Remove QTextCodec dependency from QClipboard

QClipboard used QTextCodec to convert the war clipboard data to
a QString. HTML is nowadays always encoded as utf8, and we were only
supporting utf based encodings for other text.

Add a qFromUtfEncoded() to our UTF helpers that auto detects utf16
and utf32 byte order marks, and assumes utf8 otherwise, to keep
this compatible with what we have been doing in Qt 5.

Change-Id: I5a9fccb67a88dff27cbbdecff9bd548d31aa1c6c
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>

2 files changed, 27 insertions, 0 deletions

diff --git a/src/corelib/codecs/qutfcodec.cpp b/src/corelib/codecs/qutfcodec.cpp
index 8561f908b9..9419ce8d84 100644
--- a/src/corelib/codecs/qutfcodec.cpp
+++ b/src/corelib/codecs/qutfcodec.cpp
@@ -972,6 +972,25 @@ QString QUtf32::convertToUnicode(const char *chars, int len, QTextCodec::Convert
     return result;
 }
 
+QString qFromUtfEncoded(const QByteArray &ba)
+{
+    const int arraySize = ba.size();
+    const uchar *buf = reinterpret_cast<const uchar *>(ba.constData());
+    const uint bom = 0xfeff;
+
+    if (arraySize > 3) {
+        uint uc = qFromUnaligned<uint>(buf);
+        if (uc == qToBigEndian(bom) || uc == qToLittleEndian(bom))
+            return QUtf32::convertToUnicode(ba.constData(), ba.length(), nullptr); // utf-32
+    }
+
+    if (arraySize > 1) {
+        ushort uc = qFromUnaligned<ushort>(buf);
+        if (uc == qToBigEndian(ushort(bom)) || qToLittleEndian(ushort(bom)))
+            return QUtf16::convertToUnicode(ba.constData(), ba.length(), nullptr); // utf-16
+    }
+    return QUtf8::convertToUnicode(ba.constData(), ba.length());
+}
 
 #if QT_CONFIG(textcodec)
 
diff --git a/src/corelib/codecs/qutfcodec_p.h b/src/corelib/codecs/qutfcodec_p.h
index 083e16317a..b1c7a23d4f 100644
--- a/src/corelib/codecs/qutfcodec_p.h
+++ b/src/corelib/codecs/qutfcodec_p.h
@@ -54,6 +54,7 @@
 
 #include <QtCore/qstring.h>
 #include <QtCore/qlist.h>
+#include <QtCore/qendian.h>
 
 #if QT_CONFIG(textcodec)
 #include "QtCore/qtextcodec.h"
@@ -317,6 +318,13 @@ struct QUtf32
     static QByteArray convertFromUnicode(const QChar *, int, QTextCodec::ConverterState *, DataEndianness = DetectEndianness);
 };
 
+/*
+ Converts from different utf encodings looking at a possible byte order mark at the
+ beginning of the string. If no BOM exists, utf-8 is assumed.
+ */
+QString Q_CORE_EXPORT qFromUtfEncoded(const QByteArray &ba);
+
+
 #if QT_CONFIG(textcodec)
 
 class QUtf8Codec : public QTextCodec {