executor/xeXMLParser.cpp - third_party/vulkan-cts - Git at Google

 /*-------------------------------------------------------------------------
  * drawElements Quality Program Test Executor
  * ------------------------------------------
  *
  * Copyright 2014 The Android Open Source Project
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
  * You may obtain a copy of the License at
  *
  *      http://www.apache.org/licenses/LICENSE-2.0
  *
  * Unless required by applicable law or agreed to in writing, software
  * distributed under the License is distributed on an "AS IS" BASIS,
  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  * See the License for the specific language governing permissions and
  * limitations under the License.
  *
  *//*!
  * \file
  * \brief XML Parser.
  *//*--------------------------------------------------------------------*/

 #include "xeXMLParser.hpp"
 #include "deInt32.h"

 namespace xe
 {
 namespace xml
 {

 enum
 {
 	TOKENIZER_INITIAL_BUFFER_SIZE	= 1024
 };

 static inline bool isIdentifierStartChar (int ch)
 {
 	return de::inRange<int>(ch, 'a', 'z') || de::inRange<int>(ch, 'A', 'Z');
 }

 static inline bool isIdentifierChar (int ch)
 {
 	return isIdentifierStartChar(ch) || de::inRange<int>(ch, '0', '9') || (ch == '-') || (ch == '_');
 }

 static inline bool isWhitespaceChar (int ch)
 {
 	return ch == ' ' || ch == '\t' || ch == '\r' || ch == '\n';
 }

 static int getNextBufferSize (int curSize, int minNewSize)
 {
 	return de::max(curSize*2, 1<<deLog2Ceil32(minNewSize));
 }

 Tokenizer::Tokenizer (void)
 	: m_curToken	(TOKEN_INCOMPLETE)
 	, m_curTokenLen	(0)
 	, m_state		(STATE_DATA)
 	, m_buf			(TOKENIZER_INITIAL_BUFFER_SIZE)
 {
 }

 Tokenizer::~Tokenizer (void)
 {
 }

 void Tokenizer::clear (void)
 {
 	m_curToken		= TOKEN_INCOMPLETE;
 	m_curTokenLen	= 0;
 	m_state			= STATE_DATA;
 	m_buf.clear();
 }

 void Tokenizer::error (const std::string& what)
 {
 	throw ParseError(what);
 }

 void Tokenizer::feed (const deUint8* bytes, int numBytes)
 {
 	// Grow buffer if necessary.
 	if (m_buf.getNumFree() < numBytes)
 	{
 		m_buf.resize(getNextBufferSize(m_buf.getSize(), m_buf.getNumElements()+numBytes));
 	}

 	// Append to front.
 	m_buf.pushFront(bytes, numBytes);

 	// If we haven't parsed complete token, re-try after data feed.
 	if (m_curToken == TOKEN_INCOMPLETE)
 		advance();
 }

 int Tokenizer::getChar (int offset) const
 {
 	DE_ASSERT(de::inRange(offset, 0, m_buf.getNumElements()));

 	if (offset < m_buf.getNumElements())
 		return m_buf.peekBack(offset);
 	else
 		return END_OF_BUFFER;
 }

 void Tokenizer::advance (void)
 {
 	if (m_curToken != TOKEN_INCOMPLETE)
 	{
 		// Parser should not try to advance beyond end of string.
 		DE_ASSERT(m_curToken != TOKEN_END_OF_STRING);

 		// If current token is tag end, change state to data.
 		if (m_curToken == TOKEN_TAG_END						||
 			m_curToken == TOKEN_EMPTY_ELEMENT_END			||
 			m_curToken == TOKEN_PROCESSING_INSTRUCTION_END	||
 			m_curToken == TOKEN_COMMENT						||
 			m_curToken == TOKEN_ENTITY)
 			m_state = STATE_DATA;

 		// Advance buffer by length of last token.
 		m_buf.popBack(m_curTokenLen);

 		// Reset state.
 		m_curToken		= TOKEN_INCOMPLETE;
 		m_curTokenLen	= 0;

 		// If we hit end of string here, report it as end of string.
 		if (getChar(0) == END_OF_STRING)
 		{
 			m_curToken		= TOKEN_END_OF_STRING;
 			m_curTokenLen	= 1;
 			return;
 		}
 	}

 	int curChar = getChar(m_curTokenLen);

 	for (;;)
 	{
 		if (m_state == STATE_DATA)
 		{
 			// Advance until we hit end of buffer or tag start and treat that as data token.
 			if (curChar == END_OF_STRING || curChar == (int)END_OF_BUFFER || curChar == '<' || curChar == '&')
 			{
 				if (curChar == '<')
 					m_state = STATE_TAG;
 				else if (curChar == '&')
 					m_state = STATE_ENTITY;

 				if (m_curTokenLen > 0)
 				{
 					// Report data token.
 					m_curToken = TOKEN_DATA;
 					return;
 				}
 				else if (curChar == END_OF_STRING || curChar == (int)END_OF_BUFFER)
 				{
 					// Just return incomplete token, no data parsed.
 					return;
 				}
 				else
 				{
 					DE_ASSERT(m_state == STATE_TAG || m_state == STATE_ENTITY);
 					continue;
 				}
 			}
 		}
 		else
 		{
 			// Eat all whitespace if present.
 			if (m_curTokenLen == 0)
 			{
 				while (isWhitespaceChar(curChar))
 				{
 					m_buf.popBack();
 					curChar = getChar(0);
 				}
 			}

 			// Handle end of string / buffer.
 			if (curChar == END_OF_STRING)
 				error("Unexpected end of string");
 			else if (curChar == (int)END_OF_BUFFER)
 			{
 				DE_ASSERT(m_curToken == TOKEN_INCOMPLETE);
 				return;
 			}

 			if (m_curTokenLen == 0)
 			{
 				// Expect start of identifier, value or special tag token.
 				if (curChar == '\'' || curChar == '"')
 					m_state = STATE_VALUE;
 				else if (isIdentifierStartChar(curChar))
 					m_state = STATE_IDENTIFIER;
 				else if (curChar == '<' || curChar == '?' || curChar == '/')
 					m_state = STATE_TAG;
 				else if (curChar == '&')
 					DE_ASSERT(m_state == STATE_ENTITY);
 				else if (curChar == '=')
 				{
 					m_curToken		= TOKEN_EQUAL;
 					m_curTokenLen	= 1;
 					return;
 				}
 				else if (curChar == '>')
 				{
 					m_curToken		= TOKEN_TAG_END;
 					m_curTokenLen	= 1;
 					return;
 				}
 				else
 					error("Unexpected character");
 			}
 			else if (m_state == STATE_IDENTIFIER)
 			{
 				if (!isIdentifierChar(curChar))
 				{
 					m_curToken = TOKEN_IDENTIFIER;
 					return;
 				}
 			}
 			else if (m_state == STATE_VALUE)
 			{
 				// \todo [2012-06-07 pyry] Escapes.
 				if (curChar == '\'' || curChar == '"')
 				{
 					// \todo [2012-10-17 pyry] Should we actually do the check against getChar(0)?
 					if (curChar != getChar(0))
 						error("Mismatched quote");
 					m_curToken		 = TOKEN_STRING;
 					m_curTokenLen	+= 1;
 					return;
 				}
 			}
 			else if (m_state == STATE_COMMENT)
 			{
 				DE_ASSERT(m_curTokenLen >= 2); // 2 characters have been parsed if we are in comment state.

 				if (m_curTokenLen <= 3)
 				{
 					if (curChar != '-')
 						error("Invalid comment start");
 				}
 				else
 				{
 					int prev2 = m_curTokenLen > 5 ? getChar(m_curTokenLen-2) : 0;
 					int prev1 = m_curTokenLen > 4 ? getChar(m_curTokenLen-1) : 0;

 					if (prev2 == '-' && prev1 == '-')
 					{
 						if (curChar != '>')
 							error("Invalid comment end");
 						m_curToken		 = TOKEN_COMMENT;
 						m_curTokenLen	+= 1;
 						return;
 					}
 				}
 			}
 			else if (m_state == STATE_ENTITY)
 			{
 				if (m_curTokenLen >= 1)
 				{
 					if (curChar == ';')
 					{
 						m_curToken		 = TOKEN_ENTITY;
 						m_curTokenLen	+= 1;
 						return;
 					}
 					else if (!de::inRange<int>(curChar, '0', '9')	&&
 							 !de::inRange<int>(curChar, 'a', 'z')	&&
 							 !de::inRange<int>(curChar, 'A', 'Z'))
 						error("Invalid entity");
 				}
 			}
 			else
 			{
 				// Special tokens are at most 2 characters.
 				DE_ASSERT(m_state == STATE_TAG && m_curTokenLen == 1);

 				int prevChar = getChar(m_curTokenLen-1);

 				if (prevChar == '<')
 				{
 					// Tag start.
 					if (curChar == '/')
 					{
 						m_curToken		= TOKEN_END_TAG_START;
 						m_curTokenLen	= 2;
 						return;
 					}
 					else if (curChar == '?')
 					{
 						m_curToken		= TOKEN_PROCESSING_INSTRUCTION_START;
 						m_curTokenLen	= 2;
 						return;
 					}
 					else if (curChar == '!')
 					{
 						m_state = STATE_COMMENT;
 					}
 					else
 					{
 						m_curToken		= TOKEN_TAG_START;
 						m_curTokenLen	= 1;
 						return;
 					}
 				}
 				else if (prevChar == '?')
 				{
 					if (curChar != '>')
 						error("Invalid processing instruction end");
 					m_curToken		= TOKEN_PROCESSING_INSTRUCTION_END;
 					m_curTokenLen	= 2;
 					return;
 				}
 				else if (prevChar == '/')
 				{
 					if (curChar != '>')
 						error("Invalid empty element end");
 					m_curToken		= TOKEN_EMPTY_ELEMENT_END;
 					m_curTokenLen	= 2;
 					return;
 				}
 				else
 					error("Could not parse special token");
 			}
 		}

 		m_curTokenLen	+= 1;
 		curChar			 = getChar(m_curTokenLen);
 	}
 }

 void Tokenizer::getString (std::string& dst) const
 {
 	DE_ASSERT(m_curToken == TOKEN_STRING);
 	dst.resize(m_curTokenLen-2);
 	for (int ndx = 0; ndx < m_curTokenLen-2; ndx++)
 		dst[ndx] = m_buf.peekBack(ndx+1);
 }

 Parser::Parser (void)
 	: m_element		(ELEMENT_INCOMPLETE)
 	, m_state		(STATE_DATA)
 {
 }

 Parser::~Parser (void)
 {
 }

 void Parser::clear (void)
 {
 	m_tokenizer.clear();
 	m_elementName.clear();
 	m_attributes.clear();
 	m_attribName.clear();
 	m_entityValue.clear();

 	m_element	= ELEMENT_INCOMPLETE;
 	m_state		= STATE_DATA;
 }

 void Parser::error (const std::string& what)
 {
 	throw ParseError(what);
 }

 void Parser::feed (const deUint8* bytes, int numBytes)
 {
 	m_tokenizer.feed(bytes, numBytes);

 	if (m_element == ELEMENT_INCOMPLETE)
 		advance();
 }

 void Parser::advance (void)
 {
 	if (m_element == ELEMENT_START)
 		m_attributes.clear();

 	// \note No token is advanced when element end is reported.
 	if (m_state == STATE_YIELD_EMPTY_ELEMENT_END)
 	{
 		DE_ASSERT(m_element == ELEMENT_START);
 		m_element	= ELEMENT_END;
 		m_state		= STATE_DATA;
 		return;
 	}

 	if (m_element != ELEMENT_INCOMPLETE)
 	{
 		m_tokenizer.advance();
 		m_element = ELEMENT_INCOMPLETE;
 	}

 	for (;;)
 	{
 		Token curToken = m_tokenizer.getToken();

 		// Skip comments.
 		while (curToken == TOKEN_COMMENT)
 		{
 			m_tokenizer.advance();
 			curToken = m_tokenizer.getToken();
 		}

 		if (curToken == TOKEN_INCOMPLETE)
 		{
 			DE_ASSERT(m_element == ELEMENT_INCOMPLETE);
 			return;
 		}

 		switch (m_state)
 		{
 			case STATE_ENTITY:
 				m_state = STATE_DATA;
 				// Fall-through

 			case STATE_DATA:
 				switch (curToken)
 				{
 					case TOKEN_DATA:
 						m_element = ELEMENT_DATA;
 						return;

 					case TOKEN_END_OF_STRING:
 						m_element = ELEMENT_END_OF_STRING;
 						return;

 					case TOKEN_TAG_START:
 						m_state = STATE_START_TAG_OPEN;
 						break;

 					case TOKEN_END_TAG_START:
 						m_state = STATE_END_TAG_OPEN;
 						break;

 					case TOKEN_PROCESSING_INSTRUCTION_START:
 						m_state = STATE_IN_PROCESSING_INSTRUCTION;
 						break;

 					case TOKEN_ENTITY:
 						m_state		= STATE_ENTITY;
 						m_element	= ELEMENT_DATA;
 						parseEntityValue();
 						return;

 					default:
 						error("Unexpected token");
 				}
 				break;

 			case STATE_IN_PROCESSING_INSTRUCTION:
 				if (curToken == TOKEN_PROCESSING_INSTRUCTION_END)
 					m_state = STATE_DATA;
 				else
 					if (curToken != TOKEN_IDENTIFIER && curToken != TOKEN_EQUAL && curToken != TOKEN_STRING)
 						error("Unexpected token in processing instruction");
 				break;

 			case STATE_START_TAG_OPEN:
 				if (curToken != TOKEN_IDENTIFIER)
 					error("Expected identifier");
 				m_tokenizer.getTokenStr(m_elementName);
 				m_state = STATE_ATTRIBUTE_LIST;
 				break;

 			case STATE_END_TAG_OPEN:
 				if (curToken != TOKEN_IDENTIFIER)
 					error("Expected identifier");
 				m_tokenizer.getTokenStr(m_elementName);
 				m_state = STATE_EXPECTING_END_TAG_CLOSE;
 				break;

 			case STATE_EXPECTING_END_TAG_CLOSE:
 				if (curToken != TOKEN_TAG_END)
 					error("Expected tag end");
 				m_state		= STATE_DATA;
 				m_element	= ELEMENT_END;
 				return;

 			case STATE_ATTRIBUTE_LIST:
 				if (curToken == TOKEN_IDENTIFIER)
 				{
 					m_tokenizer.getTokenStr(m_attribName);
 					m_state = STATE_EXPECTING_ATTRIBUTE_EQ;
 				}
 				else if (curToken == TOKEN_EMPTY_ELEMENT_END)
 				{
 					m_state		= STATE_YIELD_EMPTY_ELEMENT_END;
 					m_element	= ELEMENT_START;
 					return;
 				}
 				else if (curToken == TOKEN_TAG_END)
 				{
 					m_state		= STATE_DATA;
 					m_element	= ELEMENT_START;
 					return;
 				}
 				else
 					error("Unexpected token");
 				break;

 			case STATE_EXPECTING_ATTRIBUTE_EQ:
 				if (curToken != TOKEN_EQUAL)
 					error("Expected '='");
 				m_state = STATE_EXPECTING_ATTRIBUTE_VALUE;
 				break;

 			case STATE_EXPECTING_ATTRIBUTE_VALUE:
 				if (curToken != TOKEN_STRING)
 					error("Expected value");
 				if (hasAttribute(m_attribName.c_str()))
 					error("Duplicate attribute");

 				m_tokenizer.getString(m_attributes[m_attribName]);
 				m_state = STATE_ATTRIBUTE_LIST;
 				break;

 			default:
 				DE_ASSERT(false);
 		}

 		m_tokenizer.advance();
 	}
 }

 static char getEntityValue (const std::string& entity)
 {
 	static const struct
 	{
 		const char*		name;
 		char			value;
 	} s_entities[] =
 	{
 			{ "&lt;",			'<' },
 			{ "&gt;",			'>' },
 			{ "&amp;",			'&' },
 			{ "&apos;",			'\''},
 			{ "&quot;",			'"' },
 	};

 	for (int ndx = 0; ndx < DE_LENGTH_OF_ARRAY(s_entities); ndx++)
 	{
 		if (entity == s_entities[ndx].name)
 			return s_entities[ndx].value;
 	}

 	return 0;
 }

 void Parser::parseEntityValue (void)
 {
 	DE_ASSERT(m_state == STATE_ENTITY && m_tokenizer.getToken() == TOKEN_ENTITY);

 	std::string entity;
 	m_tokenizer.getTokenStr(entity);

 	const char value = getEntityValue(entity);
 	if (value == 0)
 		error("Invalid entity '" + entity + "'");

 	m_entityValue.resize(1);
 	m_entityValue[0] = value;
 }

 } // xml
 } // xe
	/*-------------------------------------------------------------------------
	* drawElements Quality Program Test Executor
	* ------------------------------------------
	*
	* Copyright 2014 The Android Open Source Project
	*
	* Licensed under the Apache License, Version 2.0 (the "License");
	* you may not use this file except in compliance with the License.
	* You may obtain a copy of the License at
	*
	* http://www.apache.org/licenses/LICENSE-2.0
	*
	* Unless required by applicable law or agreed to in writing, software
	* distributed under the License is distributed on an "AS IS" BASIS,
	* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
	* See the License for the specific language governing permissions and
	* limitations under the License.
	*
	//!
	* \file
	* \brief XML Parser.
	//--------------------------------------------------------------------*/

	#include "xeXMLParser.hpp"
	#include "deInt32.h"

	namespace xe
	{
	namespace xml
	{

	enum
	{
	TOKENIZER_INITIAL_BUFFER_SIZE = 1024
	};

	static inline bool isIdentifierStartChar (int ch)
	{
	return de::inRange<int>(ch, 'a', 'z') \|\| de::inRange<int>(ch, 'A', 'Z');
	}

	static inline bool isIdentifierChar (int ch)
	{
	return isIdentifierStartChar(ch) \|\| de::inRange<int>(ch, '0', '9') \|\| (ch == '-') \|\| (ch == '_');
	}

	static inline bool isWhitespaceChar (int ch)
	{
	return ch == ' ' \|\| ch == '\t' \|\| ch == '\r' \|\| ch == '\n';
	}

	static int getNextBufferSize (int curSize, int minNewSize)
	{
	return de::max(curSize*2, 1<<deLog2Ceil32(minNewSize));
	}

	Tokenizer::Tokenizer (void)
	: m_curToken (TOKEN_INCOMPLETE)
	, m_curTokenLen (0)
	, m_state (STATE_DATA)
	, m_buf (TOKENIZER_INITIAL_BUFFER_SIZE)
	{
	}

	Tokenizer::~Tokenizer (void)
	{
	}

	void Tokenizer::clear (void)
	{
	m_curToken = TOKEN_INCOMPLETE;
	m_curTokenLen = 0;
	m_state = STATE_DATA;
	m_buf.clear();
	}

	void Tokenizer::error (const std::string& what)
	{
	throw ParseError(what);
	}

	void Tokenizer::feed (const deUint8* bytes, int numBytes)
	{
	// Grow buffer if necessary.
	if (m_buf.getNumFree() < numBytes)
	{
	m_buf.resize(getNextBufferSize(m_buf.getSize(), m_buf.getNumElements()+numBytes));
	}

	// Append to front.
	m_buf.pushFront(bytes, numBytes);

	// If we haven't parsed complete token, re-try after data feed.
	if (m_curToken == TOKEN_INCOMPLETE)
	advance();
	}

	int Tokenizer::getChar (int offset) const
	{
	DE_ASSERT(de::inRange(offset, 0, m_buf.getNumElements()));

	if (offset < m_buf.getNumElements())
	return m_buf.peekBack(offset);
	else
	return END_OF_BUFFER;
	}

	void Tokenizer::advance (void)
	{
	if (m_curToken != TOKEN_INCOMPLETE)
	{
	// Parser should not try to advance beyond end of string.
	DE_ASSERT(m_curToken != TOKEN_END_OF_STRING);

	// If current token is tag end, change state to data.
	if (m_curToken == TOKEN_TAG_END \|\|
	m_curToken == TOKEN_EMPTY_ELEMENT_END \|\|
	m_curToken == TOKEN_PROCESSING_INSTRUCTION_END \|\|
	m_curToken == TOKEN_COMMENT \|\|
	m_curToken == TOKEN_ENTITY)
	m_state = STATE_DATA;

	// Advance buffer by length of last token.
	m_buf.popBack(m_curTokenLen);

	// Reset state.
	m_curToken = TOKEN_INCOMPLETE;
	m_curTokenLen = 0;

	// If we hit end of string here, report it as end of string.
	if (getChar(0) == END_OF_STRING)
	{
	m_curToken = TOKEN_END_OF_STRING;
	m_curTokenLen = 1;
	return;
	}
	}

	int curChar = getChar(m_curTokenLen);

	for (;;)
	{
	if (m_state == STATE_DATA)
	{
	// Advance until we hit end of buffer or tag start and treat that as data token.
	if (curChar == END_OF_STRING \|\| curChar == (int)END_OF_BUFFER \|\| curChar == '<' \|\| curChar == '&')
	{
	if (curChar == '<')
	m_state = STATE_TAG;
	else if (curChar == '&')
	m_state = STATE_ENTITY;

	if (m_curTokenLen > 0)
	{
	// Report data token.
	m_curToken = TOKEN_DATA;
	return;
	}
	else if (curChar == END_OF_STRING \|\| curChar == (int)END_OF_BUFFER)
	{
	// Just return incomplete token, no data parsed.
	return;
	}
	else
	{
	DE_ASSERT(m_state == STATE_TAG \|\| m_state == STATE_ENTITY);
	continue;
	}
	}
	}
	else
	{
	// Eat all whitespace if present.
	if (m_curTokenLen == 0)
	{
	while (isWhitespaceChar(curChar))
	{
	m_buf.popBack();
	curChar = getChar(0);
	}
	}

	// Handle end of string / buffer.
	if (curChar == END_OF_STRING)
	error("Unexpected end of string");
	else if (curChar == (int)END_OF_BUFFER)
	{
	DE_ASSERT(m_curToken == TOKEN_INCOMPLETE);
	return;
	}

	if (m_curTokenLen == 0)
	{
	// Expect start of identifier, value or special tag token.
	if (curChar == '\'' \|\| curChar == '"')
	m_state = STATE_VALUE;
	else if (isIdentifierStartChar(curChar))
	m_state = STATE_IDENTIFIER;
	else if (curChar == '<' \|\| curChar == '?' \|\| curChar == '/')
	m_state = STATE_TAG;
	else if (curChar == '&')
	DE_ASSERT(m_state == STATE_ENTITY);
	else if (curChar == '=')
	{
	m_curToken = TOKEN_EQUAL;
	m_curTokenLen = 1;
	return;
	}
	else if (curChar == '>')
	{
	m_curToken = TOKEN_TAG_END;
	m_curTokenLen = 1;
	return;
	}
	else
	error("Unexpected character");
	}
	else if (m_state == STATE_IDENTIFIER)
	{
	if (!isIdentifierChar(curChar))
	{
	m_curToken = TOKEN_IDENTIFIER;
	return;
	}
	}
	else if (m_state == STATE_VALUE)
	{
	// \todo [2012-06-07 pyry] Escapes.
	if (curChar == '\'' \|\| curChar == '"')
	{
	// \todo [2012-10-17 pyry] Should we actually do the check against getChar(0)?
	if (curChar != getChar(0))
	error("Mismatched quote");
	m_curToken = TOKEN_STRING;
	m_curTokenLen += 1;
	return;
	}
	}
	else if (m_state == STATE_COMMENT)
	{
	DE_ASSERT(m_curTokenLen >= 2); // 2 characters have been parsed if we are in comment state.

	if (m_curTokenLen <= 3)
	{
	if (curChar != '-')
	error("Invalid comment start");
	}
	else
	{
	int prev2 = m_curTokenLen > 5 ? getChar(m_curTokenLen-2) : 0;
	int prev1 = m_curTokenLen > 4 ? getChar(m_curTokenLen-1) : 0;

	if (prev2 == '-' && prev1 == '-')
	{
	if (curChar != '>')
	error("Invalid comment end");
	m_curToken = TOKEN_COMMENT;
	m_curTokenLen += 1;
	return;
	}
	}
	}
	else if (m_state == STATE_ENTITY)
	{
	if (m_curTokenLen >= 1)
	{
	if (curChar == ';')
	{
	m_curToken = TOKEN_ENTITY;
	m_curTokenLen += 1;
	return;
	}
	else if (!de::inRange<int>(curChar, '0', '9') &&
	!de::inRange<int>(curChar, 'a', 'z') &&
	!de::inRange<int>(curChar, 'A', 'Z'))
	error("Invalid entity");
	}
	}
	else
	{
	// Special tokens are at most 2 characters.
	DE_ASSERT(m_state == STATE_TAG && m_curTokenLen == 1);

	int prevChar = getChar(m_curTokenLen-1);

	if (prevChar == '<')
	{
	// Tag start.
	if (curChar == '/')
	{
	m_curToken = TOKEN_END_TAG_START;
	m_curTokenLen = 2;
	return;
	}
	else if (curChar == '?')
	{
	m_curToken = TOKEN_PROCESSING_INSTRUCTION_START;
	m_curTokenLen = 2;
	return;
	}
	else if (curChar == '!')
	{
	m_state = STATE_COMMENT;
	}
	else
	{
	m_curToken = TOKEN_TAG_START;
	m_curTokenLen = 1;
	return;
	}
	}
	else if (prevChar == '?')
	{
	if (curChar != '>')
	error("Invalid processing instruction end");
	m_curToken = TOKEN_PROCESSING_INSTRUCTION_END;
	m_curTokenLen = 2;
	return;
	}
	else if (prevChar == '/')
	{
	if (curChar != '>')
	error("Invalid empty element end");
	m_curToken = TOKEN_EMPTY_ELEMENT_END;
	m_curTokenLen = 2;
	return;
	}
	else
	error("Could not parse special token");
	}
	}

	m_curTokenLen += 1;
	curChar = getChar(m_curTokenLen);
	}
	}

	void Tokenizer::getString (std::string& dst) const
	{
	DE_ASSERT(m_curToken == TOKEN_STRING);
	dst.resize(m_curTokenLen-2);
	for (int ndx = 0; ndx < m_curTokenLen-2; ndx++)
	dst[ndx] = m_buf.peekBack(ndx+1);
	}

	Parser::Parser (void)
	: m_element (ELEMENT_INCOMPLETE)
	, m_state (STATE_DATA)
	{
	}

	Parser::~Parser (void)
	{
	}

	void Parser::clear (void)
	{
	m_tokenizer.clear();
	m_elementName.clear();
	m_attributes.clear();
	m_attribName.clear();
	m_entityValue.clear();

	m_element = ELEMENT_INCOMPLETE;
	m_state = STATE_DATA;
	}

	void Parser::error (const std::string& what)
	{
	throw ParseError(what);
	}

	void Parser::feed (const deUint8* bytes, int numBytes)
	{
	m_tokenizer.feed(bytes, numBytes);

	if (m_element == ELEMENT_INCOMPLETE)
	advance();
	}

	void Parser::advance (void)
	{
	if (m_element == ELEMENT_START)
	m_attributes.clear();

	// \note No token is advanced when element end is reported.
	if (m_state == STATE_YIELD_EMPTY_ELEMENT_END)
	{
	DE_ASSERT(m_element == ELEMENT_START);
	m_element = ELEMENT_END;
	m_state = STATE_DATA;
	return;
	}

	if (m_element != ELEMENT_INCOMPLETE)
	{
	m_tokenizer.advance();
	m_element = ELEMENT_INCOMPLETE;
	}

	for (;;)
	{
	Token curToken = m_tokenizer.getToken();

	// Skip comments.
	while (curToken == TOKEN_COMMENT)
	{
	m_tokenizer.advance();
	curToken = m_tokenizer.getToken();
	}

	if (curToken == TOKEN_INCOMPLETE)
	{
	DE_ASSERT(m_element == ELEMENT_INCOMPLETE);
	return;
	}

	switch (m_state)
	{
	case STATE_ENTITY:
	m_state = STATE_DATA;
	// Fall-through

	case STATE_DATA:
	switch (curToken)
	{
	case TOKEN_DATA:
	m_element = ELEMENT_DATA;
	return;

	case TOKEN_END_OF_STRING:
	m_element = ELEMENT_END_OF_STRING;
	return;

	case TOKEN_TAG_START:
	m_state = STATE_START_TAG_OPEN;
	break;

	case TOKEN_END_TAG_START:
	m_state = STATE_END_TAG_OPEN;
	break;

	case TOKEN_PROCESSING_INSTRUCTION_START:
	m_state = STATE_IN_PROCESSING_INSTRUCTION;
	break;

	case TOKEN_ENTITY:
	m_state = STATE_ENTITY;
	m_element = ELEMENT_DATA;
	parseEntityValue();
	return;

	default:
	error("Unexpected token");
	}
	break;

	case STATE_IN_PROCESSING_INSTRUCTION:
	if (curToken == TOKEN_PROCESSING_INSTRUCTION_END)
	m_state = STATE_DATA;
	else
	if (curToken != TOKEN_IDENTIFIER && curToken != TOKEN_EQUAL && curToken != TOKEN_STRING)
	error("Unexpected token in processing instruction");
	break;

	case STATE_START_TAG_OPEN:
	if (curToken != TOKEN_IDENTIFIER)
	error("Expected identifier");
	m_tokenizer.getTokenStr(m_elementName);
	m_state = STATE_ATTRIBUTE_LIST;
	break;

	case STATE_END_TAG_OPEN:
	if (curToken != TOKEN_IDENTIFIER)
	error("Expected identifier");
	m_tokenizer.getTokenStr(m_elementName);
	m_state = STATE_EXPECTING_END_TAG_CLOSE;
	break;

	case STATE_EXPECTING_END_TAG_CLOSE:
	if (curToken != TOKEN_TAG_END)
	error("Expected tag end");
	m_state = STATE_DATA;
	m_element = ELEMENT_END;
	return;

	case STATE_ATTRIBUTE_LIST:
	if (curToken == TOKEN_IDENTIFIER)
	{
	m_tokenizer.getTokenStr(m_attribName);
	m_state = STATE_EXPECTING_ATTRIBUTE_EQ;
	}
	else if (curToken == TOKEN_EMPTY_ELEMENT_END)
	{
	m_state = STATE_YIELD_EMPTY_ELEMENT_END;
	m_element = ELEMENT_START;
	return;
	}
	else if (curToken == TOKEN_TAG_END)
	{
	m_state = STATE_DATA;
	m_element = ELEMENT_START;
	return;
	}
	else
	error("Unexpected token");
	break;

	case STATE_EXPECTING_ATTRIBUTE_EQ:
	if (curToken != TOKEN_EQUAL)
	error("Expected '='");
	m_state = STATE_EXPECTING_ATTRIBUTE_VALUE;
	break;

	case STATE_EXPECTING_ATTRIBUTE_VALUE:
	if (curToken != TOKEN_STRING)
	error("Expected value");
	if (hasAttribute(m_attribName.c_str()))
	error("Duplicate attribute");

	m_tokenizer.getString(m_attributes[m_attribName]);
	m_state = STATE_ATTRIBUTE_LIST;
	break;

	default:
	DE_ASSERT(false);
	}

	m_tokenizer.advance();
	}
	}

	static char getEntityValue (const std::string& entity)
	{
	static const struct
	{
	const char* name;
	char value;
	} s_entities[] =
	{
	{ "<", '<' },
	{ ">", '>' },
	{ "&", '&' },
	{ "'", '\''},
	{ """, '"' },
	};

	for (int ndx = 0; ndx < DE_LENGTH_OF_ARRAY(s_entities); ndx++)
	{
	if (entity == s_entities[ndx].name)
	return s_entities[ndx].value;
	}

	return 0;
	}

	void Parser::parseEntityValue (void)
	{
	DE_ASSERT(m_state == STATE_ENTITY && m_tokenizer.getToken() == TOKEN_ENTITY);

	std::string entity;
	m_tokenizer.getTokenStr(entity);

	const char value = getEntityValue(entity);
	if (value == 0)
	error("Invalid entity '" + entity + "'");

	m_entityValue.resize(1);
	m_entityValue[0] = value;
	}

	} // xml
	} // xe