version 1.46, 2008/06/18 18:12:00
|
version 1.46.4.3, 2009/02/09 11:36:51
|
|
|
#pragma optimize( "", on ) | #pragma optimize( "", on ) |
#endif | #endif |
| |
static inline int _getCharRef(char*& p, bool hex) |
static inline int _getCharRef(char*& p) |
{ | { |
char* end; | char* end; |
unsigned long ch; | unsigned long ch; |
|
Boolean hex = false; |
| |
if (hex) |
if (*p == 'x') |
{ | { |
ch = strtoul(p, &end, 16); |
hex = true; |
|
ch = strtoul(++p, &end, 16); |
} | } |
else | else |
{ | { |
|
|
return ch; | return ch; |
} | } |
| |
static void _normalize(Uint32& line, char*& p, char end_char, char*& start) |
// Parse an entity reference or a character reference |
|
static inline int _getRef(Uint32 line, char*& p) |
{ | { |
// Skip over leading whitespace: |
int ch; |
| |
_skipWhitespace(line, p); |
if (*p == '#') |
start = p; |
{ |
|
ch = _getCharRef(++p); |
|
} |
|
else |
|
{ |
|
ch = _getEntityRef(p); |
|
} |
|
|
|
if (ch == -1) |
|
{ |
|
throw XmlException(XmlException::MALFORMED_REFERENCE, line); |
|
} |
| |
|
return ch; |
|
} |
|
|
|
static inline void _normalizeElementValue( |
|
Uint32& line, |
|
char*& p, |
|
Uint32 &textLen) |
|
{ |
// Process one character at a time: | // Process one character at a time: |
| |
char* q = p; | char* q = p; |
|
char *start = p; |
| |
while (*p && (*p != end_char)) |
while (*p && (*p != '<')) |
{ | { |
if (_isspace(*p)) | if (_isspace(*p)) |
{ | { |
// Compress sequences of whitespace characters to a single space |
// Trim whitespace from the end of the value, but do not compress |
// character. Update line number when newlines encountered. |
// whitespace within the value. |
|
|
|
const char* start = p; |
| |
if (*p++ == '\n') | if (*p++ == '\n') |
{ | { |
line++; | line++; |
} | } |
| |
*q++ = ' '; |
|
|
|
_skipWhitespace(line, p); | _skipWhitespace(line, p); |
} |
|
else if (*p == '&') |
|
{ |
|
// Process entity characters and entity references: |
|
| |
p++; |
if (*p && (*p != '<')) |
int ch; |
|
|
|
if (*p == '#') |
|
{ | { |
*p++; |
// Transfer internal whitespace to q without compressing it. |
|
const char* i = start; |
if (*p == 'x') |
while (i < p) |
{ | { |
p++; |
*q++ = *i++; |
ch = _getCharRef(p, true); |
} |
} | } |
else | else |
{ | { |
ch = _getCharRef(p, false); |
// Do not transfer trailing whitespace to q. |
|
break; |
|
} |
} | } |
|
else if (*p == '&') |
|
{ |
|
// Process an entity reference or a character reference. |
|
|
|
*q++ = _getRef(line, ++p); |
} | } |
else | else |
{ | { |
ch = _getEntityRef(p); |
*q++ = *p++; |
|
} |
} | } |
| |
if (ch == -1) |
// If q got behind p, it is safe and necessary to null-terminate q |
|
|
|
if (q != p) |
{ | { |
throw XmlException(XmlException::MALFORMED_REFERENCE, line); |
*q = '\0'; |
|
} |
|
textLen = (Uint32)(q - start); |
|
} |
|
|
|
static inline void _normalizeAttributeValue( |
|
Uint32& line, |
|
char*& p, |
|
char end_char, |
|
char*& start) |
|
{ |
|
// Skip over leading whitespace: |
|
|
|
_skipWhitespace(line, p); |
|
start = p; |
|
|
|
// Process one character at a time: |
|
|
|
char* q = p; |
|
|
|
while (*p && (*p != end_char)) |
|
{ |
|
if (_isspace(*p)) |
|
{ |
|
// Compress sequences of whitespace characters to a single space |
|
// character. Update line number when newlines encountered. |
|
|
|
if (*p++ == '\n') |
|
{ |
|
line++; |
|
} |
|
|
|
*q++ = ' '; |
|
|
|
_skipWhitespace(line, p); |
} | } |
|
else if (*p == '&') |
|
{ |
|
// Process an entity reference or a character reference. |
| |
*q++ = ch; |
*q++ = _getRef(line, ++p); |
} | } |
else | else |
{ | { |
|
|
{ | { |
// Normalize the content: | // Normalize the content: |
| |
char* start; |
char* start = _current; |
_normalize(_line, _current, '<', start); |
Uint32 textLen; |
|
_normalizeElementValue(_line, _current, textLen); |
| |
// Get the content: | // Get the content: |
| |
entry.type = XmlEntry::CONTENT; | entry.type = XmlEntry::CONTENT; |
entry.text = start; | entry.text = start; |
|
entry.textLen = textLen; |
| |
// Overwrite '<' with a null character (temporarily). | // Overwrite '<' with a null character (temporarily). |
| |
|
|
1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, | 1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, |
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, | 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, |
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, | 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, |
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, |
0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, |
}; | }; |
| |
inline Boolean _getQName(char*& p, const char*& localName) | inline Boolean _getQName(char*& p, const char*& localName) |
|
|
| |
p++; | p++; |
| |
while (*p && _isInnerElementChar[Uint8(*p)]) |
// No explicit test for NULL termination is needed. |
|
// On position 0 of the array false is returned. |
|
while (_isInnerElementChar[Uint8(*p)]) |
p++; | p++; |
| |
// We've validated the prefix, now validate the local name | // We've validated the prefix, now validate the local name |
|
|
return false; | return false; |
| |
p++; | p++; |
|
// No explicit test for NULL termination is needed. |
while (*p && _isInnerElementChar[Uint8(*p)]) |
// On position 0 of the array false is returned. |
|
while (_isInnerElementChar[Uint8(*p)]) |
p++; | p++; |
} | } |
| |
|
|
entry.type = XmlEntry::CDATA; | entry.type = XmlEntry::CDATA; |
entry.text = p; | entry.text = p; |
_getCData(p); | _getCData(p); |
|
entry.textLen = strlen(entry.text); |
return; | return; |
} | } |
else if (memcmp(p, "DOCTYPE", 7) == 0) | else if (memcmp(p, "DOCTYPE", 7) == 0) |
|
|
| |
return; | return; |
} | } |
else if ((((*p >= 'A') && (*p <= 'Z')) || |
else if (CharSet::isAlphaUnder(Uint8(*p))) |
((*p >= 'a') && (*p <= 'z')) || |
|
(*p == '_'))) |
|
{ | { |
entry.type = XmlEntry::START_TAG; | entry.type = XmlEntry::START_TAG; |
entry.text = p; | entry.text = p; |
|
|
char quote = *p++; | char quote = *p++; |
| |
char* start; | char* start; |
_normalize(_line, p, quote, start); |
_normalizeAttributeValue(_line, p, quote, start); |
attr.value = start; | attr.value = start; |
| |
if (*p != quote) | if (*p != quote) |