Bommels05/Mozilla
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

utf-8 decoder no longer accepts overlong sequences: bug 50702 r,sr=ftang

Browse Source 
git-svn-id: svn://10.0.0.236/trunk@86130 18797224-902f-48f8-a5cc-f745e15eee43
This commit is contained in:
jgmyers%netscape.com
2001-02-03 00:08:40 +00:00
parent fe3c220dae
commit 3f0e9b60a8
2 changed files with 16 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
mozilla/intl/uconv/ucvlatin/nsUTF8ToUnicode.cpp
View File

@@ -89,26 +89,31 @@ NS_IMETHODIMP nsUTF8ToUnicode::GetMaxLength(const char * aSrc,
mUcs4 = (PRUint32)(*in);
mUcs4 = (mUcs4 << 6) & 0x000007C0L;
mState=1;
mMinUcs4 = 0x00000080;
} else if( 0xE0 == (0xF0 & (*in))) {
// 3 bytes UTF8
mUcs4 = (PRUint32)(*in);
mUcs4 = (mUcs4 << 12) & 0x0000F000L;
mState=2;
mMinUcs4 = 0x00000800;
} else if( 0xF0 == (0xF8 & (*in))) {
// 4 bytes UTF8
mUcs4 = (PRUint32)(*in);
mUcs4 = (mUcs4 << 18) & 0x001F0000L;
mState=3;
mMinUcs4 = 0x00010000;
} else if( 0xF8 == (0xFC & (*in))) {
// 5 bytes UTF8
mUcs4 = (PRUint32)(*in);
mUcs4 = (mUcs4 << 24) & 0x03000000L;
mState=4;
mMinUcs4 = 0x00200000;
} else if( 0xFC == (0xFE & (*in))) {
// 6 bytes UTF8
mUcs4 = (PRUint32)(*in);
mUcs4 = (mUcs4 << 30) & 0x40000000L;
mState=5;
mMinUcs4 = 0x04000000;
} else {
//NS_ASSERTION(0, "The input string is not in utf8");
@@ -130,7 +135,16 @@ NS_IMETHODIMP nsUTF8ToUnicode::GetMaxLength(const char * aSrc,
mUcs4 |= tmp;
if(0 == --mState)
{
if(mUcs4 >= 0x00010000) {
if (mUcs4 < mMinUcs4) {
// Overlong sequence
*out++ = 0xFFFD;
} else if (mUcs4 >= 0xD800 && mUcs4 <= 0xDFFF) {
// Surrogates
*out++ = 0xFFFD;
} else if (mUcs4 == 0xFFFE || mUcs4 == 0xFFFF) {
// prohibited characters
*out++ = 0xFFFD;
} else if(mUcs4 >= 0x00010000) {
if(mUcs4 >= 0x00110000) {
*out++ = 0xFFFD;
} else {

1
mozilla/intl/uconv/ucvlatin/nsUTF8ToUnicode.h
View File

@@ -50,6 +50,7 @@ protected:
PRUint32 mState; // cached expected number of bytes per UTF8 character sequence
PRUint32 mUcs4; // cached Unicode character
PRUint32 mMinUcs4; // minimum Ucs4 value
//--------------------------------------------------------------------
// Subclassing of nsDecoderSupport class [declaration]