XS-libgeos
view release on metacpan or search on metacpan
geos-3.7.3/tests/xmltester/tinyxml/tinyxmlparser.cpp view on Meta::CPAN
}
// Was this empty?
if ( !firstChild ) {
SetError( TIXML_ERROR_DOCUMENT_EMPTY, nullptr, nullptr, encoding );
return nullptr;
}
// All is well.
return p;
}
void TiXmlDocument::SetError( int err, const char* pError, TiXmlParsingData* data, TiXmlEncoding encoding )
{
// The first error in a chain is more accurate - don't set again!
if ( error )
return;
assert( err > 0 && err < TIXML_ERROR_STRING_COUNT );
error = true;
errorId = err;
errorDesc = errorString[ errorId ];
errorLocation.Clear();
if ( pError && data )
{
data->Stamp( pError, encoding );
errorLocation = data->Cursor();
}
}
TiXmlNode* TiXmlNode::Identify( const char* p, TiXmlEncoding encoding )
{
TiXmlNode* returnNode = nullptr;
p = SkipWhiteSpace( p, encoding );
if( !p || !*p || *p != '<' )
{
return nullptr;
}
TiXmlDocument* doc = GetDocument();
p = SkipWhiteSpace( p, encoding );
if ( !p || !*p )
{
return nullptr;
}
// What is this thing?
// - Elements start with a letter or underscore, but xml is reserved.
// - Comments: <!--
// - Decleration: <?xml
// - Everthing else is unknown to tinyxml.
//
const char* xmlHeader = { "<?xml" };
const char* commentHeader = { "<!--" };
const char* dtdHeader = { "<!" };
const char* cdataHeader = { "<![CDATA[" };
if ( StringEqual( p, xmlHeader, true, encoding ) )
{
#ifdef DEBUG_PARSER
TIXML_LOG( "XML parsing Declaration\n" );
#endif
returnNode = new TiXmlDeclaration();
}
else if ( StringEqual( p, commentHeader, false, encoding ) )
{
#ifdef DEBUG_PARSER
TIXML_LOG( "XML parsing Comment\n" );
#endif
returnNode = new TiXmlComment();
}
else if ( StringEqual( p, cdataHeader, false, encoding ) )
{
#ifdef DEBUG_PARSER
TIXML_LOG( "XML parsing CDATA\n" );
#endif
TiXmlText* text = new TiXmlText( "" );
text->SetCDATA( true );
returnNode = text;
}
else if ( StringEqual( p, dtdHeader, false, encoding ) )
{
#ifdef DEBUG_PARSER
TIXML_LOG( "XML parsing Unknown(1)\n" );
#endif
returnNode = new TiXmlUnknown();
}
else if ( IsAlpha( *(p+1), encoding )
|| *(p+1) == '_' )
{
#ifdef DEBUG_PARSER
TIXML_LOG( "XML parsing Element\n" );
#endif
returnNode = new TiXmlElement( "" );
}
else
{
#ifdef DEBUG_PARSER
TIXML_LOG( "XML parsing Unknown(2)\n" );
#endif
returnNode = new TiXmlUnknown();
}
if ( returnNode )
{
// Set the parent, so it can report errors
returnNode->parent = this;
}
else
{
if ( doc )
doc->SetError( TIXML_ERROR_OUT_OF_MEMORY, nullptr, nullptr, TIXML_ENCODING_UNKNOWN );
}
return returnNode;
}
#ifdef TIXML_USE_STL
void TiXmlElement::StreamIn (std::istream * in, TIXML_STRING * tag)
{
// We're called with some amount of pre-parsing. That is, some of "this"
// element is in "tag". Go ahead and stream to the closing ">"
while( in->good() )
{
int c = in->get();
if ( c <= 0 )
{
TiXmlDocument* document = GetDocument();
if ( document )
document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN );
return;
}
(*tag) += (char) c ;
if ( c == '>' )
break;
}
if ( tag->length() < 3 ) return;
// Okay...if we are a "/>" tag, then we're done. We've read a complete tag.
// If not, identify and stream.
if ( tag->at( tag->length() - 1 ) == '>'
&& tag->at( tag->length() - 2 ) == '/' )
{
// All good!
return;
}
else if ( tag->at( tag->length() - 1 ) == '>' )
{
// There is more. Could be:
// text
// cdata text (which looks like another node)
// closing tag
// another node.
for ( ;; )
{
StreamWhiteSpace( in, tag );
// Do we have text?
if ( in->good() && in->peek() != '<' )
{
// Yep, text.
TiXmlText text( "" );
text.StreamIn( in, tag );
// What follows text is a closing tag or another node.
// Go around again and figure it out.
continue;
}
// We now have either a closing tag...or another node.
// We should be at a "<", regardless.
if ( !in->good() ) return;
assert( in->peek() == '<' );
int tagIndex = (int) tag->length();
bool closingTag = false;
bool firstCharFound = false;
for( ;; )
{
if ( !in->good() )
return;
int c = in->peek();
if ( c <= 0 )
{
TiXmlDocument* document = GetDocument();
if ( document )
document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN );
return;
}
if ( c == '>' )
break;
*tag += (char) c;
in->get();
// Early out if we find the CDATA id.
if ( c == '[' && tag->size() >= 9 )
{
size_t len = tag->size();
const char* start = tag->c_str() + len - 9;
if ( strcmp( start, "<![CDATA[" ) == 0 ) {
assert( !closingTag );
break;
}
}
if ( !firstCharFound && c != '<' && !IsWhiteSpace( c ) )
{
geos-3.7.3/tests/xmltester/tinyxml/tinyxmlparser.cpp view on Meta::CPAN
if ( !p || !*p || *p != '=' )
{
if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, p, data, encoding );
return nullptr;
}
++p; // skip '='
p = SkipWhiteSpace( p, encoding );
if ( !p || !*p )
{
if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, p, data, encoding );
return nullptr;
}
const char* end;
const char SINGLE_QUOTE = '\'';
const char DOUBLE_QUOTE = '\"';
if ( *p == SINGLE_QUOTE )
{
++p;
end = "\'"; // single quote in string
p = ReadText( p, &value, false, end, false, encoding );
}
else if ( *p == DOUBLE_QUOTE )
{
++p;
end = "\""; // double quote in string
p = ReadText( p, &value, false, end, false, encoding );
}
else
{
// All attribute values should be in single or double quotes.
// But this is such a common error that the parser will try
// its best, even without them.
value = "";
while ( p && *p // existence
&& !IsWhiteSpace( *p ) && *p != '\n' && *p != '\r' // whitespace
&& *p != '/' && *p != '>' ) // tag end
{
if ( *p == SINGLE_QUOTE || *p == DOUBLE_QUOTE ) {
// [ 1451649 ] Attribute values with trailing quotes not handled correctly
// We did not have an opening quote but seem to have a
// closing one. Give up and throw an error.
if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, p, data, encoding );
return nullptr;
}
value += *p;
++p;
}
}
return p;
}
#ifdef TIXML_USE_STL
void TiXmlText::StreamIn( std::istream * in, TIXML_STRING * tag )
{
while ( in->good() )
{
int c = in->peek();
if ( !cdata && (c == '<' ) )
{
return;
}
if ( c <= 0 )
{
TiXmlDocument* document = GetDocument();
if ( document )
document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN );
return;
}
(*tag) += (char) c;
in->get(); // "commits" the peek made above
if ( cdata && c == '>' && tag->size() >= 3 ) {
size_t len = tag->size();
if ( (*tag)[len-2] == ']' && (*tag)[len-3] == ']' ) {
// terminator of cdata.
return;
}
}
}
}
#endif
const char* TiXmlText::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding encoding )
{
value = "";
TiXmlDocument* document = GetDocument();
if ( data )
{
data->Stamp( p, encoding );
location = data->Cursor();
}
const char* const startTag = "<![CDATA[";
const char* const endTag = "]]>";
if ( cdata || StringEqual( p, startTag, false, encoding ) )
{
cdata = true;
if ( !StringEqual( p, startTag, false, encoding ) )
{
document->SetError( TIXML_ERROR_PARSING_CDATA, p, data, encoding );
return nullptr;
}
p += strlen( startTag );
// Keep all the white space, ignore the encoding, etc.
while ( p && *p
&& !StringEqual( p, endTag, false, encoding )
)
{
value += *p;
++p;
}
TIXML_STRING dummy;
p = ReadText( p, &dummy, false, endTag, false, encoding );
return p;
}
else
{
bool ignoreWhite = true;
const char* end = "<";
p = ReadText( p, &value, ignoreWhite, end, false, encoding );
if ( p )
return p-1; // don't truncate the '<'
return nullptr;
}
}
#ifdef TIXML_USE_STL
void TiXmlDeclaration::StreamIn( std::istream * in, TIXML_STRING * tag )
{
while ( in->good() )
{
int c = in->get();
if ( c <= 0 )
{
TiXmlDocument* document = GetDocument();
if ( document )
document->SetError( TIXML_ERROR_EMBEDDED_NULL, 0, 0, TIXML_ENCODING_UNKNOWN );
return;
}
(*tag) += (char) c;
if ( c == '>' )
{
// All is well.
return;
}
}
}
#endif
const char* TiXmlDeclaration::Parse( const char* p, TiXmlParsingData* data, TiXmlEncoding _encoding )
{
p = SkipWhiteSpace( p, _encoding );
( run in 0.746 second using v1.01-cache-2.11-cpan-39bf76dae61 )