diff --git a/src/xercesc/internal/DGXMLScanner.cpp b/src/xercesc/internal/DGXMLScanner.cpp
index 8875f50d8d465b7d4cdf27d61a23a35c9016d846..ef293488606b178b3ca737f23717ba695227dac0 100644
--- a/src/xercesc/internal/DGXMLScanner.cpp
+++ b/src/xercesc/internal/DGXMLScanner.cpp
@@ -2326,23 +2326,13 @@ bool DGXMLScanner::scanAttValue( const XMLAttDef* const attDef
bool firstNonWS = false;
bool gotLeadingSurrogate = false;
bool escaped;
- bool charref_expanded = false;
while (true)
{
try
{
while(true)
{
- // Get another char. Use second char if one is waiting
- if (secondCh)
- {
- nextCh = secondCh;
- secondCh = 0;
- }
- else
- {
- nextCh = fReaderMgr.getNextChar();
- }
+ nextCh = fReaderMgr.getNextChar();
if (!nextCh)
ThrowXML(UnexpectedEOFException, XMLExcepts::Gen_UnexpectedEOF);
@@ -2372,18 +2362,16 @@ bool DGXMLScanner::scanAttValue( const XMLAttDef* const attDef
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
}
-
- // Deal with surrogate pairs
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
+ // Deal with surrogate pairs
// Its a leading surrogate. If we already got one, then
// issue an error, else set leading flag to make sure that
// we look for a trailing next time.
if (gotLeadingSurrogate)
emitError(XMLErrs::Expected2ndSurrogateChar);
- else
+ else
gotLeadingSurrogate = true;
}
else
@@ -2407,22 +2395,17 @@ bool DGXMLScanner::scanAttValue( const XMLAttDef* const attDef
// Its got to at least be a valid XML character
if (!fReaderMgr.getCurrentReader()->isXMLChar(nextCh))
{
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr.getCurrentReader()->isControlChar(nextCh)))
- {
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- emitError(XMLErrs::InvalidCharacterInAttrValue, attrName, tmpBuf);
- }
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ emitError(XMLErrs::InvalidCharacterInAttrValue, attrName, tmpBuf);
}
}
- charref_expanded = false;
gotLeadingSurrogate = false;
}
@@ -2498,6 +2481,9 @@ bool DGXMLScanner::scanAttValue( const XMLAttDef* const attDef
// Else add it to the buffer
toFill.append(nextCh);
+
+ if (secondCh)
+ toFill.append(secondCh);
}
}
catch(const EndOfEntityException&)
@@ -2704,44 +2690,36 @@ void DGXMLScanner::scanCharData(XMLBuffer& toUse)
bool escaped = false;
bool gotLeadingSurrogate = false;
bool notDone = true;
- bool charref_expanded = false;
while (notDone)
{
try
{
while (true)
{
- if (secondCh)
+ // Eat through as many plain content characters as possible without
+ // needing special handling. Moving most content characters here,
+ // in this one call, rather than running the overall loop once
+ // per content character, is a speed optimization.
+ if (curState == State_Waiting && !gotLeadingSurrogate)
{
- nextCh = secondCh;
- secondCh = 0;
+ fReaderMgr.movePlainContentChars(toUse);
}
- else
- {
- // Eat through as many plain content characters as possible without
- // needing special handling. Moving most content characters here,
- // in this one call, rather than running the overall loop once
- // per content character, is a speed optimization.
- if (curState == State_Waiting && !gotLeadingSurrogate)
- {
- fReaderMgr.movePlainContentChars(toUse);
- }
- // Try to get another char from the source
- // The code from here on down covers all contengencies,
- if (!fReaderMgr.getNextCharIfNot(chOpenAngle, nextCh))
- {
- // If we were waiting for a trailing surrogate, its an error
- if (gotLeadingSurrogate)
- emitError(XMLErrs::Expected2ndSurrogateChar);
+ // Try to get another char from the source
+ // The code from here on down covers all contengencies,
+ if (!fReaderMgr.getNextCharIfNot(chOpenAngle, nextCh))
+ {
+ // If we were waiting for a trailing surrogate, its an error
+ if (gotLeadingSurrogate)
+ emitError(XMLErrs::Expected2ndSurrogateChar);
- notDone = false;
- break;
- }
+ notDone = false;
+ break;
}
// Watch for a reference. Note that the escapement mechanism
// is ignored in this content.
+ escaped = false;
if (nextCh == chAmpersand)
{
sendCharData(toUse);
@@ -2754,42 +2732,10 @@ void DGXMLScanner::scanCharData(XMLBuffer& toUse)
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
}
- else
- {
- escaped = false;
- }
-
- // Keep the state machine up to date
- if (!escaped)
- {
- if (nextCh == chCloseSquare)
- {
- if (curState == State_Waiting)
- curState = State_GotOne;
- else if (curState == State_GotOne)
- curState = State_GotTwo;
- }
- else if (nextCh == chCloseAngle)
- {
- if (curState == State_GotTwo)
- emitError(XMLErrs::BadSequenceInCharData);
- curState = State_Waiting;
- }
- else
- {
- curState = State_Waiting;
- }
- }
- else
- {
- curState = State_Waiting;
- }
-
- // Deal with surrogate pairs
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
+ // Deal with surrogate pairs
// Its a leading surrogate. If we already got one, then
// issue an error, else set leading flag to make sure that
// we look for a trailing next time.
@@ -2819,27 +2765,51 @@ void DGXMLScanner::scanCharData(XMLBuffer& toUse)
// Make sure the returned char is a valid XML char
if (!fReaderMgr.getCurrentReader()->isXMLChar(nextCh))
{
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr.getCurrentReader()->isControlChar(nextCh)))
- {
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- emitError(XMLErrs::InvalidCharacter, tmpBuf);
- }
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ emitError(XMLErrs::InvalidCharacter, tmpBuf);
}
}
- charref_expanded = false;
gotLeadingSurrogate = false;
}
+ // Keep the state machine up to date
+ if (!escaped)
+ {
+ if (nextCh == chCloseSquare)
+ {
+ if (curState == State_Waiting)
+ curState = State_GotOne;
+ else if (curState == State_GotOne)
+ curState = State_GotTwo;
+ }
+ else if (nextCh == chCloseAngle)
+ {
+ if (curState == State_GotTwo)
+ emitError(XMLErrs::BadSequenceInCharData);
+ curState = State_Waiting;
+ }
+ else
+ {
+ curState = State_Waiting;
+ }
+ }
+ else
+ {
+ curState = State_Waiting;
+ }
+
// Add this char to the buffer
toUse.append(nextCh);
+
+ if (secondCh)
+ toUse.append(secondCh);
}
}
catch(const EndOfEntityException& toCatch)
diff --git a/src/xercesc/internal/IGXMLScanner2.cpp b/src/xercesc/internal/IGXMLScanner2.cpp
index 293ac5508f1b824977f3a955178ce67b55698668..45297c2450a554c6db7f106223e755c764acc795 100644
--- a/src/xercesc/internal/IGXMLScanner2.cpp
+++ b/src/xercesc/internal/IGXMLScanner2.cpp
@@ -1542,23 +1542,13 @@ bool IGXMLScanner::basicAttrValueScan(const XMLCh* const attrName, XMLBuffer& to
XMLCh secondCh = 0;
bool gotLeadingSurrogate = false;
bool escaped;
- bool charref_expanded = false;
while (true)
{
try
{
while(true)
{
- // Get another char. Use second char if one is waiting
- if (secondCh)
- {
- nextCh = secondCh;
- secondCh = 0;
- }
- else
- {
- nextCh = fReaderMgr.getNextChar();
- }
+ nextCh = fReaderMgr.getNextChar();
if (!nextCh)
ThrowXML(UnexpectedEOFException, XMLExcepts::Gen_UnexpectedEOF);
@@ -1589,12 +1579,10 @@ bool IGXMLScanner::basicAttrValueScan(const XMLCh* const attrName, XMLBuffer& to
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
}
-
- // Deal with surrogate pairs
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
+ // Deal with surrogate pairs
// Its a leading surrogate. If we already got one, then
// issue an error, else set leading flag to make sure that
// we look for a trailing next time.
@@ -1626,23 +1614,17 @@ bool IGXMLScanner::basicAttrValueScan(const XMLCh* const attrName, XMLBuffer& to
// Its got to at least be a valid XML character
else if (!fReaderMgr.getCurrentReader()->isXMLChar(nextCh))
{
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr.getCurrentReader()->isControlChar(nextCh)))
- {
-
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- emitError(XMLErrs::InvalidCharacterInAttrValue, attrName, tmpBuf);
- }
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ emitError(XMLErrs::InvalidCharacterInAttrValue, attrName, tmpBuf);
}
}
- charref_expanded = false;
gotLeadingSurrogate = false;
}
@@ -1655,6 +1637,9 @@ bool IGXMLScanner::basicAttrValueScan(const XMLCh* const attrName, XMLBuffer& to
// Else add it to the buffer
toFill.append(nextCh);
+
+ if (secondCh)
+ toFill.append(secondCh);
}
}
catch(const EndOfEntityException&)
@@ -1705,23 +1690,13 @@ bool IGXMLScanner::scanAttValue( const XMLAttDef* const attDef
bool firstNonWS = false;
bool gotLeadingSurrogate = false;
bool escaped;
- bool charref_expanded = false;
while (true)
{
try
{
while(true)
{
- // Get another char. Use second char if one is waiting
- if (secondCh)
- {
- nextCh = secondCh;
- secondCh = 0;
- }
- else
- {
- nextCh = fReaderMgr.getNextChar();
- }
+ nextCh = fReaderMgr.getNextChar();
if (!nextCh)
ThrowXML(UnexpectedEOFException, XMLExcepts::Gen_UnexpectedEOF);
@@ -1751,12 +1726,10 @@ bool IGXMLScanner::scanAttValue( const XMLAttDef* const attDef
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
}
-
- // Deal with surrogate pairs
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
+ // Deal with surrogate pairs
// Its a leading surrogate. If we already got one, then
// issue an error, else set leading flag to make sure that
// we look for a trailing next time.
@@ -1786,22 +1759,17 @@ bool IGXMLScanner::scanAttValue( const XMLAttDef* const attDef
// Its got to at least be a valid XML character
if (!fReaderMgr.getCurrentReader()->isXMLChar(nextCh))
{
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr.getCurrentReader()->isControlChar(nextCh)))
- {
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- emitError(XMLErrs::InvalidCharacterInAttrValue, attrName, tmpBuf);
- }
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ emitError(XMLErrs::InvalidCharacterInAttrValue, attrName, tmpBuf);
}
}
- charref_expanded = false;
gotLeadingSurrogate = false;
}
@@ -1877,6 +1845,9 @@ bool IGXMLScanner::scanAttValue( const XMLAttDef* const attDef
// Else add it to the buffer
toFill.append(nextCh);
+
+ if (secondCh)
+ toFill.append(secondCh);
}
}
catch(const EndOfEntityException&)
@@ -2093,44 +2064,36 @@ void IGXMLScanner::scanCharData(XMLBuffer& toUse)
bool escaped = false;
bool gotLeadingSurrogate = false;
bool notDone = true;
- bool charref_expanded = false;
while (notDone)
{
try
{
while (true)
{
- if (secondCh)
+ // Eat through as many plain content characters as possible without
+ // needing special handling. Moving most content characters here,
+ // in this one call, rather than running the overall loop once
+ // per content character, is a speed optimization.
+ if (curState == State_Waiting && !gotLeadingSurrogate)
{
- nextCh = secondCh;
- secondCh = 0;
+ fReaderMgr.movePlainContentChars(toUse);
}
- else
- {
- // Eat through as many plain content characters as possible without
- // needing special handling. Moving most content characters here,
- // in this one call, rather than running the overall loop once
- // per content character, is a speed optimization.
- if (curState == State_Waiting && !gotLeadingSurrogate)
- {
- fReaderMgr.movePlainContentChars(toUse);
- }
- // Try to get another char from the source
- // The code from here on down covers all contengencies,
- if (!fReaderMgr.getNextCharIfNot(chOpenAngle, nextCh))
- {
- // If we were waiting for a trailing surrogate, its an error
- if (gotLeadingSurrogate)
- emitError(XMLErrs::Expected2ndSurrogateChar);
+ // Try to get another char from the source
+ // The code from here on down covers all contengencies,
+ if (!fReaderMgr.getNextCharIfNot(chOpenAngle, nextCh))
+ {
+ // If we were waiting for a trailing surrogate, its an error
+ if (gotLeadingSurrogate)
+ emitError(XMLErrs::Expected2ndSurrogateChar);
- notDone = false;
- break;
- }
+ notDone = false;
+ break;
}
// Watch for a reference. Note that the escapement mechanism
// is ignored in this content.
+ escaped = false;
if (nextCh == chAmpersand)
{
sendCharData(toUse);
@@ -2143,42 +2106,10 @@ void IGXMLScanner::scanCharData(XMLBuffer& toUse)
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
}
- else
- {
- escaped = false;
- }
-
- // Keep the state machine up to date
- if (!escaped)
- {
- if (nextCh == chCloseSquare)
- {
- if (curState == State_Waiting)
- curState = State_GotOne;
- else if (curState == State_GotOne)
- curState = State_GotTwo;
- }
- else if (nextCh == chCloseAngle)
- {
- if (curState == State_GotTwo)
- emitError(XMLErrs::BadSequenceInCharData);
- curState = State_Waiting;
- }
- else
- {
- curState = State_Waiting;
- }
- }
- else
- {
- curState = State_Waiting;
- }
-
- // Deal with surrogate pairs
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
+ // Deal with surrogate pairs
// Its a leading surrogate. If we already got one, then
// issue an error, else set leading flag to make sure that
// we look for a trailing next time.
@@ -2208,27 +2139,51 @@ void IGXMLScanner::scanCharData(XMLBuffer& toUse)
// Make sure the returned char is a valid XML char
if (!fReaderMgr.getCurrentReader()->isXMLChar(nextCh))
{
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr.getCurrentReader()->isControlChar(nextCh)))
- {
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- emitError(XMLErrs::InvalidCharacter, tmpBuf);
- }
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ emitError(XMLErrs::InvalidCharacter, tmpBuf);
}
}
- charref_expanded = false;
gotLeadingSurrogate = false;
}
+ // Keep the state machine up to date
+ if (!escaped)
+ {
+ if (nextCh == chCloseSquare)
+ {
+ if (curState == State_Waiting)
+ curState = State_GotOne;
+ else if (curState == State_GotOne)
+ curState = State_GotTwo;
+ }
+ else if (nextCh == chCloseAngle)
+ {
+ if (curState == State_GotTwo)
+ emitError(XMLErrs::BadSequenceInCharData);
+ curState = State_Waiting;
+ }
+ else
+ {
+ curState = State_Waiting;
+ }
+ }
+ else
+ {
+ curState = State_Waiting;
+ }
+
// Add this char to the buffer
toUse.append(nextCh);
+
+ if (secondCh)
+ toUse.append(secondCh);
}
}
catch(const EndOfEntityException& toCatch)
diff --git a/src/xercesc/internal/SGXMLScanner.cpp b/src/xercesc/internal/SGXMLScanner.cpp
index 552082abf948204bb36fa95f29d649219a7e41f3..f8707041bc14ded3f68bd17c8e1f39d6ce4dde4d 100644
--- a/src/xercesc/internal/SGXMLScanner.cpp
+++ b/src/xercesc/internal/SGXMLScanner.cpp
@@ -3224,23 +3224,13 @@ bool SGXMLScanner::basicAttrValueScan(const XMLCh* const attrName, XMLBuffer& to
XMLCh secondCh = 0;
bool gotLeadingSurrogate = false;
bool escaped;
- bool charref_expanded = false;
while (true)
{
try
{
while(true)
{
- // Get another char. Use second char if one is waiting
- if (secondCh)
- {
- nextCh = secondCh;
- secondCh = 0;
- }
- else
- {
- nextCh = fReaderMgr.getNextChar();
- }
+ nextCh = fReaderMgr.getNextChar();
if (!nextCh)
ThrowXML(UnexpectedEOFException, XMLExcepts::Gen_UnexpectedEOF);
@@ -3271,12 +3261,10 @@ bool SGXMLScanner::basicAttrValueScan(const XMLCh* const attrName, XMLBuffer& to
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
}
-
- // Deal with surrogate pairs
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
+ // Deal with surrogate pairs
// Its a leading surrogate. If we already got one, then
// issue an error, else set leading flag to make sure that
// we look for a trailing next time.
@@ -3308,22 +3296,17 @@ bool SGXMLScanner::basicAttrValueScan(const XMLCh* const attrName, XMLBuffer& to
// Its got to at least be a valid XML character
else if (!fReaderMgr.getCurrentReader()->isXMLChar(nextCh))
{
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr.getCurrentReader()->isControlChar(nextCh)))
- {
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- emitError(XMLErrs::InvalidCharacterInAttrValue, attrName, tmpBuf);
- }
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ emitError(XMLErrs::InvalidCharacterInAttrValue, attrName, tmpBuf);
}
}
- charref_expanded = false;
gotLeadingSurrogate = false;
}
@@ -3336,6 +3319,9 @@ bool SGXMLScanner::basicAttrValueScan(const XMLCh* const attrName, XMLBuffer& to
// Else add it to the buffer
toFill.append(nextCh);
+
+ if (secondCh)
+ toFill.append(secondCh);
}
}
catch(const EndOfEntityException&)
@@ -3553,44 +3539,36 @@ void SGXMLScanner::scanCharData(XMLBuffer& toUse)
bool escaped = false;
bool gotLeadingSurrogate = false;
bool notDone = true;
- bool charref_expanded = false;
while (notDone)
{
try
{
while (true)
{
- if (secondCh)
+ // Eat through as many plain content characters as possible without
+ // needing special handling. Moving most content characters here,
+ // in this one call, rather than running the overall loop once
+ // per content character, is a speed optimization.
+ if (curState == State_Waiting && !gotLeadingSurrogate)
{
- nextCh = secondCh;
- secondCh = 0;
+ fReaderMgr.movePlainContentChars(toUse);
}
- else
- {
- // Eat through as many plain content characters as possible without
- // needing special handling. Moving most content characters here,
- // in this one call, rather than running the overall loop once
- // per content character, is a speed optimization.
- if (curState == State_Waiting && !gotLeadingSurrogate)
- {
- fReaderMgr.movePlainContentChars(toUse);
- }
- // Try to get another char from the source
- // The code from here on down covers all contengencies,
- if (!fReaderMgr.getNextCharIfNot(chOpenAngle, nextCh))
- {
- // If we were waiting for a trailing surrogate, its an error
- if (gotLeadingSurrogate)
- emitError(XMLErrs::Expected2ndSurrogateChar);
+ // Try to get another char from the source
+ // The code from here on down covers all contengencies,
+ if (!fReaderMgr.getNextCharIfNot(chOpenAngle, nextCh))
+ {
+ // If we were waiting for a trailing surrogate, its an error
+ if (gotLeadingSurrogate)
+ emitError(XMLErrs::Expected2ndSurrogateChar);
- notDone = false;
- break;
- }
+ notDone = false;
+ break;
}
// Watch for a reference. Note that the escapement mechanism
// is ignored in this content.
+ escaped = false;
if (nextCh == chAmpersand)
{
sendCharData(toUse);
@@ -3603,42 +3581,10 @@ void SGXMLScanner::scanCharData(XMLBuffer& toUse)
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
- }
- else
- {
- escaped = false;
- }
-
- // Keep the state machine up to date
- if (!escaped)
- {
- if (nextCh == chCloseSquare)
- {
- if (curState == State_Waiting)
- curState = State_GotOne;
- else if (curState == State_GotOne)
- curState = State_GotTwo;
- }
- else if (nextCh == chCloseAngle)
- {
- if (curState == State_GotTwo)
- emitError(XMLErrs::BadSequenceInCharData);
- curState = State_Waiting;
- }
- else
- {
- curState = State_Waiting;
- }
- }
- else
- {
- curState = State_Waiting;
}
-
- // Deal with surrogate pairs
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
+ // Deal with surrogate pairs
// Its a leading surrogate. If we already got one, then
// issue an error, else set leading flag to make sure that
// we look for a trailing next time.
@@ -3668,27 +3614,51 @@ void SGXMLScanner::scanCharData(XMLBuffer& toUse)
// Make sure the returned char is a valid XML char
if (!fReaderMgr.getCurrentReader()->isXMLChar(nextCh))
{
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr.getCurrentReader()->isControlChar(nextCh)))
- {
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- emitError(XMLErrs::InvalidCharacter, tmpBuf);
- }
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ emitError(XMLErrs::InvalidCharacter, tmpBuf);
}
}
- charref_expanded = false;
gotLeadingSurrogate = false;
}
+ // Keep the state machine up to date
+ if (!escaped)
+ {
+ if (nextCh == chCloseSquare)
+ {
+ if (curState == State_Waiting)
+ curState = State_GotOne;
+ else if (curState == State_GotOne)
+ curState = State_GotTwo;
+ }
+ else if (nextCh == chCloseAngle)
+ {
+ if (curState == State_GotTwo)
+ emitError(XMLErrs::BadSequenceInCharData);
+ curState = State_Waiting;
+ }
+ else
+ {
+ curState = State_Waiting;
+ }
+ }
+ else
+ {
+ curState = State_Waiting;
+ }
+
// Add this char to the buffer
toUse.append(nextCh);
+
+ if (secondCh)
+ toUse.append(secondCh);
}
}
catch(const EndOfEntityException& toCatch)
diff --git a/src/xercesc/internal/WFXMLScanner.cpp b/src/xercesc/internal/WFXMLScanner.cpp
index 12099dc8fb111c0ef6ae84381945a6b1969eb44a..d52e0b65efe3fd7fb99dd43440f47cf2092f80d8 100644
--- a/src/xercesc/internal/WFXMLScanner.cpp
+++ b/src/xercesc/internal/WFXMLScanner.cpp
@@ -1522,23 +1522,13 @@ bool WFXMLScanner::scanAttValue(const XMLCh* const attrName
bool firstNonWS = false;
bool gotLeadingSurrogate = false;
bool escaped;
- bool charref_expanded = false;
while (true)
{
try
{
while(true)
{
- // Get another char. Use second char if one is waiting
- if (secondCh)
- {
- nextCh = secondCh;
- secondCh = 0;
- }
- else
- {
- nextCh = fReaderMgr.getNextChar();
- }
+ nextCh = fReaderMgr.getNextChar();
if (!nextCh)
ThrowXML(UnexpectedEOFException, XMLExcepts::Gen_UnexpectedEOF);
@@ -1568,12 +1558,10 @@ bool WFXMLScanner::scanAttValue(const XMLCh* const attrName
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
}
-
- // Deal with surrogate pairs
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
+ // Deal with surrogate pairs
// Its a leading surrogate. If we already got one, then
// issue an error, else set leading flag to make sure that
// we look for a trailing next time.
@@ -1605,30 +1593,24 @@ bool WFXMLScanner::scanAttValue(const XMLCh* const attrName
// Its got to at least be a valid XML character
else if (!fReaderMgr.getCurrentReader()->isXMLChar(nextCh))
{
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr.getCurrentReader()->isControlChar(nextCh)))
- {
-
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- emitError(XMLErrs::InvalidCharacterInAttrValue, attrName, tmpBuf);
- }
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ emitError(XMLErrs::InvalidCharacterInAttrValue, attrName, tmpBuf);
}
}
- charref_expanded = false;
gotLeadingSurrogate = false;
}
// If its not escaped, then make sure its not a < character, which
// is not allowed in attribute values.
if (!escaped) {
- if (nextCh == chOpenAngle)
+ if (nextCh == chOpenAngle)
emitError(XMLErrs::BracketInAttrValue, attrName);
else if (fReaderMgr.getCurrentReader()->isWhitespace(nextCh))
nextCh = chSpace;
@@ -1636,6 +1618,9 @@ bool WFXMLScanner::scanAttValue(const XMLCh* const attrName
// Else add it to the buffer
toFill.append(nextCh);
+
+ if (secondCh)
+ toFill.append(secondCh);
}
}
catch(const EndOfEntityException&)
@@ -1807,44 +1792,36 @@ void WFXMLScanner::scanCharData(XMLBuffer& toUse)
bool escaped = false;
bool gotLeadingSurrogate = false;
bool notDone = true;
- bool charref_expanded = false;
while (notDone)
{
try
{
while (true)
{
- if (secondCh)
+ // Eat through as many plain content characters as possible without
+ // needing special handling. Moving most content characters here,
+ // in this one call, rather than running the overall loop once
+ // per content character, is a speed optimization.
+ if (curState == State_Waiting && !gotLeadingSurrogate)
{
- nextCh = secondCh;
- secondCh = 0;
+ fReaderMgr.movePlainContentChars(toUse);
}
- else
- {
- // Eat through as many plain content characters as possible without
- // needing special handling. Moving most content characters here,
- // in this one call, rather than running the overall loop once
- // per content character, is a speed optimization.
- if (curState == State_Waiting && !gotLeadingSurrogate)
- {
- fReaderMgr.movePlainContentChars(toUse);
- }
- // Try to get another char from the source
- // The code from here on down covers all contengencies,
- if (!fReaderMgr.getNextCharIfNot(chOpenAngle, nextCh))
- {
- // If we were waiting for a trailing surrogate, its an error
- if (gotLeadingSurrogate)
- emitError(XMLErrs::Expected2ndSurrogateChar);
+ // Try to get another char from the source
+ // The code from here on down covers all contengencies,
+ if (!fReaderMgr.getNextCharIfNot(chOpenAngle, nextCh))
+ {
+ // If we were waiting for a trailing surrogate, its an error
+ if (gotLeadingSurrogate)
+ emitError(XMLErrs::Expected2ndSurrogateChar);
- notDone = false;
- break;
- }
+ notDone = false;
+ break;
}
// Watch for a reference. Note that the escapement mechanism
// is ignored in this content.
+ escaped = false;
if (nextCh == chAmpersand)
{
sendCharData(toUse);
@@ -1857,42 +1834,10 @@ void WFXMLScanner::scanCharData(XMLBuffer& toUse)
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
- }
- else
- {
- escaped = false;
- }
-
- // Keep the state machine up to date
- if (!escaped)
- {
- if (nextCh == chCloseSquare)
- {
- if (curState == State_Waiting)
- curState = State_GotOne;
- else if (curState == State_GotOne)
- curState = State_GotTwo;
- }
- else if (nextCh == chCloseAngle)
- {
- if (curState == State_GotTwo)
- emitError(XMLErrs::BadSequenceInCharData);
- curState = State_Waiting;
- }
- else
- {
- curState = State_Waiting;
- }
}
- else
- {
- curState = State_Waiting;
- }
-
- // Deal with surrogate pairs
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
+ // Deal with surrogate pairs
// Its a leading surrogate. If we already got one, then
// issue an error, else set leading flag to make sure that
// we look for a trailing next time.
@@ -1924,29 +1869,51 @@ void WFXMLScanner::scanCharData(XMLBuffer& toUse)
// Its got to at least be a valid XML character
else if (!fReaderMgr.getCurrentReader()->isXMLChar(nextCh))
{
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr.getCurrentReader()->isControlChar(nextCh)))
- {
-
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- emitError(XMLErrs::InvalidCharacter, tmpBuf);
- }
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ emitError(XMLErrs::InvalidCharacter, tmpBuf);
}
}
- charref_expanded = false;
gotLeadingSurrogate = false;
}
+ // Keep the state machine up to date
+ if (!escaped)
+ {
+ if (nextCh == chCloseSquare)
+ {
+ if (curState == State_Waiting)
+ curState = State_GotOne;
+ else if (curState == State_GotOne)
+ curState = State_GotTwo;
+ }
+ else if (nextCh == chCloseAngle)
+ {
+ if (curState == State_GotTwo)
+ emitError(XMLErrs::BadSequenceInCharData);
+ curState = State_Waiting;
+ }
+ else
+ {
+ curState = State_Waiting;
+ }
+ }
+ else
+ {
+ curState = State_Waiting;
+ }
// Add this char to the buffer
toUse.append(nextCh);
+
+ if (secondCh)
+ toUse.append(secondCh);
}
}
catch(const EndOfEntityException& toCatch)
diff --git a/src/xercesc/internal/XMLScanner.cpp b/src/xercesc/internal/XMLScanner.cpp
index 6be0688cef0009cb59512529e158eff105106eec..54296478b4a7d9da79b8db7e779c876993e54221 100644
--- a/src/xercesc/internal/XMLScanner.cpp
+++ b/src/xercesc/internal/XMLScanner.cpp
@@ -1725,16 +1725,27 @@ bool XMLScanner::scanCharRef(XMLCh& toFill, XMLCh& second)
}
// Return the char (or chars)
- if (value >= 0x10000)
+ // And check if the character expanded is valid or not
+ if (value >= 0x10000 && value <= 0x10FFFF)
{
value -= 0x10000;
toFill = XMLCh((value >> 10) + 0xD800);
second = XMLCh((value & 0x3FF) + 0xDC00);
}
- else
+ else if (value <= 0xFFFD)
{
toFill = XMLCh(value);
second = 0;
+ if (!fReaderMgr.getCurrentReader()->isXMLChar(toFill) && !fReaderMgr.getCurrentReader()->isControlChar(toFill)) {
+ // Character reference was not in the valid range
+ emitError(XMLErrs::InvalidCharacterRef);
+ return false;
+ }
+ }
+ else {
+ // Character reference was not in the valid range
+ emitError(XMLErrs::InvalidCharacterRef);
+ return false;
}
return true;
diff --git a/src/xercesc/validators/DTD/DTDScanner.cpp b/src/xercesc/validators/DTD/DTDScanner.cpp
index 4f0e8eb16fe03237c97286c54d585bcfb4da4cff..0631696820f3c4c9220faebcde5ac93dab2a8ebc 100644
--- a/src/xercesc/validators/DTD/DTDScanner.cpp
+++ b/src/xercesc/validators/DTD/DTDScanner.cpp
@@ -56,6 +56,9 @@
/*
* $Log$
+ * Revision 1.20 2002/12/24 16:12:19 tng
+ * For performance reason, move the character check to scancharref.
+ *
* Revision 1.19 2002/12/20 22:10:47 tng
* XML 1.1
*
@@ -918,23 +921,13 @@ bool DTDScanner::scanAttValue(const XMLCh* const attrName
bool firstNonWS = false;
bool gotLeadingSurrogate = false;
bool escaped;
- bool charref_expanded = false;
while (true)
{
try
{
while(true)
{
- // Get another char. Use second char from prevous is its there
- if (secondCh)
- {
- nextCh = secondCh;
- secondCh = 0;
- }
- else
- {
- nextCh = fReaderMgr->getNextChar();
- }
+ nextCh = fReaderMgr->getNextChar();
if (!nextCh)
ThrowXML(UnexpectedEOFException, XMLExcepts::Gen_UnexpectedEOF);
@@ -966,12 +959,10 @@ bool DTDScanner::scanAttValue(const XMLCh* const attrName
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
}
-
- // Check for correct surrogate pairs
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
+ // Check for correct surrogate pairs
if (gotLeadingSurrogate)
fScanner->emitError(XMLErrs::Expected2ndSurrogateChar);
else
@@ -985,29 +976,25 @@ bool DTDScanner::scanAttValue(const XMLCh* const attrName
fScanner->emitError(XMLErrs::Expected2ndSurrogateChar);
}
// Its got to at least be a valid XML character
- else if (!fReaderMgr->getCurrentReader()->isXMLChar(nextCh)) {
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr->getCurrentReader()->isControlChar(nextCh)))
- {
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- fScanner->emitError
- (
- XMLErrs::InvalidCharacterInAttrValue
- , attrName
- , tmpBuf
- );
- }
+ else if (!fReaderMgr->getCurrentReader()->isXMLChar(nextCh))
+ {
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ fScanner->emitError
+ (
+ XMLErrs::InvalidCharacterInAttrValue
+ , attrName
+ , tmpBuf
+ );
}
gotLeadingSurrogate = false;
- charref_expanded = false;
}
//
@@ -1063,6 +1050,9 @@ bool DTDScanner::scanAttValue(const XMLCh* const attrName
// Else add it to the buffer
toFill.append(nextCh);
+
+ if (secondCh)
+ toFill.append(secondCh);
}
}
@@ -1164,17 +1154,29 @@ bool DTDScanner::scanCharRef(XMLCh& first, XMLCh& second)
}
// Return the char (or chars)
- if (value >= 0x10000)
+ // And check if the character expanded is valid or not
+ if (value >= 0x10000 && value <= 0x10FFFF)
{
value -= 0x10000;
- first = XMLCh((value >> 10) + 0xD800);
+ first = XMLCh((value >> 10) + 0xD800);
second = XMLCh((value & 0x3FF) + 0xDC00);
}
- else
+ else if (value <= 0xFFFD)
{
- first = XMLCh(value);
+ first = XMLCh(value);
second = 0;
+ if (!fReaderMgr->getCurrentReader()->isXMLChar(first) && !fReaderMgr->getCurrentReader()->isControlChar(first)) {
+ // Character reference was not in the valid range
+ fScanner->emitError(XMLErrs::InvalidCharacterRef);
+ return false;
+ }
}
+ else {
+ // Character reference was not in the valid range
+ fScanner->emitError(XMLErrs::InvalidCharacterRef);
+ return false;
+ }
+
return true;
}
@@ -2156,19 +2158,9 @@ bool DTDScanner::scanEntityLiteral(XMLBuffer& toFill, const bool isPE)
XMLCh nextCh;
XMLCh secondCh = 0;
bool gotLeadingSurrogate = false;
- bool charref_expanded = false;
while (true)
{
- // Get the second char if we have one, else get another
- if (secondCh)
- {
- nextCh = secondCh;
- secondCh = 0;
- }
- else
- {
- nextCh = fReaderMgr->getNextChar();
- }
+ nextCh = fReaderMgr->getNextChar();
//
// Watch specifically for EOF and issue a more meaningful error
@@ -2226,8 +2218,6 @@ bool DTDScanner::scanEntityLiteral(XMLBuffer& toFill, const bool isPE)
gotLeadingSurrogate = false;
continue;
}
- charref_expanded = true;
-
}
else
{
@@ -2262,8 +2252,7 @@ bool DTDScanner::scanEntityLiteral(XMLBuffer& toFill, const bool isPE)
gotLeadingSurrogate = false;
}
}
-
- if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
+ else if ((nextCh >= 0xD800) && (nextCh <= 0xDBFF))
{
if (gotLeadingSurrogate)
fScanner->emitError(XMLErrs::Expected2ndSurrogateChar);
@@ -2279,28 +2268,26 @@ bool DTDScanner::scanEntityLiteral(XMLBuffer& toFill, const bool isPE)
}
else if (!fReaderMgr->getCurrentReader()->isXMLChar(nextCh))
{
- // if it was a character reference and is control char, then it's ok
- if (!(charref_expanded && fReaderMgr->getCurrentReader()->isControlChar(nextCh)))
- {
- XMLCh tmpBuf[9];
- XMLString::binToText
- (
- nextCh
- , tmpBuf
- , 8
- , 16
- );
- fScanner->emitError(XMLErrs::InvalidCharacter, tmpBuf);
- fReaderMgr->skipPastChar(quoteCh);
- return false;
- }
+ XMLCh tmpBuf[9];
+ XMLString::binToText
+ (
+ nextCh
+ , tmpBuf
+ , 8
+ , 16
+ );
+ fScanner->emitError(XMLErrs::InvalidCharacter, tmpBuf);
+ fReaderMgr->skipPastChar(quoteCh);
+ return false;
}
- charref_expanded = false;
gotLeadingSurrogate = false;
}
// Looks ok, so add it to the literal
toFill.append(nextCh);
+
+ if (secondCh)
+ toFill.append(secondCh);
}
//