michael@0: /* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
michael@0: /* This Source Code Form is subject to the terms of the Mozilla Public
michael@0:  * License, v. 2.0. If a copy of the MPL was not distributed with this
michael@0:  * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
michael@0: 
michael@0: 
michael@0: #ifndef MITREXSL_EXPRLEXER_H
michael@0: #define MITREXSL_EXPRLEXER_H
michael@0: 
michael@0: #include "txCore.h"
michael@0: #include "nsString.h"
michael@0: 
michael@0: /**
michael@0:  * A Token class for the ExprLexer.
michael@0:  *
michael@0:  * This class was ported from XSL:P, an open source Java based 
michael@0:  * XSLT processor, written by yours truly.
michael@0:  */
michael@0: class Token
michael@0: {
michael@0: public:
michael@0: 
michael@0:     /**
michael@0:      * Token types
michael@0:      */
michael@0:     enum Type {
michael@0:         //-- Trivial Tokens
michael@0:         NULL_TOKEN = 1,
michael@0:         LITERAL,
michael@0:         NUMBER,
michael@0:         CNAME,
michael@0:         VAR_REFERENCE,
michael@0:         PARENT_NODE,
michael@0:         SELF_NODE,
michael@0:         R_PAREN,
michael@0:         R_BRACKET, // 9
michael@0:         /**
michael@0:          * start of tokens for 3.7, bullet 1
michael@0:          * ExprLexer::nextIsOperatorToken bails if the tokens aren't
michael@0:          * consecutive.
michael@0:          */
michael@0:         COMMA,
michael@0:         AT_SIGN,
michael@0:         L_PAREN,
michael@0:         L_BRACKET,
michael@0:         AXIS_IDENTIFIER,
michael@0: 
michael@0:         // These tokens include their following left parenthesis
michael@0:         FUNCTION_NAME_AND_PAREN, // 15
michael@0:         COMMENT_AND_PAREN,
michael@0:         NODE_AND_PAREN,
michael@0:         PROC_INST_AND_PAREN,
michael@0:         TEXT_AND_PAREN,
michael@0: 
michael@0:         /**
michael@0:          * operators
michael@0:          */
michael@0:         //-- boolean ops
michael@0:         AND_OP, // 20
michael@0:         OR_OP,
michael@0: 
michael@0:         //-- relational
michael@0:         EQUAL_OP, // 22
michael@0:         NOT_EQUAL_OP,
michael@0:         LESS_THAN_OP,
michael@0:         GREATER_THAN_OP,
michael@0:         LESS_OR_EQUAL_OP,
michael@0:         GREATER_OR_EQUAL_OP,
michael@0:         //-- additive operators
michael@0:         ADDITION_OP, // 28
michael@0:         SUBTRACTION_OP,
michael@0:         //-- multiplicative
michael@0:         DIVIDE_OP, // 30
michael@0:         MULTIPLY_OP,
michael@0:         MODULUS_OP,
michael@0:         //-- path operators
michael@0:         PARENT_OP, // 33
michael@0:         ANCESTOR_OP,
michael@0:         UNION_OP,
michael@0:         /**
michael@0:          * end of tokens for 3.7, bullet 1 -/
michael@0:          */
michael@0:         //-- Special endtoken
michael@0:         END // 36
michael@0:     };
michael@0: 
michael@0: 
michael@0:     /**
michael@0:      * Constructors
michael@0:      */
michael@0:     typedef nsASingleFragmentString::const_char_iterator iterator;
michael@0: 
michael@0:     Token(iterator aStart, iterator aEnd, Type aType)
michael@0:         : mStart(aStart),
michael@0:           mEnd(aEnd),
michael@0:           mType(aType),
michael@0:           mNext(nullptr)
michael@0:     {
michael@0:     }
michael@0:     Token(iterator aChar, Type aType)
michael@0:         : mStart(aChar),
michael@0:           mEnd(aChar + 1),
michael@0:           mType(aType),
michael@0:           mNext(nullptr)
michael@0:     {
michael@0:     }
michael@0: 
michael@0:     const nsDependentSubstring Value()
michael@0:     {
michael@0:         return Substring(mStart, mEnd);
michael@0:     }
michael@0: 
michael@0:     iterator mStart, mEnd;
michael@0:     Type mType;
michael@0:     Token* mNext;
michael@0: };
michael@0: 
michael@0: /**
michael@0:  * A class for splitting an "Expr" String into tokens and
michael@0:  * performing  basic Lexical Analysis.
michael@0:  *
michael@0:  * This class was ported from XSL:P, an open source Java based XSL processor
michael@0:  */
michael@0: 
michael@0: class txExprLexer
michael@0: {
michael@0: public:
michael@0: 
michael@0:     txExprLexer();
michael@0:     ~txExprLexer();
michael@0: 
michael@0:     /**
michael@0:      * Parse the given string.
michael@0:      * returns an error result if lexing failed.
michael@0:      * The given string must outlive the use of the lexer, as the
michael@0:      * generated Tokens point to Substrings of it.
michael@0:      * mPosition points to the offending location in case of an error.
michael@0:      */
michael@0:     nsresult parse(const nsASingleFragmentString& aPattern);
michael@0: 
michael@0:     typedef nsASingleFragmentString::const_char_iterator iterator;
michael@0:     iterator mPosition;
michael@0: 
michael@0:     /**
michael@0:      * Functions for iterating over the TokenList
michael@0:      */
michael@0: 
michael@0:     Token* nextToken();
michael@0:     Token* peek()
michael@0:     {
michael@0:         NS_ASSERTION(mCurrentItem, "peek called uninitialized lexer");
michael@0:         return mCurrentItem;
michael@0:     }
michael@0:     Token* peekAhead()
michael@0:     {
michael@0:         NS_ASSERTION(mCurrentItem, "peekAhead called on uninitialized lexer");
michael@0:         // Don't peek past the end node
michael@0:         return (mCurrentItem && mCurrentItem->mNext) ? mCurrentItem->mNext : mCurrentItem;
michael@0:     }
michael@0:     bool hasMoreTokens()
michael@0:     {
michael@0:         NS_ASSERTION(mCurrentItem, "HasMoreTokens called on uninitialized lexer");
michael@0:         return (mCurrentItem && mCurrentItem->mType != Token::END);
michael@0:     }
michael@0: 
michael@0:     /**
michael@0:      * Trivial Tokens
michael@0:      */
michael@0:     //-- LF, changed to enum
michael@0:     enum _TrivialTokens {
michael@0:         D_QUOTE        = '\"',
michael@0:         S_QUOTE        = '\'',
michael@0:         L_PAREN        = '(',
michael@0:         R_PAREN        = ')',
michael@0:         L_BRACKET      = '[',
michael@0:         R_BRACKET      = ']',
michael@0:         L_ANGLE        = '<',
michael@0:         R_ANGLE        = '>',
michael@0:         COMMA          = ',',
michael@0:         PERIOD         = '.',
michael@0:         ASTERIX        = '*',
michael@0:         FORWARD_SLASH  = '/',
michael@0:         EQUAL          = '=',
michael@0:         BANG           = '!',
michael@0:         VERT_BAR       = '|',
michael@0:         AT_SIGN        = '@',
michael@0:         DOLLAR_SIGN    = '$',
michael@0:         PLUS           = '+',
michael@0:         HYPHEN         = '-',
michael@0:         COLON          = ':',
michael@0:         //-- whitespace tokens
michael@0:         SPACE          = ' ',
michael@0:         TX_TAB            = '\t',
michael@0:         TX_CR             = '\n',
michael@0:         TX_LF             = '\r'
michael@0:     };
michael@0: 
michael@0: private:
michael@0: 
michael@0:     Token* mCurrentItem;
michael@0:     Token* mFirstItem;
michael@0:     Token* mLastItem;
michael@0: 
michael@0:     int mTokenCount;
michael@0: 
michael@0:     void addToken(Token* aToken);
michael@0: 
michael@0:     /**
michael@0:      * Returns true if the following Token should be an operator.
michael@0:      * This is a helper for the first bullet of [XPath 3.7]
michael@0:      *  Lexical Structure
michael@0:      */
michael@0:     bool nextIsOperatorToken(Token* aToken);
michael@0: 
michael@0:     /**
michael@0:      * Returns true if the given character represents a numeric letter (digit)
michael@0:      * Implemented in ExprLexerChars.cpp
michael@0:      */
michael@0:     static bool isXPathDigit(char16_t ch)
michael@0:     {
michael@0:         return (ch >= '0' && ch <= '9');
michael@0:     }
michael@0: };
michael@0: 
michael@0: #endif
michael@0: