qtextcodec.cpp

奇趣公司比较新的qt/emd版本

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

#include "qplatformdefs.h"
#include "qtextcodec.h"

#ifndef QT_NO_TEXTCODEC

#include "qlist.h"
#include "qfile.h"
#ifndef QT_NO_LIBRARY
# include "qcoreapplication.h"
# include "qtextcodecplugin.h"
# include "private/qfactoryloader_p.h"
#endif
#include "qstringlist.h"

#ifdef Q_OS_UNIX
#  include "qiconvcodec_p.h"
#endif

#include "qutfcodec_p.h"
#include "qsimplecodec_p.h"
#include "qlatincodec_p.h"
#ifndef QT_NO_CODECS
#  include "qtsciicodec_p.h"
#  include "qisciicodec_p.h"
#  if defined(QT_NO_ICONV) && !defined(QT_BOOTSTRAPPED)
// no iconv(3) support, must build all codecs into the library
#    include "../../plugins/codecs/cn/qgb18030codec.h"
#    include "../../plugins/codecs/jp/qeucjpcodec.h"
#    include "../../plugins/codecs/jp/qjiscodec.h"
#    include "../../plugins/codecs/jp/qsjiscodec.h"
#    include "../../plugins/codecs/kr/qeuckrcodec.h"
#    include "../../plugins/codecs/tw/qbig5codec.h"
#  endif // QT_NO_ICONV
#  if defined(Q_WS_X11) && !defined(QT_BOOTSTRAPPED)
#    include "qfontlaocodec_p.h"
#    include "../../plugins/codecs/jp/qfontjpcodec.h"
#  endif
#endif // QT_NO_CODECS
#include "qlocale.h"
#include "private/qmutexpool_p.h"

#include <stdlib.h>
#include <ctype.h>
#ifndef Q_OS_TEMP
#include <locale.h>
#endif
#if defined (_XOPEN_UNIX) && !defined(Q_OS_QNX6) && !defined(Q_OS_OSF)
#include <langinfo.h>
#endif


#ifndef QT_NO_TEXTCODECPLUGIN
Q_GLOBAL_STATIC_WITH_ARGS(QFactoryLoader, loader,
    (QTextCodecFactoryInterface_iid, QCoreApplication::libraryPaths(), QLatin1String("/codecs")))
#endif


static bool nameMatch(const QByteArray &name, const QByteArray &test)
{
    // if they're the same, return a perfect score
    if (qstricmp(name, test) == 0)
        return true;

    const char *n = name.constData();
    const char *h = test.constData();

    // if the letters and numbers are the same, we have a match
    while (*n != '\0') {
        if (isalnum((uchar)*n)) {
            for (;;) {
                if (*h == '\0')
                    return false;
                if (isalnum((uchar)*h))
                    break;
                ++h;
            }
            if (tolower((uchar)*n) != tolower((uchar)*h))
                return false;
            ++h;
        }
        ++n;
    }
    while (*h && !isalnum((uchar)*h))
           ++h;
    return (*h == '\0');
}


static QTextCodec *createForName(const QByteArray &name)
{
#ifndef QT_NO_TEXTCODECPLUGIN
    QFactoryLoader *l = loader();
    QStringList keys = l->keys();
    for (int i = 0; i < keys.size(); ++i) {
        if (nameMatch(name, keys.at(i).toLatin1())) {
            QString realName = keys.at(i);
            if (QTextCodecFactoryInterface *factory
                = qobject_cast<QTextCodecFactoryInterface*>(l->instance(realName))) {
                return factory->create(realName);
            }
        }
    }
#else
    Q_UNUSED(name);
#endif
    return 0;
}

static QTextCodec *createForMib(int mib)
{
#ifndef QT_NO_TEXTCODECPLUGIN
    QString name = QLatin1String("MIB: ") + QString::number(mib);
    if (QTextCodecFactoryInterface *factory
        = qobject_cast<QTextCodecFactoryInterface*>(loader()->instance(name)))
        return factory->create(name);
#else
    Q_UNUSED(mib);
#endif
    return 0;
}

static QList<QTextCodec*> *all = 0;
static bool destroying_is_ok = false;

static QTextCodec *localeMapper = 0;
QTextCodec *QTextCodec::cftr = 0;


class QTextCodecCleanup
{
public:
    ~QTextCodecCleanup();
};

/*
    Deletes all the created codecs. This destructor is called just
    before exiting to delete any QTextCodec objects that may be lying
    around.
*/
QTextCodecCleanup::~QTextCodecCleanup()
{
    if (!all)
        return;

    destroying_is_ok = true;

    while (all->size())
        delete all->takeFirst();
    delete all;
    all = 0;
    localeMapper = 0;

    destroying_is_ok = false;
}

Q_GLOBAL_STATIC(QTextCodecCleanup, createQTextCodecCleanup)

#ifdef Q_OS_WIN32
class QWindowsLocalCodec: public QTextCodec
{
public:
    QWindowsLocalCodec();
    ~QWindowsLocalCodec();

    QString convertToUnicode(const char *, int, ConverterState *) const;
    QByteArray convertFromUnicode(const QChar *, int, ConverterState *) const;

    QByteArray name() const;
    int mibEnum() const;

};

QWindowsLocalCodec::QWindowsLocalCodec()
{
}

QWindowsLocalCodec::~QWindowsLocalCodec()
{
}


QString QWindowsLocalCodec::convertToUnicode(const char *chars, int len, ConverterState * /*state*/) const
{
    return qt_winMB2QString(chars, len);
}

QByteArray QWindowsLocalCodec::convertFromUnicode(const QChar *uc, int len, ConverterState *) const
{
    return qt_winQString2MB(uc, len);
}


QByteArray QWindowsLocalCodec::name() const
{
    return "System";
}

int QWindowsLocalCodec::mibEnum() const
{
    return 0;
}

#else

/* locale names mostly copied from XFree86 */
static const char * const iso8859_2locales[] = {
    "croatian", "cs", "cs_CS", "cs_CZ","cz", "cz_CZ", "czech", "hr",
    "hr_HR", "hu", "hu_HU", "hungarian", "pl", "pl_PL", "polish", "ro",
    "ro_RO", "rumanian", "serbocroatian", "sh", "sh_SP", "sh_YU", "sk",
    "sk_SK", "sl", "sl_CS", "sl_SI", "slovak", "slovene", "sr_SP", 0 };

static const char * const iso8859_3locales[] = {
    "eo", 0 };

static const char * const iso8859_4locales[] = {
    "ee", "ee_EE", 0 };

static const char * const iso8859_5locales[] = {
    "mk", "mk_MK", "sp", "sp_YU", 0 };

static const char * const cp_1251locales[] = {
    "be", "be_BY", "bg", "bg_BG", "bulgarian", 0 };

static const char * const pt_154locales[] = {
    "ba_RU", "ky", "ky_KG", "kk", "kk_KZ", 0 };

static const char * const iso8859_6locales[] = {
    "ar_AA", "ar_SA", "arabic", 0 };

static const char * const iso8859_7locales[] = {
    "el", "el_GR", "greek", 0 };

static const char * const iso8859_8locales[] = {
    "hebrew", "he", "he_IL", "iw", "iw_IL", 0 };

static const char * const iso8859_9locales[] = {
    "tr", "tr_TR", "turkish", 0 };

static const char * const iso8859_13locales[] = {
    "lt", "lt_LT", "lv", "lv_LV", 0 };

static const char * const iso8859_15locales[] = {
    "et", "et_EE",
    // Euro countries
    "br_FR", "ca_ES", "de", "de_AT", "de_BE", "de_DE", "de_LU", "en_IE",
    "es", "es_ES", "eu_ES", "fi", "fi_FI", "finnish", "fr", "fr_FR",
    "fr_BE", "fr_LU", "french", "ga_IE", "gl_ES", "it", "it_IT", "oc_FR",
    "nl", "nl_BE", "nl_NL", "pt", "pt_PT", "sv_FI", "wa_BE",
    0 };

static const char * const koi8_ulocales[] = {
    "uk", "uk_UA", "ru_UA", "ukrainian", 0 };

static const char * const tis_620locales[] = {
    "th", "th_TH", "thai", 0 };

// static const char * const tcvnlocales[] = {
//     "vi", "vi_VN", 0 };

static bool try_locale_list(const char * const locale[], const char * lang)
{
    int i;
    for(i=0; locale[i] && *locale[i] && strcmp(locale[i], lang); i++)
        ;
    return locale[i] != 0;
}

// For the probably_koi8_locales we have to look. the standard says
// these are 8859-5, but almost all Russian users use KOI8-R and
// incorrectly set $LANG to ru_RU. We'll check tolower() to see what
// it thinks ru_RU means.

// If you read the history, it seems that many Russians blame ISO and
// Perestroika for the confusion.
//
// The real bug is that some programs break if the user specifies
// ru_RU.KOI8-R.

static const char * const probably_koi8_rlocales[] = {
    "ru", "ru_SU", "ru_RU", "russian", 0 };

static QTextCodec * ru_RU_hack(const char * i) {
    QTextCodec * ru_RU_codec = 0;

    QByteArray origlocale(setlocale(LC_CTYPE, i));
    // unicode   koi8r   latin5   name
    // 0x044E    0xC0    0xEE     CYRILLIC SMALL LETTER YU
    // 0x042E    0xE0    0xCE     CYRILLIC CAPITAL LETTER YU
    int latin5 = tolower(0xCE);
    int koi8r = tolower(0xE0);
    if (koi8r == 0xC0 && latin5 != 0xEE) {
        ru_RU_codec = QTextCodec::codecForName("KOI8-R");
    } else if (koi8r != 0xC0 && latin5 == 0xEE) {
        ru_RU_codec = QTextCodec::codecForName("ISO 8859-5");
    } else {
        // something else again... let's assume... *throws dice*
        ru_RU_codec = QTextCodec::codecForName("KOI8-R");
        qWarning("QTextCodec: Using KOI8-R, probe failed (%02x %02x %s)",
                  koi8r, latin5, i);
    }
    setlocale(LC_CTYPE, origlocale);

    return ru_RU_codec;
}

#endif

#ifndef Q_OS_WIN32
static QTextCodec *checkForCodec(const char *name) {
    QTextCodec *c = QTextCodec::codecForName(name);
    if (!c) {
        const char *at = strchr(name, '@');
        if (at) {
            QByteArray n(name, at - name);
            c = QTextCodec::codecForName(n.data());
        }
    }
    return c;
}
#endif

/* the next two functions are implicitely thread safe,
   as they are only called by setup() which uses a mutex.
*/
static void setupLocaleMapper()
{
#ifdef Q_OS_WIN32
    localeMapper = QTextCodec::codecForName("System");
#else

#ifndef QT_NO_ICONV
    localeMapper = QTextCodec::codecForName("System");
#endif

#if defined (_XOPEN_UNIX) && !defined(Q_OS_QNX6) && !defined(Q_OS_OSF)
    if (!localeMapper) {
        char *charset = nl_langinfo (CODESET);
        if (charset)
            localeMapper = QTextCodec::codecForName(charset);
    }
#endif

    if (!localeMapper) {
        // Very poorly defined and followed standards causes lots of
        // code to try to get all the cases... This logic is
        // duplicated in QIconvCodec, so if you change it here, change
        // it there too.

        // Try to determine locale codeset from locale name assigned to
        // LC_CTYPE category.

        // First part is getting that locale name.  First try setlocale() which
        // definitely knows it, but since we cannot fully trust it, get ready
        // to fall back to environment variables.
        char * ctype = qstrdup(setlocale(LC_CTYPE, 0));

        // Get the first nonempty value from $LC_ALL, $LC_CTYPE, and $LANG
        // environment variables.
        char * lang = qstrdup(qgetenv("LC_ALL").constData());
        if (!lang || lang[0] == 0 || strcmp(lang, "C") == 0) {
            if (lang) delete [] lang;
            lang = qstrdup(qgetenv("LC_CTYPE").constData());
        }
        if (!lang || lang[0] == 0 || strcmp(lang, "C") == 0) {
            if (lang) delete [] lang;
            lang = qstrdup(qgetenv("LANG").constData());
        }

        // Now try these in order:
        // 1. CODESET from ctype if it contains a .CODESET part (e.g. en_US.ISO8859-15)
        // 2. CODESET from lang if it contains a .CODESET part
        // 3. ctype (maybe the locale is named "ISO-8859-1" or something)
        // 4. locale (ditto)
        // 5. check for "@euro"
        // 6. guess locale from ctype unless ctype is "C"
        // 7. guess locale from lang

        // 1. CODESET from ctype if it contains a .CODESET part (e.g. en_US.ISO8859-15)
        char * codeset = ctype ? strchr(ctype, '.') : 0;
        if (codeset && *codeset == '.')
            localeMapper = checkForCodec(codeset + 1);

        // 2. CODESET from lang if it contains a .CODESET part
        codeset = lang ? strchr(lang, '.') : 0;
        if (!localeMapper && codeset && *codeset == '.')
            localeMapper = checkForCodec(codeset + 1);

        // 3. ctype (maybe the locale is named "ISO-8859-1" or something)
        if (!localeMapper && ctype && *ctype != 0 && strcmp (ctype, "C") != 0)
            localeMapper = checkForCodec(ctype);

        // 4. locale (ditto)
        if (!localeMapper && lang && *lang != 0)
            localeMapper = checkForCodec(lang);

        // 5. "@euro"
        if (!localeMapper && ctype && strstr(ctype, "@euro") || lang && strstr(lang, "@euro"))
            localeMapper = checkForCodec("ISO 8859-15");

        // 6. guess locale from ctype unless ctype is "C"
        // 7. guess locale from lang
        char * try_by_name = ctype;
        if (ctype && *ctype != 0 && strcmp (ctype, "C") != 0)
            try_by_name = lang;

        // Now do the guessing.
        if (lang && *lang && !localeMapper && try_by_name && *try_by_name) {
            if (try_locale_list(iso8859_15locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-15");
            else if (try_locale_list(iso8859_2locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-2");
            else if (try_locale_list(iso8859_3locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-3");
            else if (try_locale_list(iso8859_4locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-4");
            else if (try_locale_list(iso8859_5locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-5");
            else if (try_locale_list(iso8859_6locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-6");
            else if (try_locale_list(iso8859_7locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-7");
            else if (try_locale_list(iso8859_8locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-8-I");
            else if (try_locale_list(iso8859_9locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-9");
            else if (try_locale_list(iso8859_13locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-13");
            else if (try_locale_list(tis_620locales, lang))
                localeMapper = QTextCodec::codecForName("ISO 8859-11");
            else if (try_locale_list(koi8_ulocales, lang))
                localeMapper = QTextCodec::codecForName("KOI8-U");
            else if (try_locale_list(cp_1251locales, lang))
                localeMapper = QTextCodec::codecForName("CP 1251");
            else if (try_locale_list(pt_154locales, lang))
                localeMapper = QTextCodec::codecForName("PT 154");
            else if (try_locale_list(probably_koi8_rlocales, lang))