www.pudn.com > ICE_1.4.zip > tinyxmlparser.cpp

/*
www.sourceforge.net/projects/tinyxml
Original code (2.0 and earlier )copyright (c) 2000-2002 Lee Thomason (www.grinninglizard.com)

This software is provided 'as-is', without any express or implied 
warranty. In no event will the authors be held liable for any 
damages arising from the use of this software.

Permission is granted to anyone to use this software for any 
purpose, including commercial applications, and to alter it and 
redistribute it freely, subject to the following restrictions:

1. The origin of this software must not be misrepresented; you must 
not claim that you wrote the original software. If you use this
software in a product, an acknowledgment in the product documentation
would be appreciated but is not required.

2. Altered source versions must be plainly marked as such, and 
must not be misrepresented as being the original software.

3. This notice may not be removed or altered from any source 
distribution.
*/

#include "tinyxml.h"
#include 

//#define DEBUG_PARSER

// Note tha "PutString" hardcodes the same list. This
// is less flexible than it appears. Changing the entries
// or order will break putstring.	
TiXmlBase::Entity TiXmlBase::entity[ NUM_ENTITY ] = 
{
	{ "&",  5, '&' },
	{ "<",   4, '<' },
	{ ">",   4, '>' },
	{ """, 6, '\"' },
	{ "'", 6, '\'' }
};


class TiXmlParsingData
{
	friend class TiXmlDocument;
  public:
	//TiXmlParsingData( const char* now, const TiXmlParsingData* prevData );
	void Stamp( const char* now );

	const TiXmlCursor& Cursor()	{ return cursor; }
	//void Update( const char* now );

  private:
	// Only used by the document!
	TiXmlParsingData( const char* start, int _tabsize, int row, int col )
	{
		assert( start );
		stamp = start;
		tabsize = _tabsize;
		cursor.row = row;
		cursor.col = col;
	}

	TiXmlCursor		cursor;
	const char*		stamp;
	int				tabsize;
};


void TiXmlParsingData::Stamp( const char* now )
{
	assert( now );

	// Do nothing if the tabsize is 0.
	if ( tabsize < 1 )
	{
		return;
	}

	// Get the current row, column.
	int row = cursor.row;
	int col = cursor.col;
	const char* p = stamp;
	assert( p );

	while ( p < now )
	{
		// Code contributed by Fletcher Dunn: (modified by lee)
		switch (*p) {
			case 0:
				// We *should* never get here, but in case we do, don't
				// advance past the terminating null character, ever
				return;

			case '\r':
				// bump down to the next line
				++row;
				col = 0;				
				// Eat the character
				++p;

				// Check for \r\n sequence, and treat this as a single character
				if (*p == '\n') {
					++p;
				}
				break;

			case '\n':
				// bump down to the next line
				++row;
				col = 0;

				// Eat the character
				++p;

				// Check for \n\r sequence, and treat this as a single
				// character.  (Yes, this bizarre thing does occur still
				// on some arcane platforms...)
				if (*p == '\r') {
					++p;
				}
				break;

			case '\t':
				// Eat the character
				++p;

				// Skip to next tab stop
				col = (col / tabsize + 1) * tabsize;
				break;

			default:
				// Eat the character
				++p;

				// Normal char - just advance one column
				++col;
				break;
		}
	}
	cursor.row = row;
	cursor.col = col;
	assert( cursor.row >= -1 );
	assert( cursor.col >= -1 );
	stamp = p;
	assert( stamp );
}


const char* TiXmlBase::SkipWhiteSpace( const char* p )
{
	if ( !p || !*p )
	{
		return 0;
	}
	while ( p && *p )
	{
		if ( isspace( *p ) || *p == '\n' || *p =='\r' )		// Still using old rules for white space.
			++p;
		else
			break;
	}

	return p;
}

#ifdef TIXML_USE_STL
/*static*/ bool TiXmlBase::StreamWhiteSpace( TIXML_ISTREAM * in, TIXML_STRING * tag )
{
	for( ;; )
	{
		if ( !in->good() ) return false;

		int c = in->peek();
		if ( !IsWhiteSpace( c ) )
			return true;
		*tag += in->get();
	}
}

/*static*/ bool TiXmlBase::StreamTo( TIXML_ISTREAM * in, int character, TIXML_STRING * tag )
{
	while ( in->good() )
	{
		int c = in->peek();
		if ( c == character )
			return true;

		in->get();
		*tag += c;
	}
	return false;
}
#endif

const char* TiXmlBase::ReadName( const char* p, TIXML_STRING * name )
{
	*name = "";
	assert( p );

	// Names start with letters or underscores.
	// After that, they can be letters, underscores, numbers,
	// hyphens, or colons. (Colons are valid ony for namespaces,
	// but tinyxml can't tell namespaces from names.)
	if (    p && *p 
		 && ( isalpha( (unsigned char) *p ) || *p == '_' ) )
	{
		while(		p && *p
				&&	(		isalnum( (unsigned char ) *p ) 
						 || *p == '_'
						 || *p == '-'
						 || *p == '.'
						 || *p == ':' ) )
		{
			(*name) += *p;
			++p;
		}
		return p;
	}
	return 0;
}

const char* TiXmlBase::GetEntity( const char* p, char* value )
{
	// Presume an entity, and pull it out.
    TIXML_STRING ent;
	int i;

	// Handle the &#x entities.
	if (    strncmp( "&#x", p, 3 ) == 0 
	     && *(p+3) 
		 && *(p+4) 
		 && ( *(p+4) == ';' || *(p+5) == ';' )
	   )
	{
		*value = 0;

		if ( *(p+4) == ';' )
		{
			// Short, one value entity.
			if ( isalpha( *(p+3) ) ) *value += ( tolower( *(p+3) ) - 'a' + 10 );
			else				     *value += ( *(p+3) - '0' );

			return p+5;
		}
		else
		{
			// two value entity
			if ( isalpha( *(p+3) ) ) *value += ( tolower( *(p+3) ) - 'a' + 10 ) * 16;
			else				     *value += ( *(p+3) - '0' ) * 16;

			if ( isalpha( *(p+4) ) ) *value += ( tolower( *(p+4) ) - 'a' + 10 );
			else				     *value += ( *(p+4) - '0' );

			return p+6;
		}
	}

	// Now try to match it.
	for( i=0; i" so the
	// sub-tag can orient itself.

	if ( !StreamTo( in, '<', tag ) ) 
	{
		SetError( TIXML_ERROR_PARSING_EMPTY, 0, 0 );
		return;
	}

	while ( in->good() )
	{
		int tagIndex = tag->length();
		while ( in->good() && in->peek() != '>' )
		{
			int c = in->get();
			(*tag) += (char) c;
		}

		if ( in->good() )
		{
			// We now have something we presume to be a node of 
			// some sort. Identify it, and call the node to
			// continue streaming.
			TiXmlNode* node = Identify( tag->c_str() + tagIndex );

			if ( node )
			{
				node->StreamIn( in, tag );
				bool isElement = node->ToElement() != 0;
				delete node;
				node = 0;

				// If this is the root element, we're done. Parsing will be
				// done by the >> operator.
				if ( isElement )
				{
					return;
				}
			}
			else
			{
				SetError( TIXML_ERROR, 0, 0 );
				return;
			}
		}
	}
	// We should have returned sooner.
	SetError( TIXML_ERROR, 0, 0 );
}

#endif

const char* TiXmlDocument::Parse( const char* p, TiXmlParsingData* prevData )
{
	ClearError();

	// Parse away, at the document level. Since a document
	// contains nothing but other tags, most of what happens
	// here is skipping white space.
	if ( !p || !*p )
	{
		SetError( TIXML_ERROR_DOCUMENT_EMPTY, 0, 0 );
		return 0;
	}

	// Note that, for a document, this needs to come
	// before the while space skip, so that parsing
	// starts from the pointer we are given.
	location.Clear();
	if ( prevData )
	{
		location.row = prevData->cursor.row;
		location.col = prevData->cursor.col;
	}
	else
	{
		location.row = 0;
		location.col = 0;
	}
	TiXmlParsingData data( p, TabSize(), location.row, location.col );
	location = data.Cursor();

    p = SkipWhiteSpace( p );
	if ( !p )
	{
		SetError( TIXML_ERROR_DOCUMENT_EMPTY, 0, 0 );
		return 0;
	}

	while ( p && *p )
	{
		TiXmlNode* node = Identify( p );
		if ( node )
		{
			p = node->Parse( p, &data );
			LinkEndChild( node );
		}
		else
		{
			break;
		}
		p = SkipWhiteSpace( p );
	}

	// All is well.
	return p;
}

void TiXmlDocument::SetError( int err, const char* pError, TiXmlParsingData* data )
{	
	// 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 )
	{
		//TiXmlParsingData data( pError, prevData );
		data->Stamp( pError );
		errorLocation = data->Cursor();
	}
}


TiXmlNode* TiXmlNode::Identify( const char* p )
{
	TiXmlNode* returnNode = 0;

	p = SkipWhiteSpace( p );
	if( !p || !*p || *p != '<' )
	{
		return 0;
	}

	TiXmlDocument* doc = GetDocument();
	p = SkipWhiteSpace( p );

	if ( !p || !*p )
	{
		return 0;
	}

	// What is this thing? 
	// - Elements start with a letter or underscore, but xml is reserved.
	// - Comments: ";

	if ( !StringEqual( p, startTag, false ) )
	{
		document->SetError( TIXML_ERROR_PARSING_COMMENT, p, data );
		return 0;
	}
	p += strlen( startTag );
	p = ReadText( p, &value, false, endTag, false );
	return p;
}


const char* TiXmlAttribute::Parse( const char* p, TiXmlParsingData* data )
{
	p = SkipWhiteSpace( p );
	if ( !p || !*p ) return 0;

	int tabsize = 4;
	if ( document )
		tabsize = document->TabSize();

//	TiXmlParsingData data( p, prevData );
	if ( data )
	{
		data->Stamp( p );
		location = data->Cursor();
	}
	// Read the name, the '=' and the value.
	const char* pErr = p;
	p = ReadName( p, &name );
	if ( !p || !*p )
	{
		if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, pErr, data );
		return 0;
	}
	p = SkipWhiteSpace( p );
	if ( !p || !*p || *p != '=' )
	{
		if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, p, data );
		return 0;
	}

	++p;	// skip '='
	p = SkipWhiteSpace( p );
	if ( !p || !*p )
	{
		if ( document ) document->SetError( TIXML_ERROR_READING_ATTRIBUTES, p, data );
		return 0;
	}
	
	const char* end;

	if ( *p == '\'' )
	{
		++p;
		end = "\'";
		p = ReadText( p, &value, false, end, false );
	}
	else if ( *p == '"' )
	{
		++p;
		end = "\"";
		p = ReadText( p, &value, false, end, false );
	}
	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
				&& !isspace( *p ) && *p != '\n' && *p != '\r'	// whitespace
				&& *p != '/' && *p != '>' )						// tag end
		{
			value += *p;
			++p;
		}
	}
	return p;
}

#ifdef TIXML_USE_STL
void TiXmlText::StreamIn( TIXML_ISTREAM * in, TIXML_STRING * tag )
{
	while ( in->good() )
	{
		int c = in->peek();	
		if ( c == '<' )
			return;

		(*tag) += c;
		in->get();
	}
}
#endif

const char* TiXmlText::Parse( const char* p, TiXmlParsingData* data )
{
	value = "";
//	TiXmlParsingData data( p, prevData );
	if ( data )
	{
		data->Stamp( p );
		location = data->Cursor();
	}
	bool ignoreWhite = true;

	const char* end = "<";
	p = ReadText( p, &value, ignoreWhite, end, false );
	if ( p )
		return p-1;	// don't truncate the '<'
	return 0;
}

#ifdef TIXML_USE_STL
void TiXmlDeclaration::StreamIn( TIXML_ISTREAM * in, TIXML_STRING * tag )
{
	while ( in->good() )
	{
		int c = in->get();
		(*tag) += c;

		if ( c == '>' )
		{
			// All is well.
			return;
		}
	}
}
#endif

const char* TiXmlDeclaration::Parse( const char* p, TiXmlParsingData* data )
{
	p = SkipWhiteSpace( p );
	// Find the beginning, find the end, and look for
	// the stuff in-between.
	TiXmlDocument* document = GetDocument();
	if ( !p || !*p || !StringEqual( p, "SetError( TIXML_ERROR_PARSING_DECLARATION, 0, 0 );
		return 0;
	}
//	TiXmlParsingData data( p, prevData );
	if ( data )
	{
		data->Stamp( p );
		location = data->Cursor();
	}
	p += 5;

	version = "";
	encoding = "";
	standalone = "";

	while ( p && *p )
	{
		if ( *p == '>' )
		{
			++p;
			return p;
		}

		p = SkipWhiteSpace( p );
		if ( StringEqual( p, "version", true ) )
		{
			TiXmlAttribute attrib;
			p = attrib.Parse( p, data );		
			version = attrib.Value();
		}
		else if ( StringEqual( p, "encoding", true ) )
		{
			TiXmlAttribute attrib;
			p = attrib.Parse( p, data );		
			encoding = attrib.Value();
		}
		else if ( StringEqual( p, "standalone", true ) )
		{
			TiXmlAttribute attrib;
			p = attrib.Parse( p, data );		
			standalone = attrib.Value();
		}
		else
		{
			// Read over whatever it is.
			while( p && *p && *p != '>' && !isspace( *p ) )
				++p;
		}
	}
	return 0;
}

bool TiXmlText::Blank() const
{
	for ( unsigned i=0; i