tinyxmlparser.cpp

lame 3.97源码.最好的mp3解码和压缩软件

/*
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 <ctype.h>
#include <strstream>
using namespace std;

//#define DEBUG_PARSER

TiXmlBase::Entity TiXmlBase::entity[ NUM_ENTITY ] = 
{
	{ "&amp;",  5, '&' },
	{ "&lt;",   4, '<' },
	{ "&gt;",   4, '>' },
	{ "&quot;", 6, '\"' },
	{ "&apos;", 6, '\'' }
};


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;
}


/*static*/ bool TiXmlBase::StreamWhiteSpace( std::istream* in, std::string* tag )
{
	for( ;; )
	{
		if ( !in->good() ) return false;

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


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

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


const char* TiXmlBase::ReadName( const char* p, 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 == ':' ) )
		{
			(*name) += *p;
			++p;
		}
		return p;
	}
	return 0;
}


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

	// Ignore the &#x entities.
	if ( strncmp( "&#x", p, 3 ) == 0 )
	{
		*value = *p;
		return p+1;
	}

	// Now try to match it.
	for( i=0; i<NUM_ENTITY; ++i )
	{
		if ( strncmp( entity[i].str, p, entity[i].strLength ) == 0 )
		{
			assert( strlen( entity[i].str ) == entity[i].strLength );
			*value = entity[i].chr;
			return ( p + entity[i].strLength );
		}
	}

	// So it wasn't an entity, its unrecognized, or something like that.
	*value = *p;	// Don't put back the last one, since we return it!
	return p+1;
}


bool TiXmlBase::StringEqual( const char* p,
							 const char* tag,
							 bool ignoreCase )
{
	assert( p );
	if ( !p || !*p )
	{
		assert( 0 );
		return false;
	}

    if ( tolower( *p ) == tolower( *tag ) )
	{
		const char* q = p;

		if (ignoreCase)
		{
			while ( *q && *tag && *q == *tag )
			{
				++q;
				++tag;
			}

			if ( *tag == 0 )		// Have we found the end of the tag, and everything equal?
			{
				return true;
			}
		}
		else
		{
			while ( *q && *tag && tolower( *q ) == tolower( *tag ) )
			{
				++q;
				++tag;
			}

			if ( *tag == 0 )
			{
				return true;
			}
		}
	}
	return false;
}


const char* TiXmlBase::ReadText(	const char* p, 
									string* text, 
									bool trimWhiteSpace, 
									const char* endTag, 
									bool caseInsensitive )
{
	*text = "";

	if (    !trimWhiteSpace			// certain tags always keep whitespace
		 || !condenseWhiteSpace )	// if true, whitespace is always kept
	{
		// Keep all the white space.
		while (	   p && *p
				&& !StringEqual( p, endTag, caseInsensitive )
			  )
		{
			char c;
			p = GetChar( p, &c );
			text->append( &c, 1 );
		}
	}
	else
	{
		bool whitespace = false;

		// Remove leading white space:
		p = SkipWhiteSpace( p );
		while (	   p && *p
				&& !StringEqual( p, endTag, caseInsensitive ) )
		{
			if ( *p == '\r' || *p == '\n' )
			{
				whitespace = true;
				++p;
			}
			else if ( isspace( *p ) )
			{
				whitespace = true;
				++p;
			}
			else
			{
				// If we've found whitespace, add it before the
				// new character. Any whitespace just becomes a space.
				if ( whitespace )
				{
					text->append( " ", 1 );
					whitespace = false;
				}
				char c;
				p = GetChar( p, &c );
				text->append( &c, 1 );
			}
		}
	}
	return p + strlen( endTag );
}


void TiXmlDocument::StreamIn( std::istream* in, std::string* tag )
{
	// The basic issue with a document is that we don't know what we're
	// streaming. Read something presumed to be a tag (and hope), then
	// identify it, and call the appropriate stream method on the tag.
	//
	// This "pre-streaming" will never read the closing ">" so the
	// sub-tag can orient itself.

	if ( !StreamTo( in, '<', tag ) ) 
	{
		SetError( TIXML_ERROR_PARSING_EMPTY );
		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 );
				return;
			}
		}
	}
	// We should have returned sooner. 
	SetError( TIXML_ERROR );
}


const char* TiXmlDocument::Parse( const char* p )
{
	// Parse away, at the document level. Since a document
	// contains nothing but other tags, most of what happens
	// here is skipping white space.
	//
	// In this variant (as opposed to stream and Parse) we
	// read everything we can.


	if ( !p || !*p  || !( p = SkipWhiteSpace( p ) ) )
	{
		SetError( TIXML_ERROR_DOCUMENT_EMPTY );
		return false;
	}
	
	while ( p && *p )
	{
		TiXmlNode* node = Identify( p );
		if ( node )
		{				
			p = node->Parse( p );
			LinkEndChild( node );
		}		
		else
		{
			break;
		}		
		p = SkipWhiteSpace( p );
	}
	// All is well.
	return p;
}


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: <!--
	// - Decleration: <?xml
	// - Everthing else is unknown to tinyxml.
	//

	const char* xmlHeader = { "<?xml" };
	const char* commentHeader = { "<!--" };

	if ( StringEqual( p, xmlHeader, true ) )
	{
		#ifdef DEBUG_PARSER
			TIXML_LOG( "XML parsing Declaration\n" );
		#endif
		returnNode = new TiXmlDeclaration();
	}
	else if (    isalpha( *(p+1) )
			  || *(p+1) == '_' )
	{
		#ifdef DEBUG_PARSER
			TIXML_LOG( "XML parsing Element\n" );
		#endif
		returnNode = new TiXmlElement( "" );
	}
	else if ( StringEqual( p, commentHeader, false ) )
	{
		#ifdef DEBUG_PARSER
			TIXML_LOG( "XML parsing Comment\n" );
		#endif
		returnNode = new TiXmlComment();
	}
	else
	{
		#ifdef DEBUG_PARSER
			TIXML_LOG( "XML parsing Unknown\n" );
		#endif
		returnNode = new TiXmlUnknown();
	}

	if ( returnNode )
	{
		// Set the parent, so it can report errors
		returnNode->parent = this;
		//p = returnNode->Parse( p );
	}
	else
	{
		if ( doc )
			doc->SetError( TIXML_ERROR_OUT_OF_MEMORY );
	}
	return returnNode;
}


void TiXmlElement::StreamIn( std::istream* in, std::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();
		(*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
		//		closing tag
		//		another node.
		for ( ;; )
		{
			StreamWhiteSpace( in, tag );

			// Do we have text?
			if ( 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 = tag->length();