in variant contains an integer between 1 and 7 (Monday - Sunday) \value DayNameShort a QString specifying the short name of a weekday. the in variant contains an integer between 1 and 7 (Monday - Sunday) \value MonthNameLong a QString specifying the name of a month. the in variant contains an integer between 1 and 12 \value MonthNameShort a QString specifying the short name of a month. the in variant contains an integer between 1 and 12 \value DateToStringLong converts the QDate stored in the in variant to a QString using the long date format \value DateToStringShort converts the QDate stored in the in variant to a QString using the short date format \value TimeToStringLong converts the QTime stored in the in variant to a QString using the long time format \value TimeToStringShort converts the QTime stored in the in variant to a QString using the short time format \value DateTimeFormatLong a QString specifying the long date time format \value DateTimeFormatShort a QString specifying the short date time format \value DateTimeToStringLong converts the QDateTime in the in variant to a QString using the long datetime format \value DateTimeToStringShort converts the QDateTime in the in variant to a QString using the short datetime format \value MeasurementSystem a QLocale::MeasurementSystem enum specifying the measurement system \value AMText a string that represents the system AM designator associated with a 12-hour clock. \value PMText a string that represents the system PM designator associated with a 12-hour clock. */ /*! Constructs a QSystemLocale object. The constructor will automatically install this object as the system locale and remove any earlier installed system locales. */ QSystemLocale::QSystemLocale() { delete _systemLocale; _systemLocale = this; if (system_lp) system_lp->m_language_id = 0; } /*! \internal */ QSystemLocale::QSystemLocale(bool) { } /*! Deletes the object. */ QSystemLocale::~QSystemLocale() { if (_systemLocale == this) { _systemLocale = 0; if (system_lp) system_lp->m_language_id = 0; } } static const QSystemLocale *systemLocale() { if (_systemLocale) return _systemLocale; return QSystemLocale_globalSystemLocale(); } void QLocalePrivate::updateSystemPrivate() { const QSystemLocale *sys_locale = systemLocale(); if (!system_lp) system_lp = globalLocalePrivate(); *system_lp = *sys_locale->fallbackLocale().d(); QVariant res = sys_locale->query(QSystemLocale::LanguageId, QVariant()); if (!res.isNull()) system_lp->m_language_id = res.toInt(); res = sys_locale->query(QSystemLocale::CountryId, QVariant()); if (!res.isNull()) system_lp->m_country_id = res.toInt(); res = sys_locale->query(QSystemLocale::DecimalPoint, QVariant()); if (!res.isNull()) system_lp->m_decimal = res.toString().at(0).unicode(); res = sys_locale->query(QSystemLocale::GroupSeparator, QVariant()); if (!res.isNull()) system_lp->m_group = res.toString().at(0).unicode(); res = sys_locale->query(QSystemLocale::ZeroDigit, QVariant()); if (!res.isNull()) system_lp->m_zero = res.toString().at(0).unicode(); res = sys_locale->query(QSystemLocale::NegativeSign, QVariant()); if (!res.isNull()) system_lp->m_minus = res.toString().at(0).unicode(); res = sys_locale->query(QSystemLocale::PositiveSign, QVariant()); if (!res.isNull()) system_lp->m_plus = res.toString().at(0).unicode(); } #endif static const QLocalePrivate *systemPrivate() { #ifndef QT_NO_SYSTEMLOCALE // copy over the information from the fallback locale and modify if (!system_lp || system_lp->m_language_id == 0) QLocalePrivate::updateSystemPrivate(); return system_lp; #else return locale_data; #endif } static const QLocalePrivate *defaultPrivate() { if (!default_lp) default_lp = systemPrivate(); return default_lp; } static QString getLocaleListData(const ushort *data, int size, int index) { static const ushort separator = ';'; while (index && size > 0) { while (*data != separator) ++data, --size; --index; ++data; --size; } const ushort *end = data; while (size > 0 && *end != separator) ++end, --size; return QString::fromRawData(reinterpret_cast(data), end-data); } static inline QString getLocaleData(const ushort *data, int size) { return QString::fromRawData(reinterpret_cast(data), size); } #ifndef QT_NO_DATASTREAM QDataStream &operator<<(QDataStream &ds, const QLocale &l) { ds << l.name(); return ds; } QDataStream &operator>>(QDataStream &ds, QLocale &l) { QString s; ds >> s; l = QLocale(s); return ds; } #endif /*! \class QLocale \brief The QLocale class converts between numbers and their string representations in various languages. \reentrant \ingroup i18n \ingroup text \ingroup shared \mainclass QLocale is initialized with a language/country pair in its constructor and offers number-to-string and string-to-number conversion functions similar to those in QString. Example: \snippet doc/src/snippets/code/src_corelib_tools_qlocale.cpp 0 QLocale supports the concept of a default locale, which is determined from the system's locale settings at application startup. The default locale can be changed by calling the static member setDefault(). Setting the default locale has the following effects: \list \i If a QLocale object is constructed with the default constructor, it will use the default locale's settings. \i QString::toInt(), QString::toDouble(), etc., interpret the string according to the default locale. If this fails, it falls back on the "C" locale. \i QString::arg() uses the default locale to format a number when its position specifier in the format string contains an 'L', e.g. "%L1". \endlist The following example illustrates how to use QLocale directly: \snippet doc/src/snippets/code/src_corelib_tools_qlocale.cpp 1 When a language/country pair is specified in the constructor, one of three things can happen: \list \i If the language/country pair is found in the database, it is used. \i If the language is found but the country is not, or if the country is \c AnyCountry, the language is used with the most appropriate available country (for example, Germany for German), \i If neither the language nor the country are found, QLocale defaults to the default locale (see setDefault()). \endlist The "C" locale is identical in behavior to \l{English}/\l{UnitedStates}. Use language() and country() to determine the actual language and country values used. An alternative method for constructing a QLocale object is by specifying the locale name. \snippet doc/src/snippets/code/src_corelib_tools_qlocale.cpp 2 This constructor converts the locale name to a language/country pair; it does not use the system locale database. The double-to-string and string-to-double conversion functions are covered by the following licenses: \legalese Copyright (c) 1991 by AT&T. Permission to use, copy, modify, and distribute this software for any purpose without fee is hereby granted, provided that this entire notice is included in all copies of any software which is or includes a copy or modification of this software and in all copies of the supporting documentation for such software. THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED WARRANTY. IN PARTICULAR, NEITHER THE AUTHOR NOR AT&T MAKES ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE. This product includes software developed by the University of California, Berkeley and its contributors. QLocale's data is based on Common Locale Data Repository v1.6.1. \sa QString::arg(), QString::toInt(), QString::toDouble() */ /*! \enum QLocale::Language This enumerated type is used to specify a language. \value C The "C" locale is identical in behavior to English/UnitedStates. \value Abkhazian \value Afan \value Afar \value Afrikaans \value Albanian \value Amharic \value Arabic \value Armenian \value Assamese \value Aymara \value Azerbaijani \value Bashkir \value Basque \value Bengali \value Bhutani \value Bihari \value Bislama \value Bosnian \value Breton \value Bulgarian \value Burmese \value Byelorussian \value Cambodian \value Catalan \value Chinese \value Cornish \value Corsican \value Croatian \value Czech \value Danish \value Divehi \value Dutch \value English \value Esperanto \value Estonian \value Faroese \value FijiLanguage \value Finnish \value French \value Frisian \value Gaelic \value Galician \value Georgian \value German \value Greek \value Greenlandic \value Guarani \value Gujarati \value Hausa \value Hebrew \value Hindi \value Hungarian \value Icelandic \value Indonesian \value Interlingua \value Interlingue \value Inuktitut \value Inupiak \value Irish \value Italian \value Japanese \value Javanese \value Kannada \value Kashmiri \value Kazakh \value Kinyarwanda \value Kirghiz \value Korean \value Kurdish \value Kurundi \value Laothian \value Latin \value Latvian \value Lingala \value Lithuanian \value Macedonian \value Malagasy \value Malay \value Malayalam \value Maltese \value Manx \value Maori \value Marathi \value Moldavian \value Mongolian \value NauruLanguage \value Nepali \value Norwegian \value NorwegianBokmal \value Nynorsk Obsolete, please use NorwegianNynorsk \value NorwegianNynorsk \value Occitan \value Oriya \value Pashto \value Persian \value Polish \value Portuguese \value Punjabi \value Quechua \value RhaetoRomance \value Romanian \value Russian \value Samoan \value Sangho \value Sanskrit \value Serbian \value SerboCroatian \value Sesotho \value Setswana \value Shona \value Sindhi \value Singhalese \value Siswati \value Slovak \value Slovenian \value Somali \value Spanish \value Sundanese \value Swahili \value Swedish \value Tagalog \value Tajik \value Tamil \value Tatar \value Telugu \value Thai \value Tibetan \value Tigrinya \value TongaLanguage \value Tsonga \value Turkish \value Turkmen \value Twi \value Uigur \value Ukrainian \value Urdu \value Uzbek \value Vietnamese \value Volapuk \value Welsh \value Wolof \value Xhosa \value Yiddish \value Yoruba \value Zhuang \value Zulu \value Bosnian \value Divehi \value Manx \value Cornish \value Akan \value Konkani \value Ga \value Igbo \value Kamba \value Syriac \value Blin \value Geez \value Koro \value Sidamo \value Atsam \value Tigre \value Jju \value Friulian \value Venda \value Ewe \value Walamo \value Hawaiian \value Tyap \value Chewa \omitvalue LastLanguage \sa language() */ /*! \enum QLocale::Country This enumerated type is used to specify a country. \value AnyCountry \value Afghanistan \value Albania \value Algeria \value AmericanSamoa \value Andorra \value Angola \value Anguilla \value Antarctica \value AntiguaAndBarbuda \value Argentina \value Armenia \value Aruba \value Australia \value Austria \value Azerbaijan \value Bahamas \value Bahrain \value Bangladesh \value Barbados \value Belarus \value Belgium \value Belize \value Benin \value Bermuda \value Bhutan \value Bolivia \value BosniaAndHerzegowina \value Botswana \value BouvetIsland \value Brazil \value BritishIndianOceanTerritory \value BruneiDarussalam \value Bulgaria \value BurkinaFaso \value Burundi \value Cambodia \value Cameroon \value Canada \value CapeVerde \value CaymanIslands \value CentralAfricanRepublic \value Chad \value Chile \value China \value ChristmasIsland \value CocosIslands \value Colombia \value Comoros \value DemocraticRepublicOfCongo \value PeoplesRepublicOfCongo \value CookIslands \value CostaRica \value IvoryCoast \value Croatia \value Cuba \value Cyprus \value CzechRepublic \value Denmark \value Djibouti \value Dominica \value DominicanRepublic \value EastTimor \value Ecuador \value Egypt \value ElSalvador \value EquatorialGuinea \value Eritrea \value Estonia \value Ethiopia \value FalklandIslands \value FaroeIslands \value FijiCountry \value Finland \value France \value MetropolitanFrance \value FrenchGuiana \value FrenchPolynesia \value FrenchSouthernTerritories \value Gabon \value Gambia \value Georgia \value Germany \value Ghana \value Gibraltar \value Greece \value Greenland \value Grenada \value Guadeloupe \value Guam \value Guatemala \value Guinea \value GuineaBissau \value Guyana \value Haiti \value HeardAndMcDonaldIslands \value Honduras \value HongKong \value Hungary \value Iceland \value India \value Indonesia \value Iran \value Iraq \value Ireland \value Israel \value Italy \value Jamaica \value Japan \value Jordan \value Kazakhstan \value Kenya \value Kiribati \value DemocraticRepublicOfKorea \value RepublicOfKorea \value Kuwait \value Kyrgyzstan \value Lao \value Latvia \value Lebanon \value Lesotho \value Liberia \value LibyanArabJamahiriya \value Liechtenstein \value Lithuania \value Luxembourg \value Macau \value Macedonia \value Madagascar \value Malawi \value Malaysia \value Maldives \value Mali \value Malta \value MarshallIslands \value Martinique \value Mauritania \value Mauritius \value Mayotte \value Mexico \value Micronesia \value Moldova \value Monaco \value Mongolia \value Montserrat \value Morocco \value Mozambique \value Myanmar \value Namibia \value NauruCountry \value Nepal \value Netherlands \value NetherlandsAntilles \value NewCaledonia \value NewZealand \value Nicaragua \value Niger \value Nigeria \value Niue \value NorfolkIsland \value NorthernMarianaIslands \value Norway \value Oman \value Pakistan \value Palau \value PalestinianTerritory \value Panama \value PapuaNewGuinea \value Paraguay \value Peru \value Philippines \value Pitcairn \value Poland \value Portugal \value PuertoRico \value Qatar \value Reunion \value Romania \value RussianFederation \value Rwanda \value SaintKittsAndNevis \value StLucia \value StVincentAndTheGrenadines \value Samoa \value SanMarino \value SaoTomeAndPrincipe \value SaudiArabia \value Senegal \value SerbiaAndMontenegro \value Seychelles \value SierraLeone \value Singapore \value Slovakia \value Slovenia \value SolomonIslands \value Somalia \value SouthAfrica \value SouthGeorgiaAndTheSouthSandwichIslands \value Spain \value SriLanka \value StHelena \value StPierreAndMiquelon \value Sudan \value Suriname \value SvalbardAndJanMayenIslands \value Swaziland \value Sweden \value Switzerland \value SyrianArabRepublic \value Taiwan \value Tajikistan \value Tanzania \value Thailand \value Togo \value Tokelau \value TongaCountry \value TrinidadAndTobago \value Tunisia \value Turkey \value Turkmenistan \value TurksAndCaicosIslands \value Tuvalu \value Uganda \value Ukraine \value UnitedArabEmirates \value UnitedKingdom \value UnitedStates \value UnitedStatesMinorOutlyingIslands \value Uruguay \value Uzbekistan \value Vanuatu \value VaticanCityState \value Venezuela \value VietNam \value BritishVirginIslands \value USVirginIslands \value WallisAndFutunaIslands \value WesternSahara \value Yemen \value Yugoslavia \value Zambia \value Zimbabwe \omitvalue LastCountry \sa country() */ /*! \enum QLocale::FormatType This enum describes the types of format that can be used when converting QDate and QTime objects to strings. \value LongFormat The long version of day and month names; for example, returning "January" as a month name. \value ShortFormat The short version of day and month names; for example, returning "Jan" as a month name. \value NarrowFormat A special version of day and month names for use when space is limited; for example, returning "J" as a month name. Note that the narrow format might contain the same text for different months and days or it can even be an empty string if the locale doesn't support narrow names, so you should avoid using it for date formatting. Also, for the system locale this format is the same as ShortFormat. */ /*! \enum QLocale::NumberOption This enum defines a set of options for number-to-string and string-to-number conversions. They can be retrieved with numberOptions() and set with setNumberOptions(). \value OmitGroupSeparator If this option is set, the number-to-string functions will not insert group separators in their return values. The default is to insert group separators. \value RejectGroupSeparator If this option is set, the string-to-number functions will fail if they encounter group separators in their input. The default is to accept numbers containing correctly placed group separators. \sa setNumberOptions() numberOptions() */ /*! \enum QLocale::MeasurementSystem This enum defines which units are used for measurement. \value MetricSystem This value indicates metric units, such as meters, centimeters and millimeters. \value ImperialSystem This value indicates imperial units, such as inches and miles. There are several distinct imperial systems in the world; this value stands for the official United States imperial units. \since 4.4 */ /*! \fn bool QLocale::operator==(const QLocale &other) const Returns true if the QLocale object is the same as the \a other locale specified; otherwise returns false. */ /*! \fn bool QLocale::operator!=(const QLocale &other) const Returns true if the QLocale object is not the same as the \a other locale specified; otherwise returns false. */ static const int locale_data_size = sizeof(locale_data)/sizeof(QLocalePrivate) - 1; static const QLocalePrivate *dataPointerHelper(quint16 index) { #ifndef QT_NO_SYSTEMLOCALE Q_ASSERT(index <= locale_data_size); if (index == locale_data_size) return system_lp; #else Q_ASSERT(index < locale_data_size); #endif return &locale_data[index]; } static quint16 localePrivateIndex(const QLocalePrivate *p) { #ifndef QT_NO_SYSTEMLOCALE Q_ASSERT((p >= locale_data && p - locale_data < locale_data_size) || (p != 0 && p == system_lp)); quint16 index = p == system_lp ? locale_data_size : p - locale_data; #else Q_ASSERT(p >= locale_data && p - locale_data < locale_data_size); quint16 index = p - locale_data; #endif return index; } /*! Constructs a QLocale object with the specified \a name, which has the format "language[_country][.codeset][@modifier]" or "C", where: \list \i language is a lowercase, two-letter, ISO 639 language code, \i territory is an uppercase, two-letter, ISO 3166 country code, \i and codeset and modifier are ignored. \endlist If the string violates the locale format, or language is not a valid ISO 369 code, the "C" locale is used instead. If country is not present, or is not a valid ISO 3166 code, the most appropriate country is chosen for the specified language. The language and country codes are converted to their respective \c Language and \c Country enums. After this conversion is performed the constructor behaves exactly like QLocale(Country, Language). This constructor is much slower than QLocale(Country, Language). \sa name() */ QLocale::QLocale(const QString &name) : v(0) { p.numberOptions = 0; p.index = localePrivateIndex(findLocale(name)); } /*! Constructs a QLocale object initialized with the default locale. If no default locale was set using setDefaultLocale(), this locale will be the same as the one returned by system(). \sa setDefault() */ QLocale::QLocale() : v(0) { p.numberOptions = default_number_options; p.index = localePrivateIndex(defaultPrivate()); } /*! Constructs a QLocale object with the specified \a language and \a country. \list \i If the language/country pair is found in the database, it is used. \i If the language is found but the country is not, or if the country is \c AnyCountry, the language is used with the most appropriate available country (for example, Germany for German), \i If neither the language nor the country are found, QLocale defaults to the default locale (see setDefault()). \endlist The language and country that are actually used can be queried using language() and country(). \sa setDefault() language() country() */ QLocale::QLocale(Language language, Country country) : v(0) { const QLocalePrivate *d = findLocale(language, country); // If not found, should default to system if (d->languageId() == QLocale::C && language != QLocale::C) { p.numberOptions = default_number_options; p.index = localePrivateIndex(defaultPrivate()); } else { p.numberOptions = 0; p.index = localePrivateIndex(d); } } /*! Constructs a QLocale object as a copy of \a other. */ QLocale::QLocale(const QLocale &other) { v = other.v; } const QLocalePrivate *QLocale::d() const { return dataPointerHelper(p.index); } /*! Assigns \a other to this QLocale object and returns a reference to this QLocale object. */ QLocale &QLocale::operator=(const QLocale &other) { v = other.v; return *this; } /*! \since 4.2 Sets the \a options related to number conversions for this QLocale instance. */ void QLocale::setNumberOptions(NumberOptions options) { p.numberOptions = options; } /*! \since 4.2 Returns the options related to number conversions for this QLocale instance. By default, no options are set for the standard locales. */ QLocale::NumberOptions QLocale::numberOptions() const { return static_cast(p.numberOptions); } /*! \nonreentrant Sets the global default locale to \a locale. These values are used when a QLocale object is constructed with no arguments. If this function is not called, the system's locale is used. \warning In a multithreaded application, the default locale should be set at application startup, before any non-GUI threads are created. \sa system() c() */ void QLocale::setDefault(const QLocale &locale) { default_lp = locale.d(); default_number_options = locale.numberOptions(); } /*! Returns the language of this locale. \sa country(), languageToString(), name() */ QLocale::Language QLocale::language() const { return Language(d()->languageId()); } /*! Returns the country of this locale. \sa language(), countryToString(), name() */ QLocale::Country QLocale::country() const { return Country(d()->countryId()); } /*! Returns the language and country of this locale as a string of the form "language_country", where language is a lowercase, two-letter ISO 639 language code, and country is an uppercase, two-letter ISO 3166 country code. \sa language(), country() */ QString QLocale::name() const { Language l = language(); QString result = languageToCode(l); if (l == C) return result; Country c = country(); if (c == AnyCountry) return result; result.append(QLatin1Char('_')); result.append(countryToCode(c)); return result; } /*! Returns a QString containing the name of \a language. \sa countryToString(), name() */ QString QLocale::languageToString(Language language) { if (uint(language) > uint(QLocale::LastLanguage)) return QLatin1String("Unknown"); return QLatin1String(language_name_list + language_name_index[language]); } /*! Returns a QString containing the name of \a country. \sa country(), name() */ QString QLocale::countryToString(Country country) { if (uint(country) > uint(QLocale::LastCountry)) return QLatin1String("Unknown"); return QLatin1String(country_name_list + country_name_index[country]); } /*! Returns the short int represented by the localized string \a s, using base \a base. If \a base is 0 the base is determined automatically using the following rules: If the string begins with "0x", it is assumed to be hexadecimal; if it begins with "0", it is assumed to be octal; otherwise it is assumed to be decimal. If the conversion fails the function returns 0. If \a ok is not 0, failure is reported by setting *ok to false, and success by setting *ok to true. This function ignores leading and trailing whitespace. \sa toUShort(), toString() */ short QLocale::toShort(const QString &s, bool *ok, int base) const { qlonglong i = toLongLong(s, ok, base); if (i < SHRT_MIN || i > SHRT_MAX) { if (ok != 0) *ok = false; return 0; } return short(i); } /*! Returns the unsigned short int represented by the localized string \a s, using base \a base. If \a base is 0 the base is determined automatically using the following rules: If the string begins with "0x", it is assumed to be hexadecimal; if it begins with "0", it is assumed to be octal; otherwise it is assumed to be decimal. If the conversion fails the function returns 0. If \a ok is not 0, failure is reported by setting *ok to false, and success by setting *ok to true. This function ignores leading and trailing whitespace. \sa toShort(), toString() */ ushort QLocale::toUShort(const QString &s, bool *ok, int base) const { qulonglong i = toULongLong(s, ok, base); if (i > USHRT_MAX) { if (ok != 0) *ok = false; return 0; } return ushort(i); } /*! Returns the int represented by the localized string \a s, using base \a base. If \a base is 0 the base is determined automatically using the following rules: If the string begins with "0x", it is assumed to be hexadecimal; if it begins with "0", it is assumed to be octal; otherwise it is assumed to be decimal. If the conversion fails the function returns 0. If \a ok is not 0, failure is reported by setting *ok to false, and success by setting *ok to true. This function ignores leading and trailing whitespace. \sa toUInt(), toString() */ int QLocale::toInt(const QString &s, bool *ok, int base) const { qlonglong i = toLongLong(s, ok, base); if (i < INT_MIN || i > INT_MAX) { if (ok != 0) *ok = false; return 0; } return int(i); } /*! Returns the unsigned int represented by the localized string \a s, using base \a base. If \a base is 0 the base is determined automatically using the following rules: If the string begins with "0x", it is assumed to be hexadecimal; if it begins with "0", it is assumed to be octal; otherwise it is assumed to be decimal. If the conversion fails the function returns 0. If \a ok is not 0, failure is reported by setting *ok to false, and success by setting *ok to true. This function ignores leading and trailing whitespace. \sa toInt(), toString() */ uint QLocale::toUInt(const QString &s, bool *ok, int base) const { qulonglong i = toULongLong(s, ok, base); if (i > UINT_MAX) { if (ok != 0) *ok = false; return 0; } return uint(i); } /*! Returns the long long int represented by the localized string \a s, using base \a base. If \a base is 0 the base is determined automatically using the following rules: If the string begins with "0x", it is assumed to be hexadecimal; if it begins with "0", it is assumed to be octal; otherwise it is assumed to be decimal. If the conversion fails the function returns 0. If \a ok is not 0, failure is reported by setting *ok to false, and success by setting *ok to true. This function ignores leading and trailing whitespace. \sa toInt(), toULongLong(), toDouble(), toString() */ qlonglong QLocale::toLongLong(const QString &s, bool *ok, int base) const { QLocalePrivate::GroupSeparatorMode mode = p.numberOptions & RejectGroupSeparator ? QLocalePrivate::FailOnGroupSeparators : QLocalePrivate::ParseGroupSeparators; return d()->stringToLongLong(s, base, ok, mode); } // ### Qt5: make the return type for toULongLong() qulonglong. /*! Returns the unsigned long long int represented by the localized string \a s, using base \a base. If \a base is 0 the base is determined automatically using the following rules: If the string begins with "0x", it is assumed to be hexadecimal; if it begins with "0", it is assumed to be octal; otherwise it is assumed to be decimal. If the conversion fails the function returns 0. If \a ok is not 0, failure is reported by setting *ok to false, and success by setting *ok to true. This function ignores leading and trailing whitespace. \sa toLongLong(), toInt(), toDouble(), toString() */ qlonglong QLocale::toULongLong(const QString &s, bool *ok, int base) const { QLocalePrivate::GroupSeparatorMode mode = p.numberOptions & RejectGroupSeparator ? QLocalePrivate::FailOnGroupSeparators : QLocalePrivate::ParseGroupSeparators; return d()->stringToUnsLongLong(s, base, ok, mode); } /*! Returns the float represented by the localized string \a s, or 0.0 if the conversion failed. If \a ok is not 0, reports failure by setting *ok to false and success by setting *ok to true. This function ignores leading and trailing whitespace. \sa toDouble(), toInt(), toString() */ #define QT_MAX_FLOAT 3.4028234663852886e+38 float QLocale::toFloat(const QString &s, bool *ok) const { bool myOk; double d = toDouble(s, &myOk); if (!myOk || d > QT_MAX_FLOAT || d < -QT_MAX_FLOAT) { if (ok != 0) *ok = false; return 0.0; } if (ok != 0) *ok = true; return float(d); } /*! Returns the double represented by the localized string \a s, or 0.0 if the conversion failed. If \a ok is not 0, reports failure by setting *ok to false and success by setting *ok to true. Unlike QString::toDouble(), this function does not fall back to the "C" locale if the string cannot be interpreted in this locale. \snippet doc/src/snippets/code/src_corelib_tools_qlocale.cpp 3 Notice that the last conversion returns 1234.0, because '.' is the thousands group separator in the German locale. This function ignores leading and trailing whitespace. \sa toFloat(), toInt(), toString() */ double QLocale::toDouble(const QString &s, bool *ok) const { QLocalePrivate::GroupSeparatorMode mode = p.numberOptions & RejectGroupSeparator ? QLocalePrivate::FailOnGroupSeparators : QLocalePrivate::ParseGroupSeparators; return d()->stringToDouble(s, ok, mode); } /*! Returns a localized string representation of \a i. \sa toLongLong() */ QString QLocale::toString(qlonglong i) const { int flags = p.numberOptions & OmitGroupSeparator ? 0 : QLocalePrivate::ThousandsGroup; return d()->longLongToString(i, -1, 10, -1, flags); } /*! \overload \sa toULongLong() */ QString QLocale::toString(qulonglong i) const { int flags = p.numberOptions & OmitGroupSeparator ? 0 : QLocalePrivate::ThousandsGroup; return d()->unsLongLongToString(i, -1, 10, -1, flags); } /*! Returns a localized string representation of the given \a date in the specified \a format. If \a format is an empty string, an empty string is returned. */ QString QLocale::toString(const QDate &date, const QString &format) const { return d()->dateTimeToString(format, &date, 0, this); } /*! Returns a localized string representation of the given \a date according to the specified \a format. */ QString QLocale::toString(const QDate &date, FormatType format) const { if (!date.isValid()) return QString(); #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(format == LongFormat ? QSystemLocale::DateToStringLong : QSystemLocale::DateToStringShort, date); if (!res.isNull()) return res.toString(); } #endif QString format_str = dateFormat(format); return toString(date, format_str); } static bool timeFormatContainsAP(const QString &format) { int i = 0; while (i < format.size()) { if (format.at(i).unicode() == '\'') { readEscapedFormatString(format, &i); continue; } if (format.at(i).toLower().unicode() == 'a') return true; ++i; } return false; } static QString timeZone() { #if defined(Q_OS_WINCE) TIME_ZONE_INFORMATION info; DWORD res = GetTimeZoneInformation(&info); if (res == TIME_ZONE_ID_UNKNOWN) return QString(); return QString::fromUtf16(reinterpret_cast (info.StandardName)); #elif defined(Q_OS_WIN) _tzset(); # if defined(_MSC_VER) && _MSC_VER >= 1400 size_t returnSize = 0; char timeZoneName[512]; if (_get_tzname(&returnSize, timeZoneName, 512, 1)) return QString(); return QString::fromLocal8Bit(timeZoneName); # else return QString::fromLocal8Bit(_tzname[1]); # endif #else tzset(); return QString::fromLocal8Bit(tzname[1]); #endif } /*! Returns a localized string representation of the given \a time according to the specified \a format. If \a format is an empty string, an empty string is returned. */ QString QLocale::toString(const QTime &time, const QString &format) const { return d()->dateTimeToString(format, 0, &time, this); } /*! \since 4.4 Returns a localized string representation of the given \a dateTime according to the specified \a format. If \a format is an empty string, an empty string is returned. */ QString QLocale::toString(const QDateTime &dateTime, const QString &format) const { const QDate dt = dateTime.date(); const QTime tm = dateTime.time(); return d()->dateTimeToString(format, &dt, &tm, this); } /*! \since 4.4 Returns a localized string representation of the given \a dateTime according to the specified \a format. */ QString QLocale::toString(const QDateTime &dateTime, FormatType format) const { if (!dateTime.isValid()) return QString(); #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(format == LongFormat ? QSystemLocale::DateTimeToStringLong : QSystemLocale::DateTimeToStringShort, dateTime); if (!res.isNull()) return res.toString(); } #endif const QString format_str = dateTimeFormat(format); return toString(dateTime, format_str); } /*! Returns a localized string representation of the given \a time in the specified \a format. */ QString QLocale::toString(const QTime &time, FormatType format) const { if (!time.isValid()) return QString(); #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(format == LongFormat ? QSystemLocale::TimeToStringLong : QSystemLocale::TimeToStringShort, time); if (!res.isNull()) return res.toString(); } #endif QString format_str = timeFormat(format); return toString(time, format_str); } /*! \since 4.1 Returns the date format used for the current locale. If \a format is LongFormat the format will be a long version. Otherwise it uses a shorter version. \sa QDate::toString(), QDate::fromString() */ QString QLocale::dateFormat(FormatType format) const { #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(format == LongFormat ? QSystemLocale::DateFormatLong : QSystemLocale::DateFormatShort, QVariant()); if (!res.isNull()) return res.toString(); } #endif quint32 idx, size; switch (format) { case LongFormat: idx = d()->m_long_date_format_idx; size = d()->m_long_date_format_size; break; default: idx = d()->m_short_date_format_idx; size = d()->m_short_date_format_size; break; } return getLocaleData(date_format_data + idx, size); } /*! \since 4.1 Returns the time format used for the current locale. If \a format is LongFormat the format will be a long version. Otherwise it uses a shorter version. \sa QTime::toString(), QTime::fromString() */ QString QLocale::timeFormat(FormatType format) const { #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(format == LongFormat ? QSystemLocale::TimeFormatLong : QSystemLocale::TimeFormatShort, QVariant()); if (!res.isNull()) return res.toString(); } #endif quint32 idx, size; switch (format) { case LongFormat: idx = d()->m_long_time_format_idx; size = d()->m_long_time_format_size; break; default: idx = d()->m_short_time_format_idx; size = d()->m_short_time_format_size; break; } return getLocaleData(time_format_data + idx, size); } /*! \since 4.4 Returns the date time format used for the current locale. If \a format is ShortFormat the format will be a short version. Otherwise it uses a longer version. \sa QDateTime::toString(), QDateTime::fromString() */ QString QLocale::dateTimeFormat(FormatType format) const { #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(format == LongFormat ? QSystemLocale::DateTimeFormatLong : QSystemLocale::DateTimeFormatShort, QVariant()); if (!res.isNull()) { return res.toString(); } } #endif return dateFormat(format) + QLatin1Char(' ') + timeFormat(format); } /*! \since 4.4 Parses the time string given in \a string and returns the time. The format of the time string is chosen according to the \a format parameter (see timeFormat()). If the time could not be parsed, returns an invalid time. \sa timeFormat(), toDate(), toDateTime(), QTime::fromString() */ #ifndef QT_NO_DATESTRING QTime QLocale::toTime(const QString &string, FormatType format) const { return toTime(string, timeFormat(format)); } #endif /*! \since 4.4 Parses the date string given in \a string and returns the date. The format of the date string is chosen according to the \a format parameter (see dateFormat()). If the date could not be parsed, returns an invalid date. \sa dateFormat(), toTime(), toDateTime(), QDate::fromString() */ #ifndef QT_NO_DATESTRING QDate QLocale::toDate(const QString &string, FormatType format) const { return toDate(string, dateFormat(format)); } #endif /*! \since 4.4 Parses the date/time string given in \a string and returns the time. The format of the date/time string is chosen according to the \a format parameter (see dateTimeFormat()). If the string could not be parsed, returns an invalid QDateTime. \sa dateTimeFormat(), toTime(), toDate(), QDateTime::fromString() */ #ifndef QT_NO_DATESTRING QDateTime QLocale::toDateTime(const QString &string, FormatType format) const { return toDateTime(string, dateFormat(format)); } #endif /*! \since 4.4 Parses the time string given in \a string and returns the time. See QTime::fromString() for information on what is a valid format string. If the time could not be parsed, returns an invalid time. \sa timeFormat(), toDate(), toDateTime(), QTime::fromString() */ #ifndef QT_NO_DATESTRING QTime QLocale::toTime(const QString &string, const QString &format) const { QTime time; #ifndef QT_BOOTSTRAPPED QDateTimeParser dt(QVariant::Time, QDateTimeParser::FromString); dt.defaultLocale = *this; if (dt.parseFormat(format)) dt.fromString(string, 0, &time); #else Q_UNUSED(string); Q_UNUSED(format); #endif return time; } #endif /*! \since 4.4 Parses the date string given in \a string and returns the date. See QDate::fromString() for information on the expressions that can be used with this function. This function searches month names and the names of the days of the week in the current locale. If the date could not be parsed, returns an invalid date. \sa dateFormat(), toTime(), toDateTime(), QDate::fromString() */ #ifndef QT_NO_DATESTRING QDate QLocale::toDate(const QString &string, const QString &format) const { QDate date; #ifndef QT_BOOTSTRAPPED QDateTimeParser dt(QVariant::Date, QDateTimeParser::FromString); dt.defaultLocale = *this; if (dt.parseFormat(format)) dt.fromString(string, &date, 0); #else Q_UNUSED(string); Q_UNUSED(format); #endif return date; } #endif /*! \since 4.4 Parses the date/time string given in \a string and returns the time. See QDateTime::fromString() for information on the expressions that can be used with this function. \note The month and day names used must be given in the user's local language. If the string could not be parsed, returns an invalid QDateTime. \sa dateTimeFormat(), toTime(), toDate(), QDateTime::fromString() */ #ifndef QT_NO_DATESTRING QDateTime QLocale::toDateTime(const QString &string, const QString &format) const { #ifndef QT_BOOTSTRAPPED QTime time; QDate date; QDateTimeParser dt(QVariant::DateTime, QDateTimeParser::FromString); dt.defaultLocale = *this; if (dt.parseFormat(format) && dt.fromString(string, &date, &time)) return QDateTime(date, time); #else Q_UNUSED(string); Q_UNUSED(format); #endif return QDateTime(QDate(), QTime(-1, -1, -1)); } #endif /*! \since 4.1 Returns the decimal point character of this locale. */ QChar QLocale::decimalPoint() const { return d()->decimal(); } /*! \since 4.1 Returns the group separator character of this locale. */ QChar QLocale::groupSeparator() const { return d()->group(); } /*! \since 4.1 Returns the percent character of this locale. */ QChar QLocale::percent() const { return d()->percent(); } /*! \since 4.1 Returns the zero digit character of this locale. */ QChar QLocale::zeroDigit() const { return d()->zero(); } /*! \since 4.1 Returns the negative sign character of this locale. */ QChar QLocale::negativeSign() const { return d()->minus(); } /*! \since 4.5 Returns the positive sign character of this locale. */ QChar QLocale::positiveSign() const { return d()->plus(); } /*! \since 4.1 Returns the exponential character of this locale. */ QChar QLocale::exponential() const { return d()->exponential(); } static bool qIsUpper(char c) { return c >= 'A' && c <= 'Z'; } static char qToLower(char c) { if (c >= 'A' && c <= 'Z') return c - 'A' + 'a'; else return c; } /*! \overload \a f and \a prec have the same meaning as in QString::number(double, char, int). \sa toDouble() */ QString QLocale::toString(double i, char f, int prec) const { QLocalePrivate::DoubleForm form = QLocalePrivate::DFDecimal; uint flags = 0; if (qIsUpper(f)) flags = QLocalePrivate::CapitalEorX; f = qToLower(f); switch (f) { case 'f': form = QLocalePrivate::DFDecimal; break; case 'e': form = QLocalePrivate::DFExponent; break; case 'g': form = QLocalePrivate::DFSignificantDigits; break; default: break; } if (!(p.numberOptions & OmitGroupSeparator)) flags |= QLocalePrivate::ThousandsGroup; return d()->doubleToString(i, prec, form, -1, flags); } /*! \fn QLocale QLocale::c() Returns a QLocale object initialized to the "C" locale. \sa system() */ /*! Returns a QLocale object initialized to the system locale. On Windows and Mac, this locale will use the decimal/grouping characters and date/time formats specified in the system configuration panel. \sa QTextCodec::locale() c() */ QLocale QLocale::system() { QLocale result(C); result.p.index = localePrivateIndex(systemPrivate()); return result; } /*! \since 4.3 Returns the list of countries that have entires for \a language in Qt's locale database. If the result is an empty list, then \a language is not represented in Qt's locale database. */ QList QLocale::countriesForLanguage(Language language) { QList result; unsigned language_id = language; uint idx = locale_index[language_id]; if (language == C) { result << AnyCountry; return result; } const QLocalePrivate *d = locale_data + idx; while (d->languageId() == language_id) { result << static_cast(d->countryId()); ++d; } return result; } /*! \since 4.2 Returns the localized name of \a month, in the format specified by \a type. \sa dayName(), standaloneMonthName() */ QString QLocale::monthName(int month, FormatType type) const { if (month < 1 || month > 12) return QString(); #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(type == LongFormat ? QSystemLocale::MonthNameLong : QSystemLocale::MonthNameShort, month); if (!res.isNull()) return res.toString(); } #endif quint32 idx, size; switch (type) { case QLocale::LongFormat: idx = d()->m_long_month_names_idx; size = d()->m_long_month_names_size; break; case QLocale::ShortFormat: idx = d()->m_short_month_names_idx; size = d()->m_short_month_names_size; break; case QLocale::NarrowFormat: idx = d()->m_narrow_month_names_idx; size = d()->m_narrow_month_names_size; break; default: return QString(); } return getLocaleListData(months_data + idx, size, month - 1); } /*! \since 4.5 Returns the localized name of \a month that is used as a standalone text, in the format specified by \a type. If the locale information doesn't specify the standalone month name then return value is the same as in monthName(). \sa monthName(), standaloneDayName() */ QString QLocale::standaloneMonthName(int month, FormatType type) const { if (month < 1 || month > 12) return QString(); #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(type == LongFormat ? QSystemLocale::MonthNameLong : QSystemLocale::MonthNameShort, month); if (!res.isNull()) return res.toString(); } #endif quint32 idx, size; switch (type) { case QLocale::LongFormat: idx = d()->m_standalone_long_month_names_idx; size = d()->m_standalone_long_month_names_size; break; case QLocale::ShortFormat: idx = d()->m_standalone_short_month_names_idx; size = d()->m_standalone_short_month_names_size; break; case QLocale::NarrowFormat: idx = d()->m_standalone_narrow_month_names_idx; size = d()->m_standalone_narrow_month_names_size; break; default: return QString(); } QString name = getLocaleListData(standalone_months_data + idx, size, month - 1); if (name.isEmpty()) return monthName(month, type); return name; } /*! \since 4.2 Returns the localized name of the \a day (where 1 represents Monday, 2 represents Tuesday and so on), in the format specified by \a type. \sa monthName(), standaloneDayName() */ QString QLocale::dayName(int day, FormatType type) const { if (day < 1 || day > 7) return QString(); #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(type == LongFormat ? QSystemLocale::DayNameLong : QSystemLocale::DayNameShort, day); if (!res.isNull()) return res.toString(); } #endif if (day == 7) day = 0; quint32 idx, size; switch (type) { case QLocale::LongFormat: idx = d()->m_long_day_names_idx; size = d()->m_long_day_names_size; break; case QLocale::ShortFormat: idx = d()->m_short_day_names_idx; size = d()->m_short_day_names_size; break; case QLocale::NarrowFormat: idx = d()->m_narrow_day_names_idx; size = d()->m_narrow_day_names_size; break; default: return QString(); } return getLocaleListData(days_data + idx, size, day); } /*! \since 4.5 Returns the localized name of the \a day (where 1 represents Monday, 2 represents Tuesday and so on) that is used as a standalone text, in the format specified by \a type. If the locale information does not specify the standalone day name then return value is the same as in dayName(). \sa dayName(), standaloneMonthName() */ QString QLocale::standaloneDayName(int day, FormatType type) const { if (day < 1 || day > 7) return QString(); #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(type == LongFormat ? QSystemLocale::DayNameLong : QSystemLocale::DayNameShort, day); if (!res.isNull()) return res.toString(); } #endif if (day == 7) day = 0; quint32 idx, size; switch (type) { case QLocale::LongFormat: idx = d()->m_standalone_long_day_names_idx; size = d()->m_standalone_long_day_names_size; break; case QLocale::ShortFormat: idx = d()->m_standalone_short_day_names_idx; size = d()->m_standalone_short_day_names_size; break; case QLocale::NarrowFormat: idx = d()->m_standalone_narrow_day_names_idx; size = d()->m_standalone_narrow_day_names_size; break; default: return QString(); } QString name = getLocaleListData(days_data + idx, size, day); if (name.isEmpty()) return dayName(day == 0 ? 7 : day, type); return name; } /*! \since 4.4 Returns the measurement system for the locale. */ QLocale::MeasurementSystem QLocale::measurementSystem() const { MeasurementSystem meas = MetricSystem; bool found = false; #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(QSystemLocale::MeasurementSystem, QVariant()); if (!res.isNull()) { meas = MeasurementSystem(res.toInt()); found = true; } } #endif if (!found) { meas = d()->measurementSystem(); found = true; } return meas; } /*! \since 4.5 Returns the localized name of the "AM" suffix for times specified using the conventions of the 12-hour clock. \sa pmText() */ QString QLocale::amText() const { #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(QSystemLocale::AMText, QVariant()); if (!res.isNull()) return res.toString(); } #endif return getLocaleData(am_data + d()->m_am_idx, d()->m_am_size); } /*! \since 4.5 Returns the localized name of the "PM" suffix for times specified using the conventions of the 12-hour clock. \sa amText() */ QString QLocale::pmText() const { #ifndef QT_NO_SYSTEMLOCALE if (d() == systemPrivate()) { QVariant res = systemLocale()->query(QSystemLocale::PMText, QVariant()); if (!res.isNull()) return res.toString(); } #endif return getLocaleData(pm_data + d()->m_pm_idx, d()->m_pm_size); } /*! \fn QString QLocale::toString(short i) const \overload \sa toShort() */ /*! \fn QString QLocale::toString(ushort i) const \overload \sa toUShort() */ /*! \fn QString QLocale::toString(int i) const \overload \sa toInt() */ /*! \fn QString QLocale::toString(uint i) const \overload \sa toUInt() */ /* \fn QString QLocale::toString(long i) const \overload \sa toLong() */ /* \fn QString QLocale::toString(ulong i) const \overload \sa toULong() */ /*! \fn QString QLocale::toString(float i, char f = 'g', int prec = 6) const \overload \a f and \a prec have the same meaning as in QString::number(double, char, int). \sa toDouble() */ static QString qulltoa(qulonglong l, int base, const QLocalePrivate &locale) { ushort buff[65]; // length of MAX_ULLONG in base 2 ushort *p = buff + 65; const QChar _zero = locale.zero(); if (base != 10 || _zero.unicode() == '0') { while (l != 0) { int c = l % base; --p; if (c < 10) *p = '0' + c; else *p = c - 10 + 'a'; l /= base; } } else { while (l != 0) { int c = l % base; *(--p) = _zero.unicode() + c; l /= base; } } return QString(reinterpret_cast(p), 65 - (p - buff)); } static QString qlltoa(qlonglong l, int base, const QLocalePrivate &locale) { return qulltoa(l < 0 ? -l : l, base, locale); } enum PrecisionMode { PMDecimalDigits = 0x01, PMSignificantDigits = 0x02, PMChopTrailingZeros = 0x03 }; static QString &decimalForm(QString &digits, int decpt, uint precision, PrecisionMode pm, bool always_show_decpt, bool thousands_group, const QLocalePrivate &locale) { if (decpt < 0) { for (int i = 0; i < -decpt; ++i) digits.prepend(locale.zero()); decpt = 0; } else if (decpt > digits.length()) { for (int i = digits.length(); i < decpt; ++i) digits.append(locale.zero()); } if (pm == PMDecimalDigits) { uint decimal_digits = digits.length() - decpt; for (uint i = decimal_digits; i < precision; ++i) digits.append(locale.zero()); } else if (pm == PMSignificantDigits) { for (uint i = digits.length(); i < precision; ++i) digits.append(locale.zero()); } else { // pm == PMChopTrailingZeros } if (always_show_decpt || decpt < digits.length()) digits.insert(decpt, locale.decimal()); if (thousands_group) { for (int i = decpt - 3; i > 0; i -= 3) digits.insert(i, locale.group()); } if (decpt == 0) digits.prepend(locale.zero()); return digits; } static QString &exponentForm(QString &digits, int decpt, uint precision, PrecisionMode pm, bool always_show_decpt, const QLocalePrivate &locale) { int exp = decpt - 1; if (pm == PMDecimalDigits) { for (uint i = digits.length(); i < precision + 1; ++i) digits.append(locale.zero()); } else if (pm == PMSignificantDigits) { for (uint i = digits.length(); i < precision; ++i) digits.append(locale.zero()); } else { // pm == PMChopTrailingZeros } if (always_show_decpt || digits.length() > 1) digits.insert(1, locale.decimal()); digits.append(locale.exponential()); digits.append(locale.longLongToString(exp, 2, 10, -1, QLocalePrivate::AlwaysShowSign)); return digits; } static bool isZero(double d) { uchar *ch = (uchar *)&d; #ifdef QT_ARMFPA return !(ch[3] & 0x7F || ch[2] || ch[1] || ch[0] || ch[7] || ch[6] || ch[5] || ch[4]); #else if (QSysInfo::ByteOrder == QSysInfo::BigEndian) { return !(ch[0] & 0x7F || ch[1] || ch[2] || ch[3] || ch[4] || ch[5] || ch[6] || ch[7]); } else { return !(ch[7] & 0x7F || ch[6] || ch[5] || ch[4] || ch[3] || ch[2] || ch[1] || ch[0]); } #endif } QString QLocalePrivate::dateTimeToString(const QString &format, const QDate *date, const QTime *time, const QLocale *q) const { Q_ASSERT(date || time); if ((date && !date->isValid()) || (time && !time->isValid())) return QString(); const bool format_am_pm = time && timeFormatContainsAP(format); enum { AM, PM } am_pm = AM; int hour12 = time ? time->hour() : -1; if (time) { if (hour12 == 0) { am_pm = AM; hour12 = 12; } else if (hour12 < 12) { am_pm = AM; } else if (hour12 == 12) { am_pm = PM; } else { am_pm = PM; hour12 -= 12; } } QString result; int i = 0; while (i < format.size()) { if (format.at(i).unicode() == '\'') { result.append(readEscapedFormatString(format, &i)); continue; } const QChar c = format.at(i); int repeat = repeatCount(format, i); bool used = false; if (date) { switch (c.unicode()) { case 'y': used = true; if (repeat >= 4) repeat = 4; else if (repeat >= 2) repeat = 2; switch (repeat) { case 4: result.append(longLongToString(date->year())); break; case 2: result.append(longLongToString(date->year() % 100, -1, 10, 2, QLocalePrivate::ZeroPadded)); break; default: repeat = 1; result.append(c); break; } break; case 'M': used = true; repeat = qMin(repeat, 4); switch (repeat) { case 1: result.append(longLongToString(date->month())); break; case 2: result.append(longLongToString(date->month(), -1, 10, 2, QLocalePrivate::ZeroPadded)); break; case 3: result.append(q->monthName(date->month(), QLocale::ShortFormat)); break; case 4: result.append(q->monthName(date->month(), QLocale::LongFormat)); break; } break; case 'd': used = true; repeat = qMin(repeat, 4); switch (repeat) { case 1: result.append(longLongToString(date->day())); break; case 2: result.append(longLongToString(date->day(), -1, 10, 2, QLocalePrivate::ZeroPadded)); break; case 3: result.append(q->dayName(date->dayOfWeek(), QLocale::ShortFormat)); break; case 4: result.append(q->dayName(date->dayOfWeek(), QLocale::LongFormat)); break; } break; default: break; } } if (!used && time) { switch (c.unicode()) { case 'h': { used = true; repeat = qMin(repeat, 2); const int hour = format_am_pm ? hour12 : time->hour(); switch (repeat) { case 1: result.append(longLongToString(hour)); break; case 2: result.append(longLongToString(hour, -1, 10, 2, QLocalePrivate::ZeroPadded)); break; } break; } case 'H': used = true; repeat = qMin(repeat, 2); switch (repeat) { case 1: result.append(longLongToString(time->hour())); break; case 2: result.append(longLongToString(time->hour(), -1, 10, 2, QLocalePrivate::ZeroPadded)); break; } break; case 'm': used = true; repeat = qMin(repeat, 2); switch (repeat) { case 1: result.append(longLongToString(time->minute())); break; case 2: result.append(longLongToString(time->minute(), -1, 10, 2, QLocalePrivate::ZeroPadded)); break; } break; case 's': used = true; repeat = qMin(repeat, 2); switch (repeat) { case 1: result.append(longLongToString(time->second())); break; case 2: result.append(longLongToString(time->second(), -1, 10, 2, QLocalePrivate::ZeroPadded)); break; } break; case 'a': used = true; if (i + 1 < format.length() && format.at(i + 1).unicode() == 'p') { repeat = 2; } else { repeat = 1; } result.append(am_pm == AM ? QLatin1String("am") : QLatin1String("pm")); break; case 'A': used = true; if (i + 1 < format.length() && format.at(i + 1).unicode() == 'P') { repeat = 2; } else { repeat = 1; } result.append(am_pm == AM ? QLatin1String("AM") : QLatin1String("PM")); break; case 'z': used = true; if (repeat >= 3) { repeat = 3; } else { repeat = 1; } switch (repeat) { case 1: result.append(longLongToString(time->msec())); break; case 3: result.append(longLongToString(time->msec(), -1, 10, 3, QLocalePrivate::ZeroPadded)); break; } break; case 't': used = true; repeat = 1; result.append(timeZone()); break; default: break; } } if (!used) { result.append(QString(repeat, c)); } i += repeat; } return result; } QString QLocalePrivate::doubleToString(double d, int precision, DoubleForm form, int width, unsigned flags) const { if (precision == -1) precision = 6; if (width == -1) width = 0; bool negative = false; bool special_number = false; // nan, +/-inf QString num_str; // Detect special numbers (nan, +/-inf) if (qt_is_inf(d)) { num_str = QString::fromLatin1("inf"); special_number = true; negative = d < 0; } else if (qt_is_nan(d)) { num_str = QString::fromLatin1("nan"); special_number = true; } // Handle normal numbers if (!special_number) { int decpt, sign; QString digits; #ifdef QT_QLOCALE_USES_FCVT // NOT thread safe! if (form == DFDecimal) { digits = QLatin1String(fcvt(d, precision, &decpt, &sign)); } else { int pr = precision; if (form == DFExponent) ++pr; else if (form == DFSignificantDigits && pr == 0) pr = 1; digits = QLatin1String(ecvt(d, pr, &decpt, &sign)); // Chop trailing zeros if (digits.length() > 0) { int last_nonzero_idx = digits.length() - 1; while (last_nonzero_idx > 0 && digits.unicode()[last_nonzero_idx] == QLatin1Char('0')) --last_nonzero_idx; digits.truncate(last_nonzero_idx + 1); } } #else int mode; if (form == DFDecimal) mode = 3; else mode = 2; /* This next bit is a bit quirky. In DFExponent form, the precision is the number of digits after decpt. So that would suggest using mode=3 for qdtoa. But qdtoa behaves strangely when mode=3 and precision=0. So we get around this by using mode=2 and reasoning that we want precision+1 significant digits, since the decimal point in this mode is always after the first digit. */ int pr = precision; if (form == DFExponent) ++pr; char *rve = 0; char *buff = 0; digits = QLatin1String(qdtoa(d, mode, pr, &decpt, &sign, &rve, &buff)); if (buff != 0) free(buff); #endif // QT_QLOCALE_USES_FCVT const QChar _zero = zero(); if (_zero.unicode() != '0') { ushort z = _zero.unicode() - '0'; for (int i = 0; i < digits.length(); ++i) reinterpret_cast(digits.data())[i] += z; } bool always_show_decpt = (flags & Alternate || flags & ForcePoint); switch (form) { case DFExponent: { num_str = exponentForm(digits, decpt, precision, PMDecimalDigits, always_show_decpt, *this); break; } case DFDecimal: { num_str = decimalForm(digits, decpt, precision, PMDecimalDigits, always_show_decpt, flags & ThousandsGroup, *this); break; } case DFSignificantDigits: { PrecisionMode mode = (flags & Alternate) ? PMSignificantDigits : PMChopTrailingZeros; if (decpt != digits.length() && (decpt <= -4 || decpt > precision)) num_str = exponentForm(digits, decpt, precision, mode, always_show_decpt, *this); else num_str = decimalForm(digits, decpt, precision, mode, always_show_decpt, flags & ThousandsGroup, *this); break; } } negative = sign != 0 && !isZero(d); } // pad with zeros. LeftAdjusted overrides this flag). Also, we don't // pad special numbers if (flags & QLocalePrivate::ZeroPadded && !(flags & QLocalePrivate::LeftAdjusted) && !special_number) { int num_pad_chars = width - num_str.length(); // leave space for the sign if (negative || flags & QLocalePrivate::AlwaysShowSign || flags & QLocalePrivate::BlankBeforePositive) --num_pad_chars; for (int i = 0; i < num_pad_chars; ++i) num_str.prepend(zero()); } // add sign if (negative) num_str.prepend(minus()); else if (flags & QLocalePrivate::AlwaysShowSign) num_str.prepend(plus()); else if (flags & QLocalePrivate::BlankBeforePositive) num_str.prepend(QLatin1Char(' ')); if (flags & QLocalePrivate::CapitalEorX) num_str = num_str.toUpper(); return num_str; } QString QLocalePrivate::longLongToString(qlonglong l, int precision, int base, int width, unsigned flags) const { bool precision_not_specified = false; if (precision == -1) { precision_not_specified = true; precision = 1; } bool negative = l < 0; if (base != 10) { // these are not supported by sprintf for octal and hex flags &= ~AlwaysShowSign; flags &= ~BlankBeforePositive; negative = false; // neither are negative numbers } QString num_str; if (base == 10) num_str = qlltoa(l, base, *this); else num_str = qulltoa(l, base, *this); uint cnt_thousand_sep = 0; if (flags & ThousandsGroup && base == 10) { for (int i = num_str.length() - 3; i > 0; i -= 3) { num_str.insert(i, group()); ++cnt_thousand_sep; } } for (int i = num_str.length()/* - cnt_thousand_sep*/; i < precision; ++i) num_str.prepend(base == 10 ? zero() : QChar::fromLatin1('0')); if ((flags & Alternate || flags & ShowBase) && base == 8 && (num_str.isEmpty() || num_str[0].unicode() != QLatin1Char('0'))) num_str.prepend(QLatin1Char('0')); // LeftAdjusted overrides this flag ZeroPadded. sprintf only padds // when precision is not specified in the format string bool zero_padded = flags & ZeroPadded && !(flags & LeftAdjusted) && precision_not_specified; if (zero_padded) { int num_pad_chars = width - num_str.length(); // leave space for the sign if (negative || flags & AlwaysShowSign || flags & BlankBeforePositive) --num_pad_chars; // leave space for optional '0x' in hex form if (base == 16 && (flags & Alternate || flags & ShowBase)) num_pad_chars -= 2; // leave space for optional '0b' in binary form else if (base == 2 && (flags & Alternate || flags & ShowBase)) num_pad_chars -= 2; for (int i = 0; i < num_pad_chars; ++i) num_str.prepend(base == 10 ? zero() : QChar::fromLatin1('0')); } if (flags & CapitalEorX) num_str = num_str.toUpper(); if (base == 16 && (flags & Alternate || flags & ShowBase)) num_str.prepend(QLatin1String(flags & UppercaseBase ? "0X" : "0x")); if (base == 2 && (flags & Alternate || flags & ShowBase)) num_str.prepend(QLatin1String(flags & UppercaseBase ? "0B" : "0b")); // add sign if (negative) num_str.prepend(minus()); else if (flags & AlwaysShowSign) num_str.prepend(plus()); else if (flags & BlankBeforePositive) num_str.prepend(QLatin1Char(' ')); return num_str; } QString QLocalePrivate::unsLongLongToString(qulonglong l, int precision, int base, int width, unsigned flags) const { bool precision_not_specified = false; if (precision == -1) { precision_not_specified = true; precision = 1; } QString num_str = qulltoa(l, base, *this); uint cnt_thousand_sep = 0; if (flags & ThousandsGroup && base == 10) { for (int i = num_str.length() - 3; i > 0; i -=3) { num_str.insert(i, group()); ++cnt_thousand_sep; } } for (int i = num_str.length()/* - cnt_thousand_sep*/; i < precision; ++i) num_str.prepend(base == 10 ? zero() : QChar::fromLatin1('0')); if ((flags & Alternate || flags & ShowBase) && base == 8 && (num_str.isEmpty() || num_str[0].unicode() != QLatin1Char('0'))) num_str.prepend(QLatin1Char('0')); // LeftAdjusted overrides this flag ZeroPadded. sprintf only padds // when precision is not specified in the format string bool zero_padded = flags & ZeroPadded && !(flags & LeftAdjusted) && precision_not_specified; if (zero_padded) { int num_pad_chars = width - num_str.length(); // leave space for optional '0x' in hex form if (base == 16 && flags & Alternate) num_pad_chars -= 2; // leave space for optional '0b' in binary form else if (base == 2 && flags & Alternate) num_pad_chars -= 2; for (int i = 0; i < num_pad_chars; ++i) num_str.prepend(base == 10 ? zero() : QChar::fromLatin1('0')); } if (flags & CapitalEorX) num_str = num_str.toUpper(); if (base == 16 && (flags & Alternate || flags & ShowBase)) num_str.prepend(QLatin1String(flags & UppercaseBase ? "0X" : "0x")); else if (base == 2 && (flags & Alternate || flags & ShowBase)) num_str.prepend(QLatin1String(flags & UppercaseBase ? "0B" : "0b")); // add sign if (flags & AlwaysShowSign) num_str.prepend(plus()); else if (flags & BlankBeforePositive) num_str.prepend(QLatin1Char(' ')); return num_str; } // Removes thousand-group separators in "C" locale. static bool removeGroupSeparators(QLocalePrivate::CharBuff *num) { int group_cnt = 0; // counts number of group chars int decpt_idx = -1; char *data = num->data(); int l = qstrlen(data); // Find the decimal point and check if there are any group chars int i = 0; for (; i < l; ++i) { char c = data[i]; if (c == ',') { if (i == 0 || data[i - 1] < '0' || data[i - 1] > '9') return false; if (i == l - 1 || data[i + 1] < '0' || data[i + 1] > '9') return false; ++group_cnt; } else if (c == '.') { // Fail if more than one decimal points if (decpt_idx != -1) return false; decpt_idx = i; } else if (c == 'e' || c == 'E') { // an 'e' or 'E' - if we have not encountered a decimal // point, this is where it "is". if (decpt_idx == -1) decpt_idx = i; } } // If no group chars, we're done if (group_cnt == 0) return true; // No decimal point means that it "is" at the end of the string if (decpt_idx == -1) decpt_idx = l; i = 0; while (i < l && group_cnt > 0) { char c = data[i]; if (c == ',') { // Don't allow group chars after the decimal point if (i > decpt_idx) return false; // Check that it is placed correctly relative to the decpt if ((decpt_idx - i) % 4 != 0) return false; // Remove it memmove(data + i, data + i + 1, l - i - 1); data[--l] = '\0'; --group_cnt; --decpt_idx; } else { // Check that we are not missing a separator if (i < decpt_idx && (decpt_idx - i) % 4 == 0 && !(i == 0 && c == '-')) // check for negative sign at start of string return false; ++i; } } return true; } /* Converts a number in locale to its representation in the C locale. Only has to guarantee that a string that is a correct representation of a number will be converted. If junk is passed in, junk will be passed out and the error will be detected during the actual conversion to a number. We can't detect junk here, since we don't even know the base of the number. */ bool QLocalePrivate::numberToCLocale(const QString &num, GroupSeparatorMode group_sep_mode, CharBuff *result) const { const QChar *uc = num.unicode(); int l = num.length(); int idx = 0; // Skip whitespace while (idx < l && uc[idx].isSpace()) ++idx; if (idx == l) return false; const QChar _group = group(); while (idx < l) { const QChar &in = uc[idx]; char out = digitToCLocale(in); if (out == 0) { if (in == list()) out = ';'; else if (in == percent()) out = '%'; // for handling base-x numbers else if (in.unicode() >= 'A' && in.unicode() <= 'Z') out = in.toLower().toLatin1(); else if (in.unicode() >= 'a' && in.unicode() <= 'z') out = in.toLatin1(); else break; } result->append(out); ++idx; } // Check trailing whitespace for (; idx < l; ++idx) { if (!uc[idx].isSpace()) return false; } result->append('\0'); // Check separators if (group_sep_mode == ParseGroupSeparators && !removeGroupSeparators(result)) return false; return true; } bool QLocalePrivate::validateChars(const QString &str, NumberMode numMode, QByteArray *buff, int decDigits) const { buff->clear(); buff->reserve(str.length()); const bool scientific = numMode == DoubleScientificMode; bool lastWasE = false; int eCnt = 0; int decPointCnt = 0; bool dec = false; int decDigitCnt = 0; for (int i = 0; i < str.length(); ++i) { char c = digitToCLocale(str.at(i)); if (c >= '0' && c <= '9') { if (numMode != IntegerMode) { // If a double has too many digits after decpt, it shall be Invalid. if (dec && decDigits != -1 && decDigits < ++decDigitCnt) return false; } } else { switch (c) { case '.': if (numMode == IntegerMode) { // If an integer has a decimal point, it shall be Invalid. return false; } else { // If a double has more than one decimal point, it shall be Invalid. if (++decPointCnt > 1) return false; #if 0 // If a double with no decimal digits has a decimal point, it shall be // Invalid. if (decDigits == 0) return false; #endif // On second thoughts, it shall be Valid. dec = true; } break; case '+': case '-': if (scientific) { // If a scientific has a sign that's not at the beginning or after // an 'e', it shall be Invalid. if (i != 0 && !lastWasE) return false; } else { // If a non-scientific has a sign that's not at the beginning, // it shall be Invalid. if (i != 0) return false; } break; case ',': return false; case 'e': if (scientific) { // If a scientific has more than one 'e', it shall be Invalid. if (++eCnt > 1) return false; dec = false; } else { // If a non-scientific has an 'e', it shall be Invalid. return false; } break; default: // If it's not a valid digit, it shall be Invalid. return false; } } lastWasE = c == 'e'; buff->append(c); } return true; } double QLocalePrivate::stringToDouble(const QString &number, bool *ok, GroupSeparatorMode group_sep_mode) const { CharBuff buff; if (!numberToCLocale(group().unicode() == 0xa0 ? number.trimmed() : number, group_sep_mode, &buff)) { if (ok != 0) *ok = false; return 0.0; } return bytearrayToDouble(buff.constData(), ok); } qlonglong QLocalePrivate::stringToLongLong(const QString &number, int base, bool *ok, GroupSeparatorMode group_sep_mode) const { CharBuff buff; if (!numberToCLocale(group().unicode() == 0xa0 ? number.trimmed() : number, group_sep_mode, &buff)) { if (ok != 0) *ok = false; return 0; } return bytearrayToLongLong(buff.constData(), base, ok); } qulonglong QLocalePrivate::stringToUnsLongLong(const QString &number, int base, bool *ok, GroupSeparatorMode group_sep_mode) const { CharBuff buff; if (!numberToCLocale(group().unicode() == 0xa0 ? number.trimmed() : number, group_sep_mode, &buff)) { if (ok != 0) *ok = false; return 0; } return bytearrayToUnsLongLong(buff.constData(), base, ok); } double QLocalePrivate::bytearrayToDouble(const char *num, bool *ok, bool *overflow) { if (ok != 0) *ok = true; if (overflow != 0) *overflow = false; if (*num == '\0') { if (ok != 0) *ok = false; return 0.0; } if (qstrcmp(num, "nan") == 0) return qt_snan(); if (qstrcmp(num, "+inf") == 0 || qstrcmp(num, "inf") == 0) return qt_inf(); if (qstrcmp(num, "-inf") == 0) return -qt_inf(); bool _ok; const char *endptr; double d = qstrtod(num, &endptr, &_ok); if (!_ok) { // the only way strtod can fail with *endptr != '\0' on a non-empty // input string is overflow if (ok != 0) *ok = false; if (overflow != 0) *overflow = *endptr != '\0'; return 0.0; } if (*endptr != '\0') { // we stopped at a non-digit character after converting some digits if (ok != 0) *ok = false; if (overflow != 0) *overflow = false; return 0.0; } if (ok != 0) *ok = true; if (overflow != 0) *overflow = false; return d; } qlonglong QLocalePrivate::bytearrayToLongLong(const char *num, int base, bool *ok, bool *overflow) { bool _ok; const char *endptr; if (*num == '\0') { if (ok != 0) *ok = false; if (overflow != 0) *overflow = false; return 0; } qlonglong l = qstrtoll(num, &endptr, base, &_ok); if (!_ok) { if (ok != 0) *ok = false; if (overflow != 0) { // the only way qstrtoll can fail with *endptr != '\0' on a non-empty // input string is overflow *overflow = *endptr != '\0'; } return 0; } if (*endptr != '\0') { // we stopped at a non-digit character after converting some digits if (ok != 0) *ok = false; if (overflow != 0) *overflow = false; return 0; } if (ok != 0) *ok = true; if (overflow != 0) *overflow = false; return l; } qulonglong QLocalePrivate::bytearrayToUnsLongLong(const char *num, int base, bool *ok) { bool _ok; const char *endptr; qulonglong l = qstrtoull(num, &endptr, base, &_ok); if (!_ok || *endptr != '\0') { if (ok != 0) *ok = false; return 0; } if (ok != 0) *ok = true; return l; } /*- * Copyright (c) 1992, 1993 * The Regents of the University of California. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgment: * This product includes software developed by the University of * California, Berkeley and its contributors. * 4. Neither the name of the University nor the names of its contributors * may be used to endorse or promote products derived from this software * without specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. */ // static char sccsid[] = "@(#)strtouq.c 8.1 (Berkeley) 6/4/93"; // "$FreeBSD: src/lib/libc/stdlib/strtoull.c,v 1.5.2.1 2001/03/02 09:45:20 obrien Exp $"; /* * Convert a string to an unsigned long long integer. * * Ignores `locale' stuff. Assumes that the upper and lower case * alphabets and digits are each contiguous. */ static qulonglong qstrtoull(const char *nptr, const char **endptr, register int base, bool *ok) { register const char *s = nptr; register qulonglong acc; register unsigned char c; register qulonglong qbase, cutoff; register int any, cutlim; if (ok != 0) *ok = true; /* * See strtoq for comments as to the logic used. */ s = nptr; do { c = *s++; } while (isspace(c)); if (c == '-') { if (ok != 0) *ok = false; if (endptr != 0) *endptr = s - 1; return 0; } else { if (c == '+') c = *s++; } if ((base == 0 || base == 16) && c == '0' && (*s == 'x' || *s == 'X')) { c = s[1]; s += 2; base = 16; } if (base == 0) base = c == '0' ? 8 : 10; qbase = unsigned(base); cutoff = qulonglong(ULLONG_MAX) / qbase; cutlim = qulonglong(ULLONG_MAX) % qbase; for (acc = 0, any = 0;; c = *s++) { if (!isascii(c)) break; if (isdigit(c)) c -= '0'; else if (isalpha(c)) c -= isupper(c) ? 'A' - 10 : 'a' - 10; else break; if (c >= base) break; if (any < 0 || acc > cutoff || (acc == cutoff && c > cutlim)) any = -1; else { any = 1; acc *= qbase; acc += c; } } if (any == 0) { if (ok != 0) *ok = false; } else if (any < 0) { acc = ULLONG_MAX; if (ok != 0) *ok = false; } if (endptr != 0) *endptr = (any ? s - 1 : nptr); return acc; } // "$FreeBSD: src/lib/libc/stdlib/strtoll.c,v 1.5.2.1 2001/03/02 09:45:20 obrien Exp $"; /* * Convert a string to a long long integer. * * Ignores `locale' stuff. Assumes that the upper and lower case * alphabets and digits are each contiguous. */ Q_CORE_EXPORT qlonglong qstrtoll(const char *nptr, const char **endptr, register int base, bool *ok) { register const char *s; register qulonglong acc; register unsigned char c; register qulonglong qbase, cutoff; register int neg, any, cutlim; /* * Skip white space and pick up leading +/- sign if any. * If base is 0, allow 0x for hex and 0 for octal, else * assume decimal; if base is already 16, allow 0x. */ s = nptr; do { c = *s++; } while (isspace(c)); if (c == '-') { neg = 1; c = *s++; } else { neg = 0; if (c == '+') c = *s++; } if ((base == 0 || base == 16) && c == '0' && (*s == 'x' || *s == 'X')) { c = s[1]; s += 2; base = 16; } if (base == 0) base = c == '0' ? 8 : 10; /* * Compute the cutoff value between legal numbers and illegal * numbers. That is the largest legal value, divided by the * base. An input number that is greater than this value, if * followed by a legal input character, is too big. One that * is equal to this value may be valid or not; the limit * between valid and invalid numbers is then based on the last * digit. For instance, if the range for quads is * [-9223372036854775808..9223372036854775807] and the input base * is 10, cutoff will be set to 922337203685477580 and cutlim to * either 7 (neg==0) or 8 (neg==1), meaning that if we have * accumulated a value > 922337203685477580, or equal but the * next digit is > 7 (or 8), the number is too big, and we will * return a range error. * * Set any if any `digits' consumed; make it negative to indicate * overflow. */ qbase = unsigned(base); cutoff = neg ? qulonglong(0-(LLONG_MIN + LLONG_MAX)) + LLONG_MAX : LLONG_MAX; cutlim = cutoff % qbase; cutoff /= qbase; for (acc = 0, any = 0;; c = *s++) { if (!isascii(c)) break; if (isdigit(c)) c -= '0'; else if (isalpha(c)) c -= isupper(c) ? 'A' - 10 : 'a' - 10; else break; if (c >= base) break; if (any < 0 || acc > cutoff || (acc == cutoff && c > cutlim)) any = -1; else { any = 1; acc *= qbase; acc += c; } } if (any < 0) { acc = neg ? LLONG_MIN : LLONG_MAX; if (ok != 0) *ok = false; } else if (neg) { acc = (~acc) + 1; } if (endptr != 0) *endptr = (any >= 0 ? s - 1 : nptr); if (ok != 0) *ok = any > 0; return acc; } #ifndef QT_QLOCALE_USES_FCVT /* From: NetBSD: strtod.c,v 1.26 1998/02/03 18:44:21 perry Exp */ /* $FreeBSD: src/lib/libc/stdlib/netbsd_strtod.c,v 1.2.2.2 2001/03/02 17:14:15 tegge Exp $ */ /* Please send bug reports to David M. Gay AT&T Bell Laboratories, Room 2C-463 600 Mountain Avenue Murray Hill, NJ 07974-2070 U.S.A. dmg@research.att.com or research!dmg */ /* strtod for IEEE-, VAX-, and IBM-arithmetic machines. * * This strtod returns a nearest machine number to the input decimal * string (or sets errno to ERANGE). With IEEE arithmetic, ties are * broken by the IEEE round-even rule. Otherwise ties are broken by * biased rounding (add half and chop). * * Inspired loosely by William D. Clinger's paper "How to Read Floating * Point Numbers Accurately" [Proc. ACM SIGPLAN '90, pp. 92-101]. * * Modifications: * * 1. We only require IEEE, IBM, or VAX double-precision * arithmetic (not IEEE double-extended). * 2. We get by with floating-point arithmetic in a case that * Clinger missed -- when we're computing d * 10^n * for a small integer d and the integer n is not too * much larger than 22 (the maximum integer k for which * we can represent 10^k exactly), we may be able to * compute (d*10^k) * 10^(e-k) with just one roundoff. * 3. Rather than a bit-at-a-time adjustment of the binary * result in the hard case, we use floating-point * arithmetic to determine the adjustment to within * one bit; only in really hard cases do we need to * compute a second residual. * 4. Because of 3., we don't need a large table of powers of 10 * for ten-to-e (just some small tables, e.g. of 10^k * for 0 <= k <= 22). */ /* * #define IEEE_LITTLE_ENDIAN for IEEE-arithmetic machines where the least * significant byte has the lowest address. * #define IEEE_BIG_ENDIAN for IEEE-arithmetic machines where the most * significant byte has the lowest address. * #define Long int on machines with 32-bit ints and 64-bit longs. * #define Sudden_Underflow for IEEE-format machines without gradual * underflow (i.e., that flush to zero on underflow). * #define IBM for IBM mainframe-style floating-point arithmetic. * #define VAX for VAX-style floating-point arithmetic. * #define Unsigned_Shifts if >> does treats its left operand as unsigned. * #define No_leftright to omit left-right logic in fast floating-point * computation of dtoa. * #define Check_FLT_ROUNDS if FLT_ROUNDS can assume the values 2 or 3. * #define RND_PRODQUOT to use rnd_prod and rnd_quot (assembly routines * that use extended-precision instructions to compute rounded * products and quotients) with IBM. * #define ROUND_BIASED for IEEE-format with biased rounding. * #define Inaccurate_Divide for IEEE-format with correctly rounded * products but inaccurate quotients, e.g., for Intel i860. * #define Just_16 to store 16 bits per 32-bit Long when doing high-precision * integer arithmetic. Whether this speeds things up or slows things * down depends on the machine and the number being converted. * #define KR_headers for old-style C function headers. * #define Bad_float_h if your system lacks a float.h or if it does not * define some or all of DBL_DIG, DBL_MAX_10_EXP, DBL_MAX_EXP, * FLT_RADIX, FLT_ROUNDS, and DBL_MAX. * #define MALLOC your_malloc, where your_malloc(n) acts like malloc(n) * if memory is available and otherwise does something you deem * appropriate. If MALLOC is undefined, malloc will be invoked * directly -- and assumed always to succeed. */ #if defined(LIBC_SCCS) && !defined(lint) __RCSID("$NetBSD: strtod.c,v 1.26 1998/02/03 18:44:21 perry Exp $"); #endif /* LIBC_SCCS and not lint */ /* #if defined(__m68k__) || defined(__sparc__) || defined(__i386__) || \ defined(__mips__) || defined(__ns32k__) || defined(__alpha__) || \ defined(__powerpc__) || defined(Q_OS_WIN) || defined(Q_OS_DARWIN) || defined(Q_OS_MAC) || \ defined(mips) || defined(Q_OS_AIX) || defined(Q_OS_SOLARIS) # define IEEE_BIG_OR_LITTLE_ENDIAN 1 #endif */ // *All* of our architectures have IEEE arithmetic, don't they? #define IEEE_BIG_OR_LITTLE_ENDIAN 1 #ifdef __arm32__ /* * Although the CPU is little endian the FP has different * byte and word endianness. The byte order is still little endian * but the word order is big endian. */ #define IEEE_BIG_OR_LITTLE_ENDIAN #endif #ifdef vax #define VAX #endif #define Long qint32 #define ULong quint32 #define MALLOC malloc #ifdef BSD_QDTOA_DEBUG QT_BEGIN_INCLUDE_NAMESPACE #include QT_END_INCLUDE_NAMESPACE #define Bug(x) {fprintf(stderr, "%s\n", x); exit(1);} #endif #ifdef Unsigned_Shifts #define Sign_Extend(a,b) if (b < 0) a |= 0xffff0000; #else #define Sign_Extend(a,b) /*no-op*/ #endif #if (defined(IEEE_BIG_OR_LITTLE_ENDIAN) + defined(VAX) + defined(IBM)) != 1 #error Exactly one of IEEE_BIG_OR_LITTLE_ENDIAN, VAX, or IBM should be defined. #endif static inline ULong _getWord0(const NEEDS_VOLATILE double x) { const NEEDS_VOLATILE uchar *ptr = reinterpret_cast(&x); if (QSysInfo::ByteOrder == QSysInfo::BigEndian) { return (ptr[0]<<24) + (ptr[1]<<16) + (ptr[2]<<8) + ptr[3]; } else { return (ptr[7]<<24) + (ptr[6]<<16) + (ptr[5]<<8) + ptr[4]; } } static inline void _setWord0(NEEDS_VOLATILE double *x, ULong l) { NEEDS_VOLATILE uchar *ptr = reinterpret_cast(x); if (QSysInfo::ByteOrder == QSysInfo::BigEndian) { ptr[0] = uchar(l>>24); ptr[1] = uchar(l>>16); ptr[2] = uchar(l>>8); ptr[3] = uchar(l); } else { ptr[7] = uchar(l>>24); ptr[6] = uchar(l>>16); ptr[5] = uchar(l>>8); ptr[4] = uchar(l); } } static inline ULong _getWord1(const NEEDS_VOLATILE double x) { const NEEDS_VOLATILE uchar *ptr = reinterpret_cast(&x); if (QSysInfo::ByteOrder == QSysInfo::BigEndian) { return (ptr[4]<<24) + (ptr[5]<<16) + (ptr[6]<<8) + ptr[7]; } else { return (ptr[3]<<24) + (ptr[2]<<16) + (ptr[1]<<8) + ptr[0]; } } static inline void _setWord1(NEEDS_VOLATILE double *x, ULong l) { NEEDS_VOLATILE uchar *ptr = reinterpret_cast(x); if (QSysInfo::ByteOrder == QSysInfo::BigEndian) { ptr[4] = uchar(l>>24); ptr[5] = uchar(l>>16); ptr[6] = uchar(l>>8); ptr[7] = uchar(l); } else { ptr[3] = uchar(l>>24); ptr[2] = uchar(l>>16); ptr[1] = uchar(l>>8); ptr[0] = uchar(l); } } static inline ULong getWord0(const NEEDS_VOLATILE double x) { #ifdef QT_ARMFPA return _getWord1(x); #else return _getWord0(x); #endif } static inline void setWord0(NEEDS_VOLATILE double *x, ULong l) { #ifdef QT_ARMFPA _setWord1(x, l); #else _setWord0(x, l); #endif } static inline ULong getWord1(const NEEDS_VOLATILE double x) { #ifdef QT_ARMFPA return _getWord0(x); #else return _getWord1(x); #endif } static inline void setWord1(NEEDS_VOLATILE double *x, ULong l) { #ifdef QT_ARMFPA _setWord0(x, l); #else _setWord1(x, l); #endif } static inline void Storeinc(ULong *&a, const ULong &b, const ULong &c) { *a = (ushort(b) << 16) | ushort(c); ++a; } /* #define P DBL_MANT_DIG */ /* Ten_pmax = floor(P*log(2)/log(5)) */ /* Bletch = (highest power of 2 < DBL_MAX_10_EXP) / 16 */ /* Quick_max = floor((P-1)*log(FLT_RADIX)/log(10) - 1) */ /* Int_max = floor(P*log(FLT_RADIX)/log(10) - 1) */ #if defined(IEEE_BIG_OR_LITTLE_ENDIAN) #define Exp_shift 20 #define Exp_shift1 20 #define Exp_msk1 0x100000 #define Exp_msk11 0x100000 #define Exp_mask 0x7ff00000 #define P 53 #define Bias 1023 #define IEEE_Arith #define Emin (-1022) #define Exp_1 0x3ff00000 #define Exp_11 0x3ff00000 #define Ebits 11 #define Frac_mask 0xfffff #define Frac_mask1 0xfffff #define Ten_pmax 22 #define Bletch 0x10 #define Bndry_mask 0xfffff #define Bndry_mask1 0xfffff #define LSB 1 #define Sign_bit 0x80000000 #define Log2P 1 #define Tiny0 0 #define Tiny1 1 #define Quick_max 14 #define Int_max 14 #define Infinite(x) (getWord0(x) == 0x7ff00000) /* sufficient test for here */ #else #undef Sudden_Underflow #define Sudden_Underflow #ifdef IBM #define Exp_shift 24 #define Exp_shift1 24 #define Exp_msk1 0x1000000 #define Exp_msk11 0x1000000 #define Exp_mask 0x7f000000 #define P 14 #define Bias 65 #define Exp_1 0x41000000 #define Exp_11 0x41000000 #define Ebits 8 /* exponent has 7 bits, but 8 is the right value in b2d */ #define Frac_mask 0xffffff #define Frac_mask1 0xffffff #define Bletch 4 #define Ten_pmax 22 #define Bndry_mask 0xefffff #define Bndry_mask1 0xffffff #define LSB 1 #define Sign_bit 0x80000000 #define Log2P 4 #define Tiny0 0x100000 #define Tiny1 0 #define Quick_max 14 #define Int_max 15 #else /* VAX */ #define Exp_shift 23 #define Exp_shift1 7 #define Exp_msk1 0x80 #define Exp_msk11 0x800000 #define Exp_mask 0x7f80 #define P 56 #define Bias 129 #define Exp_1 0x40800000 #define Exp_11 0x4080 #define Ebits 8 #define Frac_mask 0x7fffff #define Frac_mask1 0xffff007f #define Ten_pmax 24 #define Bletch 2 #define Bndry_mask 0xffff007f #define Bndry_mask1 0xffff007f #define LSB 0x10000 #define Sign_bit 0x8000 #define Log2P 1 #define Tiny0 0x80 #define Tiny1 0 #define Quick_max 15 #define Int_max 15 #endif #endif #ifndef IEEE_Arith #define ROUND_BIASED #endif #ifdef RND_PRODQUOT #define rounded_product(a,b) a = rnd_prod(a, b) #define rounded_quotient(a,b) a = rnd_quot(a, b) extern double rnd_prod(double, double), rnd_quot(double, double); #else #define rounded_product(a,b) a *= b #define rounded_quotient(a,b) a /= b #endif #define Big0 (Frac_mask1 | Exp_msk1*(DBL_MAX_EXP+Bias-1)) #define Big1 0xffffffff #ifndef Just_16 /* When Pack_32 is not defined, we store 16 bits per 32-bit Long. * This makes some inner loops simpler and sometimes saves work * during multiplications, but it often seems to make things slightly * slower. Hence the default is now to store 32 bits per Long. */ #ifndef Pack_32 #define Pack_32 #endif #endif #define Kmax 15 struct Bigint { struct Bigint *next; int k, maxwds, sign, wds; ULong x[1]; }; typedef struct Bigint Bigint; static Bigint *Balloc(int k) { int x; Bigint *rv; x = 1 << k; rv = static_cast(MALLOC(sizeof(Bigint) + (x-1)*sizeof(Long))); rv->k = k; rv->maxwds = x; rv->sign = rv->wds = 0; return rv; } static void Bfree(Bigint *v) { free(v); } #define Bcopy(x,y) memcpy(reinterpret_cast(&x->sign), reinterpret_cast(&y->sign), \ y->wds*sizeof(Long) + 2*sizeof(int)) /* multiply by m and add a */ static Bigint *multadd(Bigint *b, int m, int a) { int i, wds; ULong *x, y; #ifdef Pack_32 ULong xi, z; #endif Bigint *b1; wds = b->wds; x = b->x; i = 0; do { #ifdef Pack_32 xi = *x; y = (xi & 0xffff) * m + a; z = (xi >> 16) * m + (y >> 16); a = (z >> 16); *x++ = (z << 16) + (y & 0xffff); #else y = *x * m + a; a = (y >> 16); *x++ = y & 0xffff; #endif } while(++i < wds); if (a) { if (wds >= b->maxwds) { b1 = Balloc(b->k+1); Bcopy(b1, b); Bfree(b); b = b1; } b->x[wds++] = a; b->wds = wds; } return b; } static Bigint *s2b(const char *s, int nd0, int nd, ULong y9) { Bigint *b; int i, k; Long x, y; x = (nd + 8) / 9; for(k = 0, y = 1; x > y; y <<= 1, k++) ; #ifdef Pack_32 b = Balloc(k); b->x[0] = y9; b->wds = 1; #else b = Balloc(k+1); b->x[0] = y9 & 0xffff; b->wds = (b->x[1] = y9 >> 16) ? 2 : 1; #endif i = 9; if (9 < nd0) { s += 9; do b = multadd(b, 10, *s++ - '0'); while(++i < nd0); s++; } else s += 10; for(; i < nd; i++) b = multadd(b, 10, *s++ - '0'); return b; } static int hi0bits(ULong x) { int k = 0; if (!(x & 0xffff0000)) { k = 16; x <<= 16; } if (!(x & 0xff000000)) { k += 8; x <<= 8; } if (!(x & 0xf0000000)) { k += 4; x <<= 4; } if (!(x & 0xc0000000)) { k += 2; x <<= 2; } if (!(x & 0x80000000)) { k++; if (!(x & 0x40000000)) return 32; } return k; } static int lo0bits(ULong *y) { int k; ULong x = *y; if (x & 7) { if (x & 1) return 0; if (x & 2) { *y = x >> 1; return 1; } *y = x >> 2; return 2; } k = 0; if (!(x & 0xffff)) { k = 16; x >>= 16; } if (!(x & 0xff)) { k += 8; x >>= 8; } if (!(x & 0xf)) { k += 4; x >>= 4; } if (!(x & 0x3)) { k += 2; x >>= 2; } if (!(x & 1)) { k++; x >>= 1; if (!x & 1) return 32; } *y = x; return k; } static Bigint *i2b(int i) { Bigint *b; b = Balloc(1); b->x[0] = i; b->wds = 1; return b; } static Bigint *mult(Bigint *a, Bigint *b) { Bigint *c; int k, wa, wb, wc; ULong carry, y, z; ULong *x, *xa, *xae, *xb, *xbe, *xc, *xc0; #ifdef Pack_32 ULong z2; #endif if (a->wds < b->wds) { c = a; a = b; b = c; } k = a->k; wa = a->wds; wb = b->wds; wc = wa + wb; if (wc > a->maxwds) k++; c = Balloc(k); for(x = c->x, xa = x + wc; x < xa; x++) *x = 0; xa = a->x; xae = xa + wa; xb = b->x; xbe = xb + wb; xc0 = c->x; #ifdef Pack_32 for(; xb < xbe; xb++, xc0++) { if ((y = *xb & 0xffff) != 0) { x = xa; xc = xc0; carry = 0; do { z = (*x & 0xffff) * y + (*xc & 0xffff) + carry; carry = z >> 16; z2 = (*x++ >> 16) * y + (*xc >> 16) + carry; carry = z2 >> 16; Storeinc(xc, z2, z); } while(x < xae); *xc = carry; } if ((y = *xb >> 16) != 0) { x = xa; xc = xc0; carry = 0; z2 = *xc; do { z = (*x & 0xffff) * y + (*xc >> 16) + carry; carry = z >> 16; Storeinc(xc, z, z2); z2 = (*x++ >> 16) * y + (*xc & 0xffff) + carry; carry = z2 >> 16; } while(x < xae); *xc = z2; } } #else for(; xb < xbe; xc0++) { if (y = *xb++) { x = xa; xc = xc0; carry = 0; do { z = *x++ * y + *xc + carry; carry = z >> 16; *xc++ = z & 0xffff; } while(x < xae); *xc = carry; } } #endif for(xc0 = c->x, xc = xc0 + wc; wc > 0 && !*--xc; --wc) ; c->wds = wc; return c; } static Bigint *p5s; static Bigint *pow5mult(Bigint *b, int k) { Bigint *b1, *p5, *p51; int i; static const int p05[3] = { 5, 25, 125 }; if ((i = k & 3) != 0) #if defined(Q_OS_IRIX) && defined(Q_CC_GNU) { // work around a bug on 64 bit IRIX gcc int *p = (int *) p05; b = multadd(b, p[i-1], 0); } #else b = multadd(b, p05[i-1], 0); #endif if (!(k >>= 2)) return b; if (!(p5 = p5s)) { /* first time */ p5 = p5s = i2b(625); p5->next = 0; } for(;;) { if (k & 1) { b1 = mult(b, p5); Bfree(b); b = b1; } if (!(k >>= 1)) break; if (!(p51 = p5->next)) { p51 = p5->next = mult(p5,p5); p51->next = 0; } p5 = p51; } return b; } static Bigint *lshift(Bigint *b, int k) { int i, k1, n, n1; Bigint *b1; ULong *x, *x1, *xe, z; #ifdef Pack_32 n = k >> 5; #else n = k >> 4; #endif k1 = b->k; n1 = n + b->wds + 1; for(i = b->maxwds; n1 > i; i <<= 1) k1++; b1 = Balloc(k1); x1 = b1->x; for(i = 0; i < n; i++) *x1++ = 0; x = b->x; xe = x + b->wds; #ifdef Pack_32 if (k &= 0x1f) { k1 = 32 - k; z = 0; do { *x1++ = *x << k | z; z = *x++ >> k1; } while(x < xe); if ((*x1 = z) != 0) ++n1; } #else if (k &= 0xf) { k1 = 16 - k; z = 0; do { *x1++ = *x << k & 0xffff | z; z = *x++ >> k1; } while(x < xe); if (*x1 = z) ++n1; } #endif else do *x1++ = *x++; while(x < xe); b1->wds = n1 - 1; Bfree(b); return b1; } static int cmp(Bigint *a, Bigint *b) { ULong *xa, *xa0, *xb, *xb0; int i, j; i = a->wds; j = b->wds; #ifdef BSD_QDTOA_DEBUG if (i > 1 && !a->x[i-1]) Bug("cmp called with a->x[a->wds-1] == 0"); if (j > 1 && !b->x[j-1]) Bug("cmp called with b->x[b->wds-1] == 0"); #endif if (i -= j) return i; xa0 = a->x; xa = xa0 + j; xb0 = b->x; xb = xb0 + j; for(;;) { if (*--xa != *--xb) return *xa < *xb ? -1 : 1; if (xa <= xa0) break; } return 0; } static Bigint *diff(Bigint *a, Bigint *b) { Bigint *c; int i, wa, wb; Long borrow, y; /* We need signed shifts here. */ ULong *xa, *xae, *xb, *xbe, *xc; #ifdef Pack_32 Long z; #endif i = cmp(a,b); if (!i) { c = Balloc(0); c->wds = 1; c->x[0] = 0; return c; } if (i < 0) { c = a; a = b; b = c; i = 1; } else i = 0; c = Balloc(a->k); c->sign = i; wa = a->wds; xa = a->x; xae = xa + wa; wb = b->wds; xb = b->x; xbe = xb + wb; xc = c->x; borrow = 0; #ifdef Pack_32 do { y = (*xa & 0xffff) - (*xb & 0xffff) + borrow; borrow = y >> 16; Sign_Extend(borrow, y); z = (*xa++ >> 16) - (*xb++ >> 16) + borrow; borrow = z >> 16; Sign_Extend(borrow, z); Storeinc(xc, z, y); } while(xb < xbe); while(xa < xae) { y = (*xa & 0xffff) + borrow; borrow = y >> 16; Sign_Extend(borrow, y); z = (*xa++ >> 16) + borrow; borrow = z >> 16; Sign_Extend(borrow, z); Storeinc(xc, z, y); } #else do { y = *xa++ - *xb++ + borrow; borrow = y >> 16; Sign_Extend(borrow, y); *xc++ = y & 0xffff; } while(xb < xbe); while(xa < xae) { y = *xa++ + borrow; borrow = y >> 16; Sign_Extend(borrow, y); *xc++ = y & 0xffff; } #endif while(!*--xc) wa--; c->wds = wa; return c; } static double ulp(double x) { Long L; double a; L = (getWord0(x) & Exp_mask) - (P-1)*Exp_msk1; #ifndef Sudden_Underflow if (L > 0) { #endif #ifdef IBM L |= Exp_msk1 >> 4; #endif setWord0(&a, L); setWord1(&a, 0); #ifndef Sudden_Underflow } else { L = -L >> Exp_shift; if (L < Exp_shift) { setWord0(&a, 0x80000 >> L); setWord1(&a, 0); } else { setWord0(&a, 0); L -= Exp_shift; setWord1(&a, L >= 31 ? 1U : 1U << (31 - L)); } } #endif return a; } static double b2d(Bigint *a, int *e) { ULong *xa, *xa0, w, y, z; int k; double d; xa0 = a->x; xa = xa0 + a->wds; y = *--xa; #ifdef BSD_QDTOA_DEBUG if (!y) Bug("zero y in b2d"); #endif k = hi0bits(y); *e = 32 - k; #ifdef Pack_32 if (k < Ebits) { setWord0(&d, Exp_1 | y >> (Ebits - k)); w = xa > xa0 ? *--xa : 0; setWord1(&d, y << ((32-Ebits) + k) | w >> (Ebits - k)); goto ret_d; } z = xa > xa0 ? *--xa : 0; if (k -= Ebits) { setWord0(&d, Exp_1 | y << k | z >> (32 - k)); y = xa > xa0 ? *--xa : 0; setWord1(&d, z << k | y >> (32 - k)); } else { setWord0(&d, Exp_1 | y); setWord1(&d, z); } #else if (k < Ebits + 16) { z = xa > xa0 ? *--xa : 0; setWord0(&d, Exp_1 | y << k - Ebits | z >> Ebits + 16 - k); w = xa > xa0 ? *--xa : 0; y = xa > xa0 ? *--xa : 0; setWord1(&d, z << k + 16 - Ebits | w << k - Ebits | y >> 16 + Ebits - k); goto ret_d; } z = xa > xa0 ? *--xa : 0; w = xa > xa0 ? *--xa : 0; k -= Ebits + 16; setWord0(&d, Exp_1 | y << k + 16 | z << k | w >> 16 - k); y = xa > xa0 ? *--xa : 0; setWord1(&d, w << k + 16 | y << k); #endif ret_d: return d; } static Bigint *d2b(double d, int *e, int *bits) { Bigint *b; int de, i, k; ULong *x, y, z; #ifdef Pack_32 b = Balloc(1); #else b = Balloc(2); #endif x = b->x; z = getWord0(d) & Frac_mask; setWord0(&d, getWord0(d) & 0x7fffffff); /* clear sign bit, which we ignore */ #ifdef Sudden_Underflow de = (int)(getWord0(d) >> Exp_shift); #ifndef IBM z |= Exp_msk11; #endif #else if ((de = int(getWord0(d) >> Exp_shift)) != 0) z |= Exp_msk1; #endif #ifdef Pack_32 if ((y = getWord1(d)) != 0) { if ((k = lo0bits(&y)) != 0) { x[0] = y | z << (32 - k); z >>= k; } else x[0] = y; i = b->wds = (x[1] = z) ? 2 : 1; } else { #ifdef BSD_QDTOA_DEBUG if (!z) Bug("Zero passed to d2b"); #endif k = lo0bits(&z); x[0] = z; i = b->wds = 1; k += 32; } #else if (y = getWord1(d)) { if (k = lo0bits(&y)) if (k >= 16) { x[0] = y | z << 32 - k & 0xffff; x[1] = z >> k - 16 & 0xffff; x[2] = z >> k; i = 2; } else { x[0] = y & 0xffff; x[1] = y >> 16 | z << 16 - k & 0xffff; x[2] = z >> k & 0xffff; x[3] = z >> k+16; i = 3; } else { x[0] = y & 0xffff; x[1] = y >> 16; x[2] = z & 0xffff; x[3] = z >> 16; i = 3; } } else { #ifdef BSD_QDTOA_DEBUG if (!z) Bug("Zero passed to d2b"); #endif k = lo0bits(&z); if (k >= 16) { x[0] = z; i = 0; } else { x[0] = z & 0xffff; x[1] = z >> 16; i = 1; } k += 32; } while(!x[i]) --i; b->wds = i + 1; #endif #ifndef Sudden_Underflow if (de) { #endif #ifdef IBM *e = (de - Bias - (P-1) << 2) + k; *bits = 4*P + 8 - k - hi0bits(getWord0(d) & Frac_mask); #else *e = de - Bias - (P-1) + k; *bits = P - k; #endif #ifndef Sudden_Underflow } else { *e = de - Bias - (P-1) + 1 + k; #ifdef Pack_32 *bits = 32*i - hi0bits(x[i-1]); #else *bits = (i+2)*16 - hi0bits(x[i]); #endif } #endif return b; } static double ratio(Bigint *a, Bigint *b) { double da, db; int k, ka, kb; da = b2d(a, &ka); db = b2d(b, &kb); #ifdef Pack_32 k = ka - kb + 32*(a->wds - b->wds); #else k = ka - kb + 16*(a->wds - b->wds); #endif #ifdef IBM if (k > 0) { setWord0(&da, getWord0(da) + (k >> 2)*Exp_msk1); if (k &= 3) da *= 1 << k; } else { k = -k; setWord0(&db, getWord0(db) + (k >> 2)*Exp_msk1); if (k &= 3) db *= 1 << k; } #else if (k > 0) setWord0(&da, getWord0(da) + k*Exp_msk1); else { k = -k; setWord0(&db, getWord0(db) + k*Exp_msk1); } #endif return da / db; } static const double tens[] = { 1e0, 1e1, 1e2, 1e3, 1e4, 1e5, 1e6, 1e7, 1e8, 1e9, 1e10, 1e11, 1e12, 1e13, 1e14, 1e15, 1e16, 1e17, 1e18, 1e19, 1e20, 1e21, 1e22 #ifdef VAX , 1e23, 1e24 #endif }; #ifdef IEEE_Arith static const double bigtens[] = { 1e16, 1e32, 1e64, 1e128, 1e256 }; static const double tinytens[] = { 1e-16, 1e-32, 1e-64, 1e-128, 1e-256 }; #define n_bigtens 5 #else #ifdef IBM static const double bigtens[] = { 1e16, 1e32, 1e64 }; static const double tinytens[] = { 1e-16, 1e-32, 1e-64 }; #define n_bigtens 3 #else static const double bigtens[] = { 1e16, 1e32 }; static const double tinytens[] = { 1e-16, 1e-32 }; #define n_bigtens 2 #endif #endif /* The pre-release gcc3.3 shipped with SuSE 8.2 has a bug which causes the comparison 1e-100 == 0.0 to return true. As a workaround, we compare it to a global variable containing 0.0, which produces correct assembler output. ### consider detecting the broken compilers and using the static ### double for these, and use a #define for all working compilers */ static double g_double_zero = 0.0; Q_CORE_EXPORT double qstrtod(const char *s00, const char **se, bool *ok) { int bb2, bb5, bbe, bd2, bd5, bbbits, bs2, c, dsign, e, e1, esign, i, j, k, nd, nd0, nf, nz, nz0, sign; const char *s, *s0, *s1; double aadj, aadj1, adj, rv, rv0; Long L; ULong y, z; Bigint *bb1, *bd0; Bigint *bb = NULL, *bd = NULL, *bs = NULL, *delta = NULL;/* pacify gcc */ /* #ifndef KR_headers const char decimal_point = localeconv()->decimal_point[0]; #else const char decimal_point = '.'; #endif */ if (ok != 0) *ok = true; const char decimal_point = '.'; sign = nz0 = nz = 0; rv = 0.; for(s = s00; isspace(uchar(*s)); s++) ; if (*s == '-') { sign = 1; s++; } else if (*s == '+') { s++; } if (*s == '\0') { s = s00; goto ret; } if (*s == '0') { nz0 = 1; while(*++s == '0') ; if (!*s) goto ret; } s0 = s; y = z = 0; for(nd = nf = 0; (c = *s) >= '0' && c <= '9'; nd++, s++) if (nd < 9) y = 10*y + c - '0'; else if (nd < 16) z = 10*z + c - '0'; nd0 = nd; if (c == decimal_point) { c = *++s; if (!nd) { for(; c == '0'; c = *++s) nz++; if (c > '0' && c <= '9') { s0 = s; nf += nz; nz = 0; goto have_dig; } goto dig_done; } for(; c >= '0' && c <= '9'; c = *++s) { have_dig: nz++; if (c -= '0') { nf += nz; for(i = 1; i < nz; i++) if (nd++ < 9) y *= 10; else if (nd <= DBL_DIG + 1) z *= 10; if (nd++ < 9) y = 10*y + c; else if (nd <= DBL_DIG + 1) z = 10*z + c; nz = 0; } } } dig_done: e = 0; if (c == 'e' || c == 'E') { if (!nd && !nz && !nz0) { s = s00; goto ret; } s00 = s; esign = 0; switch(c = *++s) { case '-': esign = 1; case '+': c = *++s; } if (c >= '0' && c <= '9') { while(c == '0') c = *++s; if (c > '0' && c <= '9') { L = c - '0'; s1 = s; while((c = *++s) >= '0' && c <= '9') L = 10*L + c - '0'; if (s - s1 > 8 || L > 19999) /* Avoid confusion from exponents * so large that e might overflow. */ e = 19999; /* safe for 16 bit ints */ else e = int(L); if (esign) e = -e; } else e = 0; } else s = s00; } if (!nd) { if (!nz && !nz0) s = s00; goto ret; } e1 = e -= nf; /* Now we have nd0 digits, starting at s0, followed by a * decimal point, followed by nd-nd0 digits. The number we're * after is the integer represented by those digits times * 10**e */ if (!nd0) nd0 = nd; k = nd < DBL_DIG + 1 ? nd : DBL_DIG + 1; rv = y; if (k > 9) #if defined(Q_OS_IRIX) && defined(Q_CC_GNU) { // work around a bug on 64 bit IRIX gcc double *t = (double *) tens; rv = t[k - 9] * rv + z; } #else rv = tens[k - 9] * rv + z; #endif bd0 = 0; if (nd <= DBL_DIG #ifndef RND_PRODQUOT && FLT_ROUNDS == 1 #endif ) { if (!e) goto ret; if (e > 0) { if (e <= Ten_pmax) { #ifdef VAX goto vax_ovfl_check; #else /* rv = */ rounded_product(rv, tens[e]); goto ret; #endif } i = DBL_DIG - nd; if (e <= Ten_pmax + i) { /* A fancier test would sometimes let us do * this for larger i values. */ e -= i; rv *= tens[i]; #ifdef VAX /* VAX exponent range is so narrow we must * worry about overflow here... */ vax_ovfl_check: setWord0(&rv, getWord0(rv) - P*Exp_msk1); /* rv = */ rounded_product(rv, tens[e]); if ((getWord0(rv) & Exp_mask) > Exp_msk1*(DBL_MAX_EXP+Bias-1-P)) goto ovfl; setWord0(&rv, getWord0(rv) + P*Exp_msk1); #else /* rv = */ rounded_product(rv, tens[e]); #endif goto ret; } } #ifndef Inaccurate_Divide else if (e >= -Ten_pmax) { /* rv = */ rounded_quotient(rv, tens[-e]); goto ret; } #endif } e1 += nd - k; /* Get starting approximation = rv * 10**e1 */ if (e1 > 0) { if ((i = e1 & 15) != 0) rv *= tens[i]; if (e1 &= ~15) { if (e1 > DBL_MAX_10_EXP) { ovfl: // errno = ERANGE; if (ok != 0) *ok = false; #ifdef __STDC__ rv = HUGE_VAL; #else /* Can't trust HUGE_VAL */ #ifdef IEEE_Arith setWord0(&rv, Exp_mask); setWord1(&rv, 0); #else setWord0(&rv, Big0); setWord1(&rv, Big1); #endif #endif if (bd0) goto retfree; goto ret; } if (e1 >>= 4) { for(j = 0; e1 > 1; j++, e1 >>= 1) if (e1 & 1) rv *= bigtens[j]; /* The last multiplication could overflow. */ setWord0(&rv, getWord0(rv) - P*Exp_msk1); rv *= bigtens[j]; if ((z = getWord0(rv) & Exp_mask) > Exp_msk1*(DBL_MAX_EXP+Bias-P)) goto ovfl; if (z > Exp_msk1*(DBL_MAX_EXP+Bias-1-P)) { /* set to largest number */ /* (Can't trust DBL_MAX) */ setWord0(&rv, Big0); setWord1(&rv, Big1); } else setWord0(&rv, getWord0(rv) + P*Exp_msk1); } } } else if (e1 < 0) { e1 = -e1; if ((i = e1 & 15) != 0) rv /= tens[i]; if (e1 &= ~15) { e1 >>= 4; if (e1 >= 1 << n_bigtens) goto undfl; for(j = 0; e1 > 1; j++, e1 >>= 1) if (e1 & 1) rv *= tinytens[j]; /* The last multiplication could underflow. */ rv0 = rv; rv *= tinytens[j]; if (rv == g_double_zero) { rv = 2.*rv0; rv *= tinytens[j]; if (rv == g_double_zero) { undfl: rv = 0.; // errno = ERANGE; if (ok != 0) *ok = false; if (bd0) goto retfree; goto ret; } setWord0(&rv, Tiny0); setWord1(&rv, Tiny1); /* The refinement below will clean * this approximation up. */ } } } /* Now the hard part -- adjusting rv to the correct value.*/ /* Put digits into bd: true value = bd * 10^e */ bd0 = s2b(s0, nd0, nd, y); for(;;) { bd = Balloc(bd0->k); Bcopy(bd, bd0); bb = d2b(rv, &bbe, &bbbits); /* rv = bb * 2^bbe */ bs = i2b(1); if (e >= 0) { bb2 = bb5 = 0; bd2 = bd5 = e; } else { bb2 = bb5 = -e; bd2 = bd5 = 0; } if (bbe >= 0) bb2 += bbe; else bd2 -= bbe; bs2 = bb2; #ifdef Sudden_Underflow #ifdef IBM j = 1 + 4*P - 3 - bbbits + ((bbe + bbbits - 1) & 3); #else j = P + 1 - bbbits; #endif #else i = bbe + bbbits - 1; /* logb(rv) */ if (i < Emin) /* denormal */ j = bbe + (P-Emin); else j = P + 1 - bbbits; #endif bb2 += j; bd2 += j; i = bb2 < bd2 ? bb2 : bd2; if (i > bs2) i = bs2; if (i > 0) { bb2 -= i; bd2 -= i; bs2 -= i; } if (bb5 > 0) { bs = pow5mult(bs, bb5); bb1 = mult(bs, bb); Bfree(bb); bb = bb1; } if (bb2 > 0) bb = lshift(bb, bb2); if (bd5 > 0) bd = pow5mult(bd, bd5); if (bd2 > 0) bd = lshift(bd, bd2); if (bs2 > 0) bs = lshift(bs, bs2); delta = diff(bb, bd); dsign = delta->sign; delta->sign = 0; i = cmp(delta, bs); if (i < 0) { /* Error is less than half an ulp -- check for * special case of mantissa a power of two. */ if (dsign || getWord1(rv) || getWord0(rv) & Bndry_mask) break; delta = lshift(delta,Log2P); if (cmp(delta, bs) > 0) goto drop_down; break; } if (i == 0) { /* exactly half-way between */ if (dsign) { if ((getWord0(rv) & Bndry_mask1) == Bndry_mask1 && getWord1(rv) == 0xffffffff) { /*boundary case -- increment exponent*/ setWord0(&rv, (getWord0(rv) & Exp_mask) + Exp_msk1 #ifdef IBM | Exp_msk1 >> 4 #endif ); setWord1(&rv, 0); break; } } else if (!(getWord0(rv) & Bndry_mask) && !getWord1(rv)) { drop_down: /* boundary case -- decrement exponent */ #ifdef Sudden_Underflow L = getWord0(rv) & Exp_mask; #ifdef IBM if (L < Exp_msk1) #else if (L <= Exp_msk1) #endif goto undfl; L -= Exp_msk1; #else L = (getWord0(rv) & Exp_mask) - Exp_msk1; #endif setWord0(&rv, L | Bndry_mask1); setWord1(&rv, 0xffffffff); #ifdef IBM goto cont; #else break; #endif } #ifndef ROUND_BIASED if (!(getWord1(rv) & LSB)) break; #endif if (dsign) rv += ulp(rv); #ifndef ROUND_BIASED else { rv -= ulp(rv); #ifndef Sudden_Underflow if (rv == g_double_zero) goto undfl; #endif } #endif break; } if ((aadj = ratio(delta, bs)) <= 2.) { if (dsign) aadj = aadj1 = 1.; else if (getWord1(rv) || getWord0(rv) & Bndry_mask) { #ifndef Sudden_Underflow if (getWord1(rv) == Tiny1 && !getWord0(rv)) goto undfl; #endif aadj = 1.; aadj1 = -1.; } else { /* special case -- power of FLT_RADIX to be */ /* rounded down... */ if (aadj < 2./FLT_RADIX) aadj = 1./FLT_RADIX; else aadj *= 0.5; aadj1 = -aadj; } } else { aadj *= 0.5; aadj1 = dsign ? aadj : -aadj; #ifdef Check_FLT_ROUNDS switch(FLT_ROUNDS) { case 2: /* towards +infinity */ aadj1 -= 0.5; break; case 0: /* towards 0 */ case 3: /* towards -infinity */ aadj1 += 0.5; } #else if (FLT_ROUNDS == 0) aadj1 += 0.5; #endif } y = getWord0(rv) & Exp_mask; /* Check for overflow */ if (y == Exp_msk1*(DBL_MAX_EXP+Bias-1)) { rv0 = rv; setWord0(&rv, getWord0(rv) - P*Exp_msk1); adj = aadj1 * ulp(rv); rv += adj; if ((getWord0(rv) & Exp_mask) >= Exp_msk1*(DBL_MAX_EXP+Bias-P)) { if (getWord0(rv0) == Big0 && getWord1(rv0) == Big1) goto ovfl; setWord0(&rv, Big0); setWord1(&rv, Big1); goto cont; } else setWord0(&rv, getWord0(rv) + P*Exp_msk1); } else { #ifdef Sudden_Underflow if ((getWord0(rv) & Exp_mask) <= P*Exp_msk1) { rv0 = rv; setWord0(&rv, getWord0(rv) + P*Exp_msk1); adj = aadj1 * ulp(rv); rv += adj; #ifdef IBM if ((getWord0(rv) & Exp_mask) < P*Exp_msk1) #else if ((getWord0(rv) & Exp_mask) <= P*Exp_msk1) #endif { if (getWord0(rv0) == Tiny0 && getWord1(rv0) == Tiny1) goto undfl; setWord0(&rv, Tiny0); setWord1(&rv, Tiny1); goto cont; } else setWord0(&rv, getWord0(rv) - P*Exp_msk1); } else { adj = aadj1 * ulp(rv); rv += adj; } #else /* Compute adj so that the IEEE rounding rules will * correctly round rv + adj in some half-way cases. * If rv * ulp(rv) is denormalized (i.e., * y <= (P-1)*Exp_msk1), we must adjust aadj to avoid * trouble from bits lost to denormalization; * example: 1.2e-307 . */ if (y <= (P-1)*Exp_msk1 && aadj >= 1.) { aadj1 = int(aadj + 0.5); if (!dsign) aadj1 = -aadj1; } adj = aadj1 * ulp(rv); rv += adj; #endif } z = getWord0(rv) & Exp_mask; if (y == z) { /* Can we stop now? */ L = Long(aadj); aadj -= L; /* The tolerances below are conservative. */ if (dsign || getWord1(rv) || getWord0(rv) & Bndry_mask) { if (aadj < .4999999 || aadj > .5000001) break; } else if (aadj < .4999999/FLT_RADIX) break; } cont: Bfree(bb); Bfree(bd); Bfree(bs); Bfree(delta); } retfree: Bfree(bb); Bfree(bd); Bfree(bs); Bfree(bd0); Bfree(delta); ret: if (se) *se = s; return sign ? -rv : rv; } static int quorem(Bigint *b, Bigint *S) { int n; Long borrow, y; ULong carry, q, ys; ULong *bx, *bxe, *sx, *sxe; #ifdef Pack_32 Long z; ULong si, zs; #endif n = S->wds; #ifdef BSD_QDTOA_DEBUG /*debug*/ if (b->wds > n) /*debug*/ Bug("oversize b in quorem"); #endif if (b->wds < n) return 0; sx = S->x; sxe = sx + --n; bx = b->x; bxe = bx + n; q = *bxe / (*sxe + 1); /* ensure q <= true quotient */ #ifdef BSD_QDTOA_DEBUG /*debug*/ if (q > 9) /*debug*/ Bug("oversized quotient in quorem"); #endif if (q) { borrow = 0; carry = 0; do { #ifdef Pack_32 si = *sx++; ys = (si & 0xffff) * q + carry; zs = (si >> 16) * q + (ys >> 16); carry = zs >> 16; y = (*bx & 0xffff) - (ys & 0xffff) + borrow; borrow = y >> 16; Sign_Extend(borrow, y); z = (*bx >> 16) - (zs & 0xffff) + borrow; borrow = z >> 16; Sign_Extend(borrow, z); Storeinc(bx, z, y); #else ys = *sx++ * q + carry; carry = ys >> 16; y = *bx - (ys & 0xffff) + borrow; borrow = y >> 16; Sign_Extend(borrow, y); *bx++ = y & 0xffff; #endif } while(sx <= sxe); if (!*bxe) { bx = b->x; while(--bxe > bx && !*bxe) --n; b->wds = n; } } if (cmp(b, S) >= 0) { q++; borrow = 0; carry = 0; bx = b->x; sx = S->x; do { #ifdef Pack_32 si = *sx++; ys = (si & 0xffff) + carry; zs = (si >> 16) + (ys >> 16); carry = zs >> 16; y = (*bx & 0xffff) - (ys & 0xffff) + borrow; borrow = y >> 16; Sign_Extend(borrow, y); z = (*bx >> 16) - (zs & 0xffff) + borrow; borrow = z >> 16; Sign_Extend(borrow, z); Storeinc(bx, z, y); #else ys = *sx++ + carry; carry = ys >> 16; y = *bx - (ys & 0xffff) + borrow; borrow = y >> 16; Sign_Extend(borrow, y); *bx++ = y & 0xffff; #endif } while(sx <= sxe); bx = b->x; bxe = bx + n; if (!*bxe) { while(--bxe > bx && !*bxe) --n; b->wds = n; } } return q; } /* dtoa for IEEE arithmetic (dmg): convert double to ASCII string. * * Inspired by "How to Print Floating-Point Numbers Accurately" by * Guy L. Steele, Jr. and Jon L. White [Proc. ACM SIGPLAN '90, pp. 92-101]. * * Modifications: * 1. Rather than iterating, we use a simple numeric overestimate * to determine k = floor(log10(d)). We scale relevant * quantities using O(log2(k)) rather than O(k) multiplications. * 2. For some modes > 2 (corresponding to ecvt and fcvt), we don't * try to generate digits strictly left to right. Instead, we * compute with fewer bits and propagate the carry if necessary * when rounding the final digit up. This is often faster. * 3. Under the assumption that input will be rounded nearest, * mode 0 renders 1e23 as 1e23 rather than 9.999999999999999e22. * That is, we allow equality in stopping tests when the * round-nearest rule will give the same floating-point value * as would satisfaction of the stopping test with strict * inequality. * 4. We remove common factors of powers of 2 from relevant * quantities. * 5. When converting floating-point integers less than 1e16, * we use floating-point arithmetic rather than resorting * to multiple-precision integers. * 6. When asked to produce fewer than 15 digits, we first try * to get by with floating-point arithmetic; we resort to * multiple-precision integer arithmetic only if we cannot * guarantee that the floating-point calculation has given * the correctly rounded result. For k requested digits and * "uniformly" distributed input, the probability is * something like 10^(k-15) that we must resort to the Long * calculation. */ /* This actually sometimes returns a pointer to a string literal cast to a char*. Do NOT try to modify the return value. */ Q_CORE_EXPORT char *qdtoa ( double d, int mode, int ndigits, int *decpt, int *sign, char **rve, char **resultp) { // Some values of the floating-point control word can cause _qdtoa to crash with an underflow. // We set a safe value here. #ifdef Q_OS_WIN _clear87(); unsigned int oldbits = _control87(0, 0); #ifndef MCW_EM # ifdef _MCW_EM # define MCW_EM _MCW_EM # else # define MCW_EM 0x0008001F # endif #endif _control87(MCW_EM, MCW_EM); #endif #if defined(Q_OS_LINUX) && !defined(__UCLIBC__) fenv_t envp; feholdexcept(&envp); #endif char *s = _qdtoa(d, mode, ndigits, decpt, sign, rve, resultp); #ifdef Q_OS_WIN _clear87(); #ifndef _M_X64 _control87(oldbits, 0xFFFFF); #else _control87(oldbits, _MCW_EM|_MCW_DN|_MCW_RC); #endif //_M_X64 #endif //Q_OS_WIN #if defined(Q_OS_LINUX) && !defined(__UCLIBC__) fesetenv(&envp); #endif return s; } static char *_qdtoa( NEEDS_VOLATILE double d, int mode, int ndigits, int *decpt, int *sign, char **rve, char **resultp) { /* Arguments ndigits, decpt, sign are similar to those of ecvt and fcvt; trailing zeros are suppressed from the returned string. If not null, *rve is set to point to the end of the return value. If d is +-Infinity or NaN, then *decpt is set to 9999. mode: 0 ==> shortest string that yields d when read in and rounded to nearest. 1 ==> like 0, but with Steele & White stopping rule; e.g. with IEEE P754 arithmetic , mode 0 gives 1e23 whereas mode 1 gives 9.999999999999999e22. 2 ==> max(1,ndigits) significant digits. This gives a return value similar to that of ecvt, except that trailing zeros are suppressed. 3 ==> through ndigits past the decimal point. This gives a return value similar to that from fcvt, except that trailing zeros are suppressed, and ndigits can be negative. 4-9 should give the same return values as 2-3, i.e., 4 <= mode <= 9 ==> same return as mode 2 + (mode & 1). These modes are mainly for debugging; often they run slower but sometimes faster than modes 2-3. 4,5,8,9 ==> left-to-right digit generation. 6-9 ==> don't try fast floating-point estimate (if applicable). Values of mode other than 0-9 are treated as mode 0. Sufficient space is allocated to the return value to hold the suppressed trailing zeros. */ int bbits, b2, b5, be, dig, i, ieps, ilim0, j, j1, k, k0, k_check, leftright, m2, m5, s2, s5, try_quick; int ilim = 0, ilim1 = 0, spec_case = 0; /* pacify gcc */ Long L; #ifndef Sudden_Underflow int denorm; ULong x; #endif Bigint *b, *b1, *delta, *mhi, *S; Bigint *mlo = NULL; /* pacify gcc */ double d2; double ds, eps; char *s, *s0; if (getWord0(d) & Sign_bit) { /* set sign for everything, including 0's and NaNs */ *sign = 1; setWord0(&d, getWord0(d) & ~Sign_bit); /* clear sign bit */ } else *sign = 0; #if defined(IEEE_Arith) + defined(VAX) #ifdef IEEE_Arith if ((getWord0(d) & Exp_mask) == Exp_mask) #else if (getWord0(d) == 0x8000) #endif { /* Infinity or NaN */ *decpt = 9999; s = #ifdef IEEE_Arith !getWord1(d) && !(getWord0(d) & 0xfffff) ? const_cast("Infinity") : #endif const_cast("NaN"); if (rve) *rve = #ifdef IEEE_Arith s[3] ? s + 8 : #endif s + 3; return s; } #endif #ifdef IBM d += 0; /* normalize */ #endif if (d == g_double_zero) { *decpt = 1; s = const_cast("0"); if (rve) *rve = s + 1; return s; } b = d2b(d, &be, &bbits); #ifdef Sudden_Underflow i = (int)(getWord0(d) >> Exp_shift1 & (Exp_mask>>Exp_shift1)); #else if ((i = int(getWord0(d) >> Exp_shift1 & (Exp_mask>>Exp_shift1))) != 0) { #endif d2 = d; setWord0(&d2, getWord0(d2) & Frac_mask1); setWord0(&d2, getWord0(d2) | Exp_11); #ifdef IBM if (j = 11 - hi0bits(getWord0(d2) & Frac_mask)) d2 /= 1 << j; #endif /* log(x) ~=~ log(1.5) + (x-1.5)/1.5 * log10(x) = log(x) / log(10) * ~=~ log(1.5)/log(10) + (x-1.5)/(1.5*log(10)) * log10(d) = (i-Bias)*log(2)/log(10) + log10(d2) * * This suggests computing an approximation k to log10(d) by * * k = (i - Bias)*0.301029995663981 * + ( (d2-1.5)*0.289529654602168 + 0.176091259055681 ); * * We want k to be too large rather than too small. * The error in the first-order Taylor series approximation * is in our favor, so we just round up the constant enough * to compensate for any error in the multiplication of * (i - Bias) by 0.301029995663981; since |i - Bias| <= 1077, * and 1077 * 0.30103 * 2^-52 ~=~ 7.2e-14, * adding 1e-13 to the constant term more than suffices. * Hence we adjust the constant term to 0.1760912590558. * (We could get a more accurate k by invoking log10, * but this is probably not worthwhile.) */ i -= Bias; #ifdef IBM i <<= 2; i += j; #endif #ifndef Sudden_Underflow denorm = 0; } else { /* d is denormalized */ i = bbits + be + (Bias + (P-1) - 1); x = i > 32 ? getWord0(d) << (64 - i) | getWord1(d) >> (i - 32) : getWord1(d) << (32 - i); d2 = x; setWord0(&d2, getWord0(d2) - 31*Exp_msk1); /* adjust exponent */ i -= (Bias + (P-1) - 1) + 1; denorm = 1; } #endif ds = (d2-1.5)*0.289529654602168 + 0.1760912590558 + i*0.301029995663981; k = int(ds); if (ds < 0. && ds != k) k--; /* want k = floor(ds) */ k_check = 1; if (k >= 0 && k <= Ten_pmax) { if (d < tens[k]) k--; k_check = 0; } j = bbits - i - 1; if (j >= 0) { b2 = 0; s2 = j; } else { b2 = -j; s2 = 0; } if (k >= 0) { b5 = 0; s5 = k; s2 += k; } else { b2 -= k; b5 = -k; s5 = 0; } if (mode < 0 || mode > 9) mode = 0; try_quick = 1; if (mode > 5) { mode -= 4; try_quick = 0; } leftright = 1; switch(mode) { case 0: case 1: ilim = ilim1 = -1; i = 18; ndigits = 0; break; case 2: leftright = 0; /* no break */ case 4: if (ndigits <= 0) ndigits = 1; ilim = ilim1 = i = ndigits; break; case 3: leftright = 0; /* no break */ case 5: i = ndigits + k + 1; ilim = i; ilim1 = i - 1; if (i <= 0) i = 1; } *resultp = static_cast(malloc(i + 1)); s = s0 = *resultp; if (ilim >= 0 && ilim <= Quick_max && try_quick) { /* Try to get by with floating-point arithmetic. */ i = 0; d2 = d; k0 = k; ilim0 = ilim; ieps = 2; /* conservative */ if (k > 0) { ds = tens[k&0xf]; j = k >> 4; if (j & Bletch) { /* prevent overflows */ j &= Bletch - 1; d /= bigtens[n_bigtens-1]; ieps++; } for(; j; j >>= 1, i++) if (j & 1) { ieps++; ds *= bigtens[i]; } d /= ds; } else if ((j1 = -k) != 0) { d *= tens[j1 & 0xf]; for(j = j1 >> 4; j; j >>= 1, i++) if (j & 1) { ieps++; d *= bigtens[i]; } } if (k_check && d < 1. && ilim > 0) { if (ilim1 <= 0) goto fast_failed; ilim = ilim1; k--; d *= 10.; ieps++; } eps = ieps*d + 7.; setWord0(&eps, getWord0(eps) - (P-1)*Exp_msk1); if (ilim == 0) { S = mhi = 0; d -= 5.; if (d > eps) goto one_digit; if (d < -eps) goto no_digits; goto fast_failed; } #ifndef No_leftright if (leftright) { /* Use Steele & White method of only * generating digits needed. */ eps = 0.5/tens[ilim-1] - eps; for(i = 0;;) { L = Long(d); d -= L; *s++ = '0' + int(L); if (d < eps) goto ret1; if (1. - d < eps) goto bump_up; if (++i >= ilim) break; eps *= 10.; d *= 10.; } } else { #endif /* Generate ilim digits, then fix them up. */ #if defined(Q_OS_IRIX) && defined(Q_CC_GNU) // work around a bug on 64 bit IRIX gcc double *t = (double *) tens; eps *= t[ilim-1]; #else eps *= tens[ilim-1]; #endif for(i = 1;; i++, d *= 10.) { L = Long(d); d -= L; *s++ = '0' + int(L); if (i == ilim) { if (d > 0.5 + eps) goto bump_up; else if (d < 0.5 - eps) { while(*--s == '0') {} s++; goto ret1; } break; } } #ifndef No_leftright } #endif fast_failed: s = s0; d = d2; k = k0; ilim = ilim0; } /* Do we have a "small" integer? */ if (be >= 0 && k <= Int_max) { /* Yes. */ ds = tens[k]; if (ndigits < 0 && ilim <= 0) { S = mhi = 0; if (ilim < 0 || d <= 5*ds) goto no_digits; goto one_digit; } for(i = 1;; i++) { L = Long(d / ds); d -= L*ds; #ifdef Check_FLT_ROUNDS /* If FLT_ROUNDS == 2, L will usually be high by 1 */ if (d < 0) { L--; d += ds; } #endif *s++ = '0' + int(L); if (i == ilim) { d += d; if (d > ds || (d == ds && L & 1)) { bump_up: while(*--s == '9') if (s == s0) { k++; *s = '0'; break; } ++*s++; } break; } if ((d *= 10.) == g_double_zero) break; } goto ret1; } m2 = b2; m5 = b5; mhi = mlo = 0; if (leftright) { if (mode < 2) { i = #ifndef Sudden_Underflow denorm ? be + (Bias + (P-1) - 1 + 1) : #endif #ifdef IBM 1 + 4*P - 3 - bbits + ((bbits + be - 1) & 3); #else 1 + P - bbits; #endif } else { j = ilim - 1; if (m5 >= j) m5 -= j; else { s5 += j -= m5; b5 += j; m5 = 0; } if ((i = ilim) < 0) { m2 -= i; i = 0; } } b2 += i; s2 += i; mhi = i2b(1); } if (m2 > 0 && s2 > 0) { i = m2 < s2 ? m2 : s2; b2 -= i; m2 -= i; s2 -= i; } if (b5 > 0) { if (leftright) { if (m5 > 0) { mhi = pow5mult(mhi, m5); b1 = mult(mhi, b); Bfree(b); b = b1; } if ((j = b5 - m5) != 0) b = pow5mult(b, j); } else b = pow5mult(b, b5); } S = i2b(1); if (s5 > 0) S = pow5mult(S, s5); /* Check for special case that d is a normalized power of 2. */ if (mode < 2) { if (!getWord1(d) && !(getWord0(d) & Bndry_mask) #ifndef Sudden_Underflow && getWord0(d) & Exp_mask #endif ) { /* The special case */ b2 += Log2P; s2 += Log2P; spec_case = 1; } else spec_case = 0; } /* Arrange for convenient computation of quotients: * shift left if necessary so divisor has 4 leading 0 bits. * * Perhaps we should just compute leading 28 bits of S once * and for all and pass them and a shift to quorem, so it * can do shifts and ors to compute the numerator for q. */ #ifdef Pack_32 if ((i = ((s5 ? 32 - hi0bits(S->x[S->wds-1]) : 1) + s2) & 0x1f) != 0) i = 32 - i; #else if (i = ((s5 ? 32 - hi0bits(S->x[S->wds-1]) : 1) + s2) & 0xf) i = 16 - i; #endif if (i > 4) { i -= 4; b2 += i; m2 += i; s2 += i; } else if (i < 4) { i += 28; b2 += i; m2 += i; s2 += i; } if (b2 > 0) b = lshift(b, b2); if (s2 > 0) S = lshift(S, s2); if (k_check) { if (cmp(b,S) < 0) { k--; b = multadd(b, 10, 0); /* we botched the k estimate */ if (leftright) mhi = multadd(mhi, 10, 0); ilim = ilim1; } } if (ilim <= 0 && mode > 2) { if (ilim < 0 || cmp(b,S = multadd(S,5,0)) <= 0) { /* no digits, fcvt style */ no_digits: k = -1 - ndigits; goto ret; } one_digit: *s++ = '1'; k++; goto ret; } if (leftright) { if (m2 > 0) mhi = lshift(mhi, m2); /* Compute mlo -- check for special case * that d is a normalized power of 2. */ mlo = mhi; if (spec_case) { mhi = Balloc(mhi->k); Bcopy(mhi, mlo); mhi = lshift(mhi, Log2P); } for(i = 1;;i++) { dig = quorem(b,S) + '0'; /* Do we yet have the shortest decimal string * that will round to d? */ j = cmp(b, mlo); delta = diff(S, mhi); j1 = delta->sign ? 1 : cmp(b, delta); Bfree(delta); #ifndef ROUND_BIASED if (j1 == 0 && !mode && !(getWord1(d) & 1)) { if (dig == '9') goto round_9_up; if (j > 0) dig++; *s++ = dig; goto ret; } #endif if (j < 0 || (j == 0 && !mode #ifndef ROUND_BIASED && !(getWord1(d) & 1) #endif )) { if (j1 > 0) { b = lshift(b, 1); j1 = cmp(b, S); if ((j1 > 0 || (j1 == 0 && dig & 1)) && dig++ == '9') goto round_9_up; } *s++ = dig; goto ret; } if (j1 > 0) { if (dig == '9') { /* possible if i == 1 */ round_9_up: *s++ = '9'; goto roundoff; } *s++ = dig + 1; goto ret; } *s++ = dig; if (i == ilim) break; b = multadd(b, 10, 0); if (mlo == mhi) mlo = mhi = multadd(mhi, 10, 0); else { mlo = multadd(mlo, 10, 0); mhi = multadd(mhi, 10, 0); } } } else for(i = 1;; i++) { *s++ = dig = quorem(b,S) + '0'; if (i >= ilim) break; b = multadd(b, 10, 0); } /* Round off last digit */ b = lshift(b, 1); j = cmp(b, S); if (j > 0 || (j == 0 && dig & 1)) { roundoff: while(*--s == '9') if (s == s0) { k++; *s++ = '1'; goto ret; } ++*s++; } else { while(*--s == '0') {} s++; } ret: Bfree(S); if (mhi) { if (mlo && mlo != mhi) Bfree(mlo); Bfree(mhi); } ret1: Bfree(b); if (s == s0) { /* don't return empty string */ *s++ = '0'; k = 0; } *s = 0; *decpt = k + 1; if (rve) *rve = s; return s0; } #else // NOT thread safe! #include Q_CORE_EXPORT char *qdtoa( double d, int mode, int ndigits, int *decpt, int *sign, char **rve, char **resultp) { if(rve) *rve = 0; char *res; if (mode == 0) ndigits = 80; if (mode == 3) res = fcvt(d, ndigits, decpt, sign); else res = ecvt(d, ndigits, decpt, sign); int n = qstrlen(res); if (mode == 0) { // remove trailing 0's const int stop = qMax(1, *decpt); int i; for (i = n-1; i >= stop; --i) { if (res[i] != '0') break; } n = i + 1; } *resultp = static_cast(malloc(n + 1)); qstrncpy(*resultp, res, n + 1); return *resultp; } Q_CORE_EXPORT double qstrtod(const char *s00, const char **se, bool *ok) { double ret = strtod((char*)s00, (char**)se); if (ok) { if((ret == 0.0l && errno == ERANGE) || ret == HUGE_VAL || ret == -HUGE_VAL) *ok = false; else *ok = true; // the result will be that we don't report underflow in this case } return ret; } #endif // QT_QLOCALE_USES_FCVT QT_END_NAMESPACE