Avoid using iconv for text conversion where possible

Try to use a builtin codec as codecForLocale() if possible
first. Fall back and instantiate the iconv codec only
if that failed.

In addition, make sure we initialize the locale correctly
before we try to setup the codec.

Change-Id: I86d635f9d11e8ff93093f162e79fb37f3d85731b
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>

diff --git a/src/corelib/codecs/qiconvcodec.cpp b/src/corelib/codecs/qiconvcodec.cpp
index be22ff9..84a8e8d 100644 (file)
--- a/src/corelib/codecs/qiconvcodec.cpp
+++ b/src/corelib/codecs/qiconvcodec.cpp
@@ -88,8 +88,6 @@ static Ptr_iconv_close ptr_iconv_close = 0;
 
 QT_BEGIN_NAMESPACE
 
-extern bool qt_locale_initialized;
-
 QIconvCodec::QIconvCodec()
     : utf16Codec(0)
 {
@@ -190,7 +188,7 @@ QString QIconvCodec::convertToUnicode(const char* chars, int len, ConverterState
         }
     } else {
         QThreadStorage<QIconvCodec::IconvState *> *ts = toUnicodeState();
-        if (!qt_locale_initialized || !ts) {
+        if (!ts) {
             // we're running after the Q_GLOBAL_STATIC has been deleted
             // or before the QCoreApplication initialization
             // bad programmer, no cookie for you
@@ -346,7 +344,7 @@ QByteArray QIconvCodec::convertFromUnicode(const QChar *uc, int len, ConverterSt
 
     IconvState *temporaryState = 0;
     QThreadStorage<QIconvCodec::IconvState *> *ts = fromUnicodeState();
-    IconvState *&state = (qt_locale_initialized && ts) ? ts->localData() : temporaryState;
+    IconvState *&state = ts ? ts->localData() : temporaryState;
     if (!state) {
         iconv_t cd = QIconvCodec::createIconv_t(0, UTF16);
         if (cd != reinterpret_cast<iconv_t>(-1)) {

diff --git a/src/corelib/codecs/qtextcodec.cpp b/src/corelib/codecs/qtextcodec.cpp
index 916aefa..23cf95a 100644 (file)
--- a/src/corelib/codecs/qtextcodec.cpp
+++ b/src/corelib/codecs/qtextcodec.cpp
@@ -49,6 +49,9 @@
 #include "qfile.h"
 #include "qstringlist.h"
 #include "qvarlengtharray.h"
+#if !defined(QT_BOOTSTRAPPED)
+#include <private/qcoreapplication_p.h>
+#endif
 
 #ifdef Q_OS_UNIX
 #  include "qiconvcodec_p.h"
@@ -82,8 +85,7 @@
 #include <stdlib.h>
 #include <ctype.h>
 #include <locale.h>
-#if defined (_XOPEN_UNIX) && !defined(Q_OS_QNX) && \
-    !defined(Q_OS_OSF) && !defined(Q_OS_LINUX_ANDROID)
+#if defined (_XOPEN_UNIX) && !defined(Q_OS_QNX) && !defined(Q_OS_OSF) && !defined(Q_OS_LINUX_ANDROID)
 # include <langinfo.h>
 #endif
 
@@ -477,14 +479,11 @@ static bool try_locale_list(const char * const locale[], const QByteArray &lang)
 static const char * const probably_koi8_rlocales[] = {
     "ru", "ru_SU", "ru_RU", "russian", 0 };
 
-static QTextCodec * ru_RU_hack(const char * i) {
+static QTextCodec * ru_RU_hack(const char * i)
+{
     QTextCodec * ru_RU_codec = 0;
 
-#if !defined(QT_NO_SETLOCALE)
     QByteArray origlocale(setlocale(LC_CTYPE, i));
-#else
-    QByteArray origlocale(i);
-#endif
     // unicode   koi8r   latin5   name
     // 0x044E    0xC0    0xEE     CYRILLIC SMALL LETTER YU
     // 0x042E    0xE0    0xCE     CYRILLIC CAPITAL LETTER YU
@@ -500,16 +499,11 @@ static QTextCodec * ru_RU_hack(const char * i) {
         qWarning("QTextCodec: Using KOI8-R, probe failed (%02x %02x %s)",
                   koi8r, latin5, i);
     }
-#if !defined(QT_NO_SETLOCALE)
     setlocale(LC_CTYPE, origlocale.constData());
-#endif
 
     return ru_RU_codec;
 }
 
-#endif
-
-#if !defined(Q_OS_WIN32) && !defined(Q_OS_WINCE)
 static QTextCodec *checkForCodec(const QByteArray &name) {
     QTextCodec *c = QTextCodec::codecForName(name);
     if (!c) {
@@ -531,16 +525,25 @@ static void setupLocaleMapper()
     localeMapper = QTextCodec::codecForName("System");
 #else
 
-#ifndef QT_NO_ICONV
-    localeMapper = QTextCodec::codecForName("System");
+#if !defined(QT_BOOTSTRAPPED)
+    QCoreApplicationPrivate::initLocale();
 #endif
-
-#if defined (_XOPEN_UNIX) && !defined(Q_OS_QNX) && \
-    !defined(Q_OS_OSF) && !defined(Q_OS_LINUX_ANDROID)
+    // First try getting the codecs name from nl_langinfo and see
+    // if we have a builtin codec for it.
+    // Only fall back to using iconv if we can't find a builtin codec
+    // This is because the builtin utf8 codec is around 5 times faster
+    // then the using QIconvCodec
+
+#if defined (_XOPEN_UNIX) && !defined(Q_OS_QNX) && !defined(Q_OS_OSF) && !defined(Q_OS_LINUX_ANDROID)
+    char *charset = nl_langinfo(CODESET);
+    if (charset)
+        localeMapper = QTextCodec::codecForName(charset);
+#endif
+#if !defined(QT_NO_ICONV) && !defined(QT_BOOTSTRAPPED)
     if (!localeMapper) {
-        char *charset = nl_langinfo (CODESET);
-        if (charset)
-            localeMapper = QTextCodec::codecForName(charset);
+        // no builtin codec for the locale found, let's try using iconv
+        (void) new QIconvCodec();
+        localeMapper = QTextCodec::codecForName("System");
     }
 #endif
 
@@ -556,11 +559,7 @@ static void setupLocaleMapper()
         // 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.
-#if !defined(QT_NO_SETLOCALE)
         const QByteArray ctype = setlocale(LC_CTYPE, 0);
-#else
-        const QByteArray ctype;
-#endif
 
         // Get the first nonempty value from $LC_ALL, $LC_CTYPE, and $LANG
         // environment variables.
@@ -723,13 +722,6 @@ static void setup()
     (void)new QLatin1Codec;
     (void)new QUtf8Codec;
 
-#if !defined(Q_OS_INTEGRITY)
-#if defined(Q_OS_UNIX) && !defined(QT_NO_ICONV) && !defined(QT_BOOTSTRAPPED)
-    // QIconvCodec depends on the UTF-16 codec, so it needs to be created last
-    (void) new QIconvCodec();
-#endif
-#endif
-
     if (!localeMapper)
         setupLocaleMapper();
 }
@@ -1105,9 +1097,10 @@ void QTextCodec::setCodecForLocale(QTextCodec *c)
     Returns a pointer to the codec most suitable for this locale.
 
     On Windows, the codec will be based on a system locale. On Unix
-    systems, starting with Qt 4.2, the codec will be using the \e
-    iconv library. Note that in both cases the codec's name will be
-    "System".
+    systems, the codec will might fall back to using the \e iconv
+    library if no builtin codec for the locale can be found.
+
+    Note that in these cases the codec's name will be "System".
 */
 
 QTextCodec* QTextCodec::codecForLocale()

diff --git a/src/corelib/kernel/qcoreapplication.cpp b/src/corelib/kernel/qcoreapplication.cpp
index c453071..ad8c35b 100644 (file)
--- a/src/corelib/kernel/qcoreapplication.cpp
+++ b/src/corelib/kernel/qcoreapplication.cpp
@@ -230,7 +230,7 @@ bool QCoreApplicationPrivate::is_app_running = false;
  // app closing down if true
 bool QCoreApplicationPrivate::is_app_closing = false;
 // initialized in qcoreapplication and in qtextstream autotest when setlocale is called.
-Q_CORE_EXPORT bool qt_locale_initialized = false;
+static bool qt_locale_initialized = false;
 
 Q_CORE_EXPORT uint qGlobalPostedEventsCount()
 {
@@ -405,6 +405,17 @@ QString qAppName()
     return QCoreApplication::instance()->d_func()->appName();
 }
 
+void QCoreApplicationPrivate::initLocale()
+{
+    if (qt_locale_initialized)
+        return;
+    qt_locale_initialized = true;
+#ifdef Q_OS_UNIX
+    setlocale(LC_ALL, "");
+#endif
+}
+
+
 /*!
     \class QCoreApplication
     \brief The QCoreApplication class provides an event loop for console Qt
@@ -543,10 +554,7 @@ void QCoreApplication::init()
 {
     Q_D(QCoreApplication);
 
-#ifdef Q_OS_UNIX
-    setlocale(LC_ALL, "");                // use correct char set mapping
-    qt_locale_initialized = true;
-#endif
+    QCoreApplicationPrivate::initLocale();
 
     Q_ASSERT_X(!self, "QCoreApplication", "there should be only one application object");
     QCoreApplication::self = this;

diff --git a/src/corelib/kernel/qcoreapplication_p.h b/src/corelib/kernel/qcoreapplication_p.h
index 112b313..15aca39 100644 (file)
--- a/src/corelib/kernel/qcoreapplication_p.h
+++ b/src/corelib/kernel/qcoreapplication_p.h
@@ -89,6 +89,8 @@ public:
     static QString macMenuBarName();
 #endif
 
+    static void initLocale();
+
     QAtomicInt quitLockRef;
     void ref();
     void deref();