summaryrefslogtreecommitdiffstats
path: root/src/corelib/codecs/qutfcodec.cpp
diff options
context:
space:
mode:
authorThiago Macieira <thiago.macieira@nokia.com>2010-03-05 13:10:57 (GMT)
committerThiago Macieira <thiago.macieira@nokia.com>2010-03-05 15:57:27 (GMT)
commit8b30b72948d5f16d59f8799f3889f563e2c6e24d (patch)
tree6279bd2f3faeb0be880225dc787858b98d3a935f /src/corelib/codecs/qutfcodec.cpp
parent4f02ca5c9458299185d447da87058d6e8ca1b260 (diff)
downloadQt-8b30b72948d5f16d59f8799f3889f563e2c6e24d.zip
Qt-8b30b72948d5f16d59f8799f3889f563e2c6e24d.tar.gz
Qt-8b30b72948d5f16d59f8799f3889f563e2c6e24d.tar.bz2
Make the UTF-8 encoder/decoder not accept Unicode non-characters
Reviewed-By: Denis Dzyubenko
Diffstat (limited to 'src/corelib/codecs/qutfcodec.cpp')
-rw-r--r--src/corelib/codecs/qutfcodec.cpp28
1 files changed, 25 insertions, 3 deletions
diff --git a/src/corelib/codecs/qutfcodec.cpp b/src/corelib/codecs/qutfcodec.cpp
index 742b2e7..233bd8f 100644
--- a/src/corelib/codecs/qutfcodec.cpp
+++ b/src/corelib/codecs/qutfcodec.cpp
@@ -48,6 +48,19 @@ QT_BEGIN_NAMESPACE
enum { Endian = 0, Data = 1 };
+static inline bool isUnicodeNonCharacter(uint ucs4)
+{
+ // Unicode has a couple of "non-characters" that one can use internally,
+ // but are not allowed to be used for text interchange.
+ //
+ // Those are the last two entries each Unicode Plane (U+FFFE, U+FFFF,
+ // U+1FFFE, U+1FFFF, etc.) as well as the entries between U+FDD0 and
+ // U+FDEF (inclusive)
+
+ return (ucs4 & 0xfffe) == 0xfffe
+ || (ucs4 - 0xfdd0U) < 16;
+}
+
QByteArray QUtf8::convertFromUnicode(const QChar *uc, int len, QTextCodec::ConverterState *state)
{
uchar replacement = '?';
@@ -106,6 +119,14 @@ QByteArray QUtf8::convertFromUnicode(const QChar *uc, int len, QTextCodec::Conve
if (u < 0x0800) {
*cursor++ = 0xc0 | ((uchar) (u >> 6));
} else {
+ // is it one of the Unicode non-characters?
+ if (isUnicodeNonCharacter(u)) {
+ *cursor = replacement;
+ ++ch;
+ ++invalid;
+ continue;
+ }
+
if (u > 0xffff) {
*cursor++ = 0xf0 | ((uchar) (u >> 18));
*cursor++ = 0x80 | (((uchar) (u >> 12)) & 0x3f);
@@ -172,15 +193,16 @@ QString QUtf8::convertToUnicode(const char *chars, int len, QTextCodec::Converte
--need;
if (!need) {
// utf-8 bom composes into 0xfeff code point
+ bool nonCharacter;
if (!headerdone && uc == 0xfeff) {
// dont do anything, just skip the BOM
- } else if (uc > 0xffff && uc < 0x110000) {
+ } else if (!(nonCharacter = isUnicodeNonCharacter(uc)) && uc > 0xffff && uc < 0x110000) {
// surrogate pair
Q_ASSERT((qch - (ushort*)result.unicode()) + 2 < result.length());
*qch++ = QChar::highSurrogate(uc);
*qch++ = QChar::lowSurrogate(uc);
- } else if ((uc < min_uc) || (uc >= 0xd800 && uc <= 0xdfff) || (uc >= 0xfffe)) {
- // error: overlong sequence, UTF16 surrogate or BOM
+ } else if ((uc < min_uc) || (uc >= 0xd800 && uc <= 0xdfff) || nonCharacter || uc >= 0x110000) {
+ // error: overlong sequence, UTF16 surrogate or non-character
*qch++ = replacement;
++invalid;
} else {