ICU code page conversion support

Use ICU to do code page conversion instead of the
builtin text codecs. With this QTextCodec simply
becomes a wrapper around ICU's ucnv_* methods.

We only keep our own codecs for UTF-*, ISO-8859-1,
ISO-8859-15 for performance reasons, and for TSCII
and iscii-* because they aren't supported by ICU.

Change-Id: I4fc49eba55cf772b9772c6dac606a47a44346a60
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>

5 files changed, 791 insertions, 24 deletions

diff --git a/src/corelib/codecs/codecs.pri b/src/corelib/codecs/codecs.pri
index b08d412452..d049e52275 100644
--- a/src/corelib/codecs/codecs.pri
+++ b/src/corelib/codecs/codecs.pri
@@ -3,32 +3,43 @@
 HEADERS += \
     codecs/qisciicodec_p.h \
     codecs/qlatincodec_p.h \
-    codecs/qsimplecodec_p.h \
     codecs/qtextcodec_p.h \
     codecs/qtextcodec.h \
     codecs/qtsciicodec_p.h \
-    codecs/qutfcodec_p.h \
-    codecs/qgb18030codec_p.h \
-    codecs/qeucjpcodec_p.h \
-    codecs/qjiscodec_p.h \
-    codecs/qsjiscodec_p.h \
-    codecs/qeuckrcodec_p.h \
-    codecs/qbig5codec_p.h
+    codecs/qutfcodec_p.h
 
 SOURCES += \
     codecs/qisciicodec.cpp \
     codecs/qlatincodec.cpp \
-    codecs/qsimplecodec.cpp \
     codecs/qtextcodec.cpp \
     codecs/qtsciicodec.cpp \
-    codecs/qutfcodec.cpp \
-    codecs/qgb18030codec.cpp \
-    codecs/qjpunicode.cpp \
-    codecs/qeucjpcodec.cpp \
-    codecs/qjiscodec.cpp \
-    codecs/qsjiscodec.cpp \
-    codecs/qeuckrcodec.cpp \
-    codecs/qbig5codec.cpp
+    codecs/qutfcodec.cpp
+
+contains(QT_CONFIG,icu) {
+    HEADERS += \
+        codecs/qicucodec_p.h
+    SOURCES += \
+        codecs/qicucodec.cpp
+} else {
+    HEADERS += \
+        codecs/qsimplecodec_p.h \
+        codecs/qgb18030codec_p.h \
+        codecs/qeucjpcodec_p.h \
+        codecs/qjiscodec_p.h \
+        codecs/qsjiscodec_p.h \
+        codecs/qeuckrcodec_p.h \
+        codecs/qbig5codec_p.h
+
+    SOURCES += \
+        codecs/qsimplecodec.cpp \
+        codecs/qgb18030codec.cpp \
+        codecs/qjpunicode.cpp \
+        codecs/qeucjpcodec.cpp \
+        codecs/qjiscodec.cpp \
+        codecs/qsjiscodec.cpp \
+        codecs/qeuckrcodec.cpp \
+        codecs/qbig5codec.cpp
+}
 
 unix:!qnx:!mac:!ios:!linux-android-* {
         contains(QT_CONFIG,iconv) {
diff --git a/src/corelib/codecs/qicucodec.cpp b/src/corelib/codecs/qicucodec.cpp
new file mode 100644
index 0000000000..c91115b3b5
--- /dev/null
+++ b/src/corelib/codecs/qicucodec.cpp
@@ -0,0 +1,625 @@
+/****************************************************************************
+**
+** Copyright (C) 2012 Nokia Corporation and/or its subsidiary(-ies).
+** Contact: http://www.qt-project.org/
+**
+** This file is part of the QtCore module of the Qt Toolkit.
+**
+** $QT_BEGIN_LICENSE:LGPL$
+** GNU Lesser General Public License Usage
+** 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 1.1, 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.
+**
+** Other Usage
+** Alternatively, this file may be used in accordance with the terms and
+** conditions contained in a signed written agreement between you and Nokia.
+**
+**
+**
+**
+**
+**
+** $QT_END_LICENSE$
+**
+****************************************************************************/
+
+#include "qicucodec_p.h"
+#include "qtextcodec_p.h"
+#include "private/qcoreglobaldata_p.h"
+#include "qdebug.h"
+
+#include "unicode/ucnv.h"
+
+QT_BEGIN_NAMESPACE
+
+extern QMutex *qTextCodecsMutex();
+
+static void qIcuCodecStateFree(QTextCodec::ConverterState *state)
+{
+    ucnv_close(static_cast<UConverter *>(state->d));
+}
+
+/* The list below is generated from http://www.iana.org/assignments/character-sets/
+   using the snippet of code below:
+
+#include <QtCore>
+#include <unicode/ucnv.h>
+
+int main(int argc, char **argv)
+{
+    QCoreApplication app(argc, argv);
+
+    QFile file("character-sets.txt");
+    file.open(QFile::ReadOnly);
+    QByteArray name;
+    int mib = -1;
+    QByteArray nameList;
+    int pos = 0;
+    while (!file.atEnd()) {
+        QByteArray s = file.readLine().trimmed();
+        if (s.isEmpty()) {
+            if (mib != -1) {
+                UErrorCode error = U_ZERO_ERROR;
+                const char *standard_name = ucnv_getStandardName(name, "MIME", &error);
+                if (U_FAILURE(error) || !standard_name) {
+                    error = U_ZERO_ERROR;
+                    standard_name = ucnv_getStandardName(name, "IANA", &error);
+                }
+                UConverter *conv = ucnv_open(standard_name, &error);
+                if (!U_FAILURE(error) && conv && standard_name) {
+                    ucnv_close(conv);
+                    printf("    { %d, %d },\n", mib, pos);
+                    nameList += "\"";
+                    nameList += standard_name;
+                    nameList += "\\0\"\n";
+                    pos += strlen(standard_name) + 1;
+                }
+            }
+            name = QByteArray();
+            mib = -1;
+        }
+        if (s.startsWith("Name: ")) {
+            name = s.mid(5).trimmed();
+            if (name.indexOf(' ') > 0)
+                name = name.left(name.indexOf(' '));
+        }
+        if (s.startsWith("MIBenum:"))
+            mib = s.mid(8).trimmed().toInt();
+        if (s.startsWith("Alias:") && s.contains("MIME")) {
+            name = s.mid(6).trimmed();
+            name = name.left(name.indexOf(' ')).trimmed();
+        }
+    }
+    qDebug() << nameList;
+}
+*/
+
+struct MibToName {
+    short mib;
+    short index;
+};
+
+static MibToName mibToName[] = {
+    { 3, 0 },
+    { 4, 9 },
+    { 5, 20 },
+    { 6, 31 },
+    { 7, 42 },
+    { 8, 53 },
+    { 9, 64 },
+    { 10, 75 },
+    { 11, 86 },
+    { 12, 97 },
+    { 13, 108 },
+    { 16, 120 },
+    { 17, 134 },
+    { 18, 144 },
+    { 30, 151 },
+    { 36, 160 },
+    { 37, 167 },
+    { 38, 179 },
+    { 39, 186 },
+    { 40, 198 },
+    { 57, 212 },
+    { 81, 223 },
+    { 82, 234 },
+    { 84, 245 },
+    { 85, 256 },
+    { 104, 267 },
+    { 105, 279 },
+    { 106, 295 },
+    { 109, 301 },
+    { 110, 313 },
+    { 111, 325 },
+    { 113, 337 },
+    { 114, 341 },
+    { 1000, 349 },
+    { 1001, 356 },
+    { 1011, 363 },
+    { 1012, 368 },
+    { 1013, 374 },
+    { 1014, 383 },
+    { 1015, 392 },
+    { 1016, 399 },
+    { 1017, 406 },
+    { 1018, 413 },
+    { 1019, 422 },
+    { 1020, 431 },
+    { 2004, 438 },
+    { 2005, 448 },
+    { 2009, 472 },
+    { 2013, 479 },
+    { 2016, 486 },
+    { 2024, 495 },
+    { 2025, 505 },
+    { 2026, 512 },
+    { 2027, 517 },
+    { 2028, 527 },
+    { 2030, 534 },
+    { 2033, 541 },
+    { 2034, 548 },
+    { 2035, 555 },
+    { 2037, 562 },
+    { 2038, 569 },
+    { 2039, 576 },
+    { 2040, 583 },
+    { 2041, 590 },
+    { 2043, 597 },
+    { 2011, 604 },
+    { 2044, 611 },
+    { 2045, 618 },
+    { 2010, 624 },
+    { 2046, 631 },
+    { 2047, 638 },
+    { 2048, 645 },
+    { 2049, 652 },
+    { 2050, 659 },
+    { 2051, 666 },
+    { 2052, 673 },
+    { 2053, 680 },
+    { 2054, 687 },
+    { 2055, 694 },
+    { 2056, 701 },
+    { 2062, 708 },
+    { 2063, 715 },
+    { 2084, 723 },
+    { 2085, 730 },
+    { 2086, 741 },
+    { 2087, 748 },
+    { 2088, 755 },
+    { 2089, 762 },
+    { 2091, 771 },
+    { 2092, 780 },
+    { 2093, 789 },
+    { 2094, 798 },
+    { 2095, 807 },
+    { 2096, 816 },
+    { 2097, 825 },
+    { 2098, 834 },
+    { 2099, 843 },
+    { 2100, 852 },
+    { 2101, 861 },
+    { 2102, 872 },
+    { 2250, 880 },
+    { 2251, 893 },
+    { 2252, 906 },
+    { 2253, 919 },
+    { 2254, 932 },
+    { 2255, 945 },
+    { 2256, 958 },
+    { 2257, 971 },
+    { 2258, 984 },
+    { 2259, 997 },
+};
+int mibToNameSize = sizeof(mibToName)/sizeof(MibToName);
+
+static const char mibToNameTable[] =
+    "US-ASCII\0"
+    "ISO-8859-1\0"
+    "ISO-8859-2\0"
+    "ISO-8859-3\0"
+    "ISO-8859-4\0"
+    "ISO-8859-5\0"
+    "ISO-8859-6\0"
+    "ISO-8859-7\0"
+    "ISO-8859-8\0"
+    "ISO-8859-9\0"
+    "ISO-8859-10\0"
+    "ISO-2022-JP-1\0"
+    "Shift_JIS\0"
+    "EUC-JP\0"
+    "US-ASCII\0"
+    "EUC-KR\0"
+    "ISO-2022-KR\0"
+    "EUC-KR\0"
+    "ISO-2022-JP\0"
+    "ISO-2022-JP-2\0"
+    "GB_2312-80\0"
+    "ISO-8859-6\0"
+    "ISO-8859-6\0"
+    "ISO-8859-8\0"
+    "ISO-8859-8\0"
+    "ISO-2022-CN\0"
+    "ISO-2022-CN-EXT\0"
+    "UTF-8\0"
+    "ISO-8859-13\0"
+    "ISO-8859-14\0"
+    "ISO-8859-15\0"
+    "GBK\0"
+    "GB18030\0"
+    "UTF-16\0"
+    "UTF-32\0"
+    "SCSU\0"
+    "UTF-7\0"
+    "UTF-16BE\0"
+    "UTF-16LE\0"
+    "UTF-16\0"
+    "CESU-8\0"
+    "UTF-32\0"
+    "UTF-32BE\0"
+    "UTF-32LE\0"
+    "BOCU-1\0"
+    "hp-roman8\0"
+    "Adobe-Standard-Encoding\0"
+    "IBM850\0"
+    "IBM862\0"
+    "IBM-Thai\0"
+    "Shift_JIS\0"
+    "GB2312\0"
+    "Big5\0"
+    "macintosh\0"
+    "IBM037\0"
+    "IBM273\0"
+    "IBM277\0"
+    "IBM278\0"
+    "IBM280\0"
+    "IBM284\0"
+    "IBM285\0"
+    "IBM290\0"
+    "IBM297\0"
+    "IBM420\0"
+    "IBM424\0"
+    "IBM437\0"
+    "IBM500\0"
+    "cp851\0"
+    "IBM852\0"
+    "IBM855\0"
+    "IBM857\0"
+    "IBM860\0"
+    "IBM861\0"
+    "IBM863\0"
+    "IBM864\0"
+    "IBM865\0"
+    "IBM868\0"
+    "IBM869\0"
+    "IBM870\0"
+    "IBM871\0"
+    "IBM918\0"
+    "IBM1026\0"
+    "KOI8-R\0"
+    "HZ-GB-2312\0"
+    "IBM866\0"
+    "IBM775\0"
+    "KOI8-U\0"
+    "IBM00858\0"
+    "IBM01140\0"
+    "IBM01141\0"
+    "IBM01142\0"
+    "IBM01143\0"
+    "IBM01144\0"
+    "IBM01145\0"
+    "IBM01146\0"
+    "IBM01147\0"
+    "IBM01148\0"
+    "IBM01149\0"
+    "Big5-HKSCS\0"
+    "IBM1047\0"
+    "windows-1250\0"
+    "windows-1251\0"
+    "windows-1252\0"
+    "windows-1253\0"
+    "windows-1254\0"
+    "windows-1255\0"
+    "windows-1256\0"
+    "windows-1257\0"
+    "windows-1258\0"
+    "TIS-620\0";
+
+QList<QByteArray> QIcuCodec::availableCodecs()
+{
+    QList<QByteArray> codecs;
+    int n = ucnv_countAvailable();
+    for (int i = 0; i < n; ++i) {
+        const char *name = ucnv_getAvailableName(i);
+
+        UErrorCode error = U_ZERO_ERROR;
+        const char *standardName = ucnv_getStandardName(name, "MIME", &error);
+        if (U_FAILURE(error) || !standardName) {
+            error = U_ZERO_ERROR;
+            standardName = ucnv_getStandardName(name, "IANA", &error);
+        }
+        if (U_FAILURE(error))
+            continue;
+
+        error = U_ZERO_ERROR;
+        int ac = ucnv_countAliases(standardName, &error);
+        if (U_FAILURE(error))
+            continue;
+        for (int j = 0; j < ac; ++j) {
+            error = U_ZERO_ERROR;
+            const char *alias = ucnv_getAlias(standardName, j, &error);
+            if (!U_SUCCESS(error))
+                continue;
+            codecs += alias;
+        }
+    }
+
+    // handled by Qt and not in ICU:
+    codecs += "TSCII";
+    codecs += "System";
+
+    return codecs;
+}
+
+QList<int> QIcuCodec::availableMibs()
+{
+    QList<int> mibs;
+    for (int i = 0; i < mibToNameSize; ++i)
+        mibs += mibToName[i].mib;
+
+    // handled by Qt and not in ICU:
+    mibs += 2107; // TSCII
+
+    return mibs;
+}
+
+static inline bool nameMatch(const QByteArray &a, const char *b)
+{ return a == b; }
+
+QTextCodec *QIcuCodec::codecForName(const char *name)
+{
+    // backwards compatibility with Qt 4.x
+    if (!qstrcmp(name, "CP949"))
+        name = "windows-949";
+    // this one is broken data in ICU 4.4, and can't be resolved even though it's an alias to tis-620
+    if (!qstrcmp(name, "windows-874-2000"))
+        name = "TIS-620";
+
+    UErrorCode error = U_ZERO_ERROR;
+    // MIME gives better default names
+    const char *standardName = ucnv_getStandardName(name, "MIME", &error);
+    if (U_FAILURE(error) || !standardName) {
+        error = U_ZERO_ERROR;
+        standardName = ucnv_getStandardName(name, "IANA", &error);
+    }
+    bool qt_only = false;
+    if (U_FAILURE(error) || !standardName) {
+        standardName = name;
+        qt_only = true;
+    } else {
+        // correct some issues where the ICU data set contains duplicated entries.
+        // Where this happens it's because one data set is a subset of another. We
+        // always use the larger data set.
+
+        if (qstrcmp(standardName, "GB2312") == 0 || qstrcmp(standardName, "GB_2312-80") == 0)
+            standardName = "GBK";
+        else if (qstrcmp(standardName, "KSC_5601") == 0 || qstrcmp(standardName, "EUC-KR") == 0 || qstrcmp(standardName, "cp1363") == 0)
+            standardName = "windows-949";
+    }
+
+    QCoreGlobalData *globalData = QCoreGlobalData::instance();
+    QTextCodecCache *cache = &globalData->codecCache;
+
+    QTextCodec *codec;
+    if (cache) {
+        codec = cache->value(standardName);
+        if (codec)
+            return codec;
+    }
+
+    for (int i = 0; i < globalData->allCodecs.size(); ++i) {
+        QTextCodec *cursor = globalData->allCodecs.at(i);
+        if (nameMatch(cursor->name(), standardName)) {
+            if (cache)
+                cache->insert(standardName, cursor);
+            return cursor;
+        }
+        QList<QByteArray> aliases = cursor->aliases();
+        for (int y = 0; y < aliases.size(); ++y)
+            if (nameMatch(aliases.at(y), standardName)) {
+                if (cache)
+                    cache->insert(standardName, cursor);
+                return cursor;
+            }
+    }
+
+    if (qt_only)
+        return 0;
+
+    // check whether there is really a converter for the name available.
+    UConverter *conv = ucnv_open(standardName, &error);
+    if (!conv) {
+        qDebug() << "codecForName: ucnv_open failed" << standardName << u_errorName(error);
+        return 0;
+    }
+    //qDebug() << "QIcuCodec: Standard name for " << name << "is" << standardName;
+    ucnv_close(conv);
+
+
+    QTextCodec *c = new QIcuCodec(standardName);
+    if (cache)
+        cache->insert(standardName, c);
+    return c;
+}
+
+
+QTextCodec *QIcuCodec::codecForMib(int mib)
+{
+    for (int i = 0; i < mibToNameSize; ++i) {
+        if (mibToName[i].mib == mib)
+            return codecForName(mibToNameTable + mibToName[i].index);
+    }
+    return 0;
+}
+
+
+QIcuCodec::QIcuCodec(const char *name)
+    : m_name(name)
+{
+}
+
+QIcuCodec::~QIcuCodec()
+{
+}
+
+UConverter *QIcuCodec::getConverter(QTextCodec::ConverterState *state) const
+{
+    UConverter *conv = 0;
+    if (state) {
+        if (!state->d) {
+            // first time
+            state->flags |= QTextCodec::FreeFunction;
+            QTextCodecUnalignedPointer::encode(state->state_data, qIcuCodecStateFree);
+            UErrorCode error = U_ZERO_ERROR;
+            state->d = ucnv_open(m_name, &error);
+            ucnv_setSubstChars(static_cast<UConverter *>(state->d),
+                               state->flags & QTextCodec::ConvertInvalidToNull ? "\0" : "?", 1, &error);
+            if (U_FAILURE(error))
+                qDebug() << "getConverter(state) ucnv_open failed" << m_name << u_errorName(error);
+        }
+        conv = static_cast<UConverter *>(state->d);
+    }
+    if (!conv) {
+        // stateless conversion
+        UErrorCode error = U_ZERO_ERROR;
+        conv = ucnv_open(m_name, &error);
+        ucnv_setSubstChars(conv, "?", 1, &error);
+        if (U_FAILURE(error))
+            qDebug() << "getConverter(no state) ucnv_open failed" << m_name << u_errorName(error);
+    }
+    return conv;
+}
+
+QString QIcuCodec::convertToUnicode(const char *chars, int length, QTextCodec::ConverterState *state) const
+{
+    UConverter *conv = getConverter(state);
+
+    QString string(length + 2, Qt::Uninitialized);
+
+    const char *end = chars + length;
+    int convertedChars = 0;
+    while (1) {
+        UChar *uc = (UChar *)string.data();
+        UChar *ucEnd = uc + string.length();
+        uc += convertedChars;
+        UErrorCode error = U_ZERO_ERROR;
+        ucnv_toUnicode(conv,
+                       &uc, ucEnd,
+                       &chars, end,
+                       0, false, &error);
+        if (!U_SUCCESS(error) && error != U_BUFFER_OVERFLOW_ERROR) {
+            qDebug() << "convertToUnicode failed:" << u_errorName(error);
+            break;
+        }
+
+        convertedChars = uc - (UChar *)string.data();
+        if (chars >= end)
+            break;
+        string.resize(string.length()*2);
+    }
+    string.resize(convertedChars);
+
+    if (!state)
+        ucnv_close(conv);
+    return string;
+}
+
+
+QByteArray QIcuCodec::convertFromUnicode(const QChar *unicode, int length, QTextCodec::ConverterState *state) const
+{
+    UConverter *conv = getConverter(state);
+
+    int requiredLength = UCNV_GET_MAX_BYTES_FOR_STRING(length, ucnv_getMaxCharSize(conv));
+    QByteArray string(requiredLength, Qt::Uninitialized);
+
+    const UChar *uc = (const UChar *)unicode;
+    const UChar *end = uc + length;
+    int convertedChars = 0;
+    while (1) {
+        char *ch = (char *)string.data();
+        char *chEnd = ch + string.length();
+        ch += convertedChars;
+        UErrorCode error = U_ZERO_ERROR;
+        ucnv_fromUnicode(conv,
+                         &ch, chEnd,
+                         &uc, end,
+                         0, false, &error);
+        if (!U_SUCCESS(error))
+            qDebug() << "convertFromUnicode failed:" << u_errorName(error);
+        convertedChars = ch - string.data();
+        if (uc >= end)
+            break;
+        string.resize(string.length()*2);
+    }
+    string.resize(convertedChars);
+
+    if (!state)
+        ucnv_close(conv);
+
+    return string;
+}
+
+
+QByteArray QIcuCodec::name() const
+{
+    return m_name;
+}
+
+
+QList<QByteArray> QIcuCodec::aliases() const
+{
+    UErrorCode error = U_ZERO_ERROR;
+
+    int n = ucnv_countAliases(m_name, &error);
+
+    QList<QByteArray> aliases;
+    for (int i = 0; i < n; ++i) {
+        const char *a = ucnv_getAlias(m_name, i, &error);
+        // skip the canonical name
+        if (!a || !qstrcmp(a, m_name))
+            continue;
+        aliases += a;
+    }
+
+    return aliases;
+}
+
+
+int QIcuCodec::mibEnum() const
+{
+    for (int i = 0; i < mibToNameSize; ++i) {
+        if (m_name == (mibToNameTable + mibToName[i].index))
+            return mibToName[i].mib;
+    }
+
+    return 0;
+}
+
+QT_END_NAMESPACE
+
diff --git a/src/corelib/codecs/qicucodec_p.h b/src/corelib/codecs/qicucodec_p.h
new file mode 100644
index 0000000000..5f30eac3cc
--- /dev/null
+++ b/src/corelib/codecs/qicucodec_p.h
@@ -0,0 +1,91 @@
+/****************************************************************************
+**
+** Copyright (C) 2012 Nokia Corporation and/or its subsidiary(-ies).
+** Contact: http://www.qt-project.org/
+**
+** This file is part of the QtCore module of the Qt Toolkit.
+**
+** $QT_BEGIN_LICENSE:LGPL$
+** GNU Lesser General Public License Usage
+** 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 1.1, 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.
+**
+** Other Usage
+** Alternatively, this file may be used in accordance with the terms and
+** conditions contained in a signed written agreement between you and Nokia.
+**
+**
+**
+**
+**
+**
+** $QT_END_LICENSE$
+**
+****************************************************************************/
+
+#ifndef QICUCODEC_P_H
+#define QICUCODEC_P_H
+
+#include "QtCore/qtextcodec.h"
+
+//
+//  W A R N I N G
+//  -------------
+//
+// This file is not part of the Qt API.  It exists purely as an
+// implementation detail.  This header file may change from version to
+// version without notice, or even be removed.
+//
+// We mean it.
+//
+
+extern "C" {
+    typedef struct UConverter UConverter;
+}
+
+QT_BEGIN_NAMESPACE
+
+class QIcuCodec : public QTextCodec
+{
+public:
+    static QList<QByteArray> availableCodecs();
+    static QList<int> availableMibs();
+
+    static QTextCodec *codecForName(const char *name);
+    static QTextCodec *codecForMib(int mib);
+
+    QString convertToUnicode(const char *, int, ConverterState *) const;
+    QByteArray convertFromUnicode(const QChar *, int, ConverterState *) const;
+
+    QByteArray name() const;
+    QList<QByteArray> aliases() const;
+    int mibEnum() const;
+
+private:
+    QIcuCodec(const char *name);
+    ~QIcuCodec();
+
+    UConverter *getConverter(QTextCodec::ConverterState *state) const;
+
+    const char *m_name;
+};
+
+QT_END_NAMESPACE
+
+#endif
diff --git a/src/corelib/codecs/qtextcodec.cpp b/src/corelib/codecs/qtextcodec.cpp
index b379b2b111..b3607d421f 100644
--- a/src/corelib/codecs/qtextcodec.cpp
+++ b/src/corelib/codecs/qtextcodec.cpp
@@ -65,12 +65,15 @@
 
 #include "qutfcodec_p.h"
 #include "qlatincodec_p.h"
+
 #if !defined(QT_BOOTSTRAPPED)
-#  include "qsimplecodec_p.h"
-#endif
-#if !defined(QT_BOOTSTRAPPED) && !defined(QT_NO_BIG_CODECS)
 #  include "qtsciicodec_p.h"
 #  include "qisciicodec_p.h"
+#if defined(QT_USE_ICU)
+#include "qicucodec_p.h"
+#else
+#  include "qsimplecodec_p.h"
+#if !defined(QT_NO_BIG_CODECS)
 #  ifndef Q_OS_INTEGRITY
 #    include "qgb18030codec_p.h"
 #    include "qeucjpcodec_p.h"
@@ -79,7 +82,11 @@
 #    include "qeuckrcodec_p.h"
 #    include "qbig5codec_p.h"
 #  endif // !Q_OS_INTEGRITY
-#endif // !QT_BOOTSTRAPPED && !QT_NO_BIG_CODECS
+#endif // !QT_NO_BIG_CODECS
+
+#endif // QT_USE_ICU
+#endif // QT_BOOTSTRAPPED
+
 #include "qmutex.h"
 
 #include <stdlib.h>
@@ -140,6 +147,9 @@ static QTextCodec *checkForCodec(const QByteArray &name) {
 #endif
 
 Q_GLOBAL_STATIC_WITH_ARGS(QMutex, textCodecsMutex, (QMutex::Recursive));
+QMutex *qTextCodecsMutex() { return textCodecsMutex(); }
+
+static void setup();
 
 // this returns the codec the method sets up as locale codec to
 // avoid a race condition in codecForLocale() when
@@ -147,10 +157,12 @@ Q_GLOBAL_STATIC_WITH_ARGS(QMutex, textCodecsMutex, (QMutex::Recursive));
 static QTextCodec *setupLocaleMapper()
 {
     QCoreGlobalData *globalData = QCoreGlobalData::instance();
+
     QMutexLocker locker(textCodecsMutex());
     if (globalData->codecForLocale)
         // already setup
         return globalData->codecForLocale;
+    setup();
 
 #if !defined(QT_BOOTSTRAPPED)
     QCoreApplicationPrivate::initLocale();
@@ -250,6 +262,10 @@ static QTextCodec *setupLocaleMapper()
 static void setup()
 {
 #if !defined(QT_NO_CODECS) && !defined(QT_BOOTSTRAPPED)
+    (void)new QTsciiCodec;
+    for (int i = 0; i < 9; ++i)
+        (void)new QIsciiCodec(i);
+#if !defined(QT_USE_ICU)
     for (int i = 0; i < QSimpleTextCodec::numSimpleCodecs; ++i)
         (void)new QSimpleTextCodec(i);
 
@@ -268,13 +284,14 @@ static void setup()
     (void)new QBig5hkscsCodec;
 #    endif // !Q_OS_INTEGRITY
 
-    (void)new QTsciiCodec;
-    for (int i = 0; i < 9; ++i)
-        (void)new QIsciiCodec(i);
 #  endif // !QT_NO_BIG_CODECS
 #endif // !QT_NO_CODECS && !QT_BOOTSTRAPPED
+#endif
 
 #if !defined(QT_BOOTSTRAPPED)
+#if !defined(QT_NO_ICONV)
+    (void) new QIconvCodec;
+#endif
 #if defined(Q_OS_WIN32) || defined(Q_OS_WINCE)
     (void) new QWindowsLocalCodec;
 #endif // Q_OS_WIN32
@@ -493,9 +510,12 @@ QTextCodec *QTextCodec::codecForName(const QByteArray &name)
     QMutexLocker locker(textCodecsMutex());
 
     QCoreGlobalData *globalData = QCoreGlobalData::instance();
+    if (!globalData)
+        return 0;
     if (globalData->allCodecs.isEmpty())
         setup();
 
+#ifndef QT_USE_ICU
     QTextCodecCache *cache = &globalData->codecCache;
     QTextCodec *codec;
     if (cache) {
@@ -521,6 +541,9 @@ QTextCodec *QTextCodec::codecForName(const QByteArray &name)
     }
 
     return 0;
+#else
+    return QIcuCodec::codecForName(name);
+#endif
 }
 
 
@@ -533,6 +556,8 @@ QTextCodec* QTextCodec::codecForMib(int mib)
     QMutexLocker locker(textCodecsMutex());
 
     QCoreGlobalData *globalData = QCoreGlobalData::instance();
+    if (!globalData)
+        return 0;
     if (globalData->allCodecs.isEmpty())
         setup();
 
@@ -556,6 +581,10 @@ QTextCodec* QTextCodec::codecForMib(int mib)
         }
     }
 
+#ifdef QT_USE_ICU
+    return QIcuCodec::codecForMib(mib);
+#endif
+
     return 0;
 }
 
@@ -570,6 +599,9 @@ QTextCodec* QTextCodec::codecForMib(int mib)
 */
 QList<QByteArray> QTextCodec::availableCodecs()
 {
+#ifdef QT_USE_ICU
+    return QIcuCodec::availableCodecs();
+#else
     QMutexLocker locker(textCodecsMutex());
 
     QCoreGlobalData *globalData = QCoreGlobalData::instance();
@@ -584,6 +616,7 @@ QList<QByteArray> QTextCodec::availableCodecs()
     }
 
     return codecs;
+#endif
 }
 
 /*!
@@ -594,6 +627,9 @@ QList<QByteArray> QTextCodec::availableCodecs()
 */
 QList<int> QTextCodec::availableMibs()
 {
+#ifdef QT_USE_ICU
+    return QIcuCodec::availableMibs();
+#else
     QMutexLocker locker(textCodecsMutex());
 
     QCoreGlobalData *globalData = QCoreGlobalData::instance();
@@ -606,6 +642,7 @@ QList<int> QTextCodec::availableMibs()
         codecs += globalData->allCodecs.at(i)->mibEnum();
 
     return codecs;
+#endif
 }
 
 /*!
@@ -638,6 +675,8 @@ void QTextCodec::setCodecForLocale(QTextCodec *c)
 QTextCodec* QTextCodec::codecForLocale()
 {
     QCoreGlobalData *globalData = QCoreGlobalData::instance();
+    if (!globalData)
+        return 0;
 
     QTextCodec *codec = globalData->codecForLocale;
     if (!globalData->codecForLocale)
diff --git a/src/corelib/kernel/qcoreglobaldata_p.h b/src/corelib/kernel/qcoreglobaldata_p.h
index 39281aa6e8..1ff7d17019 100644
--- a/src/corelib/kernel/qcoreglobaldata_p.h
+++ b/src/corelib/kernel/qcoreglobaldata_p.h
@@ -59,6 +59,7 @@
 #include "QtCore/qhash.h"
 #include "QtCore/qbytearray.h"
 #include "QtCore/qtextcodec.h"
+#include "QtCore/qmutex.h"
 
 QT_BEGIN_NAMESPACE