From 99baa0d4401516cff978cab8a7102fc0a25731a9 Mon Sep 17 00:00:00 2001
From: Thiago Macieira <thiago.macieira@intel.com>
Date: Sat, 20 Oct 2018 21:45:10 -0700
Subject: Optimize QTextCodec::codecForUtfText a little

Instead of doing byte comparisons, let the compiler do 16- and 32-bit
comparisons on its own.

Change-Id: If7e743cf8476463880ccfffd155f8629991b0b87
Reviewed-by: Olivier Goffart (Woboq GmbH) <ogoffart@woboq.com>
Reviewed-by: Tobias Hunger <tobias.hunger@qt.io>
---
 src/corelib/codecs/qtextcodec.cpp | 28 ++++++++++++++--------------
 1 file changed, 14 insertions(+), 14 deletions(-)

(limited to 'src/corelib/codecs')

diff --git a/src/corelib/codecs/qtextcodec.cpp b/src/corelib/codecs/qtextcodec.cpp
index a2a128158e..e5c33d9d59 100644
--- a/src/corelib/codecs/qtextcodec.cpp
+++ b/src/corelib/codecs/qtextcodec.cpp
@@ -45,8 +45,9 @@
 #ifndef QT_NO_TEXTCODEC
 
 #include "qbytearraymatcher.h"
-#include "qlist.h"
+#include "qendian.h"
 #include "qfile.h"
+#include "qlist.h"
 #include "qstringlist.h"
 #include "qvarlengtharray.h"
 #if !defined(QT_BOOTSTRAPPED)
@@ -1183,32 +1184,31 @@ QTextCodec *QTextCodec::codecForHtml(const QByteArray &ba)
 QTextCodec *QTextCodec::codecForUtfText(const QByteArray &ba, QTextCodec *defaultCodec)
 {
     const int arraySize = ba.size();
+    const uchar *buf = reinterpret_cast<const uchar *>(ba.constData());
+    const uint bom = 0xfeff;
 
     if (arraySize > 3) {
-        if ((uchar)ba[0] == 0x00
-            && (uchar)ba[1] == 0x00
-            && (uchar)ba[2] == 0xFE
-            && (uchar)ba[3] == 0xFF)
+        uint uc = qFromUnaligned<uint>(buf);
+        if (uc == qToBigEndian(bom))
             return QTextCodec::codecForMib(1018); // utf-32 be
-        else if ((uchar)ba[0] == 0xFF
-                 && (uchar)ba[1] == 0xFE
-                 && (uchar)ba[2] == 0x00
-                 && (uchar)ba[3] == 0x00)
+        else if (uc == qToLittleEndian(bom))
             return QTextCodec::codecForMib(1019); // utf-32 le
     }
 
     if (arraySize < 2)
         return defaultCodec;
-    if ((uchar)ba[0] == 0xfe && (uchar)ba[1] == 0xff)
+
+    ushort uc = qFromUnaligned<ushort>(buf);
+    if (uc == qToBigEndian(ushort(bom)))
         return QTextCodec::codecForMib(1013); // utf16 be
-    else if ((uchar)ba[0] == 0xff && (uchar)ba[1] == 0xfe)
+    else if (uc == qToLittleEndian(ushort(bom)))
         return QTextCodec::codecForMib(1014); // utf16 le
 
     if (arraySize < 3)
         return defaultCodec;
-    if ((uchar)ba[0] == 0xef
-        && (uchar)ba[1] == 0xbb
-        && (uchar)ba[2] == 0xbf)
+
+    static const char utf8bom[] = "\xef\xbb\xbf";
+    if (memcmp(buf, utf8bom, sizeof(utf8bom) - 1) == 0)
         return QTextCodec::codecForMib(106); // utf-8
 
     return defaultCodec;
-- 
cgit v1.2.3