src/antlr/Parser.hpp

#ifndef INC_Parser_hpp__
#define INC_Parser_hpp__

/* ANTLR Translator Generator
 * Project led by Terence Parr at http://www.jGuru.com
 * Software rights: http://www.antlr.org/license.html
 *
 * $Id$
 */

#include <antlr/config.hpp>

#include <iostream>
#include <exception>

#include <antlr/BitSet.hpp>
#include <antlr/TokenBuffer.hpp>
#include <antlr/RecognitionException.hpp>
#include <antlr/MismatchedTokenException.hpp>
#include <antlr/ASTFactory.hpp>
#include <antlr/ParserSharedInputState.hpp>

#ifdef ANTLR_CXX_SUPPORTS_NAMESPACE
namespace antlr {
#endif

extern bool DEBUG_PARSER;

/** A generic ANTLR parser (LL(k) for k>=1) containing a bunch of
 * utility routines useful at any lookahead depth.  We distinguish between
 * the LL(1) and LL(k) parsers because of efficiency.  This may not be
 * necessary in the near future.
 *
 * Each parser object contains the state of the parse including a lookahead
 * cache (the form of which is determined by the subclass), whether or
 * not the parser is in guess mode, where tokens come from, etc...
 *
 * <p>
 * During <b>guess</b> mode, the current lookahead token(s) and token type(s)
 * cache must be saved because the token stream may not have been informed
 * to save the token (via <tt>mark</tt>) before the <tt>try</tt> block.
 * Guessing is started by:
 * <ol>
 * <li>saving the lookahead cache.
 * <li>marking the current position in the TokenBuffer.
 * <li>increasing the guessing level.
 * </ol>
 *
 * After guessing, the parser state is restored by:
 * <ol>
 * <li>restoring the lookahead cache.
 * <li>rewinding the TokenBuffer.
 * <li>decreasing the guessing level.
 * </ol>
 *
 * @see antlr.Token
 * @see antlr.TokenBuffer
 * @see antlr.TokenStream
 * @see antlr.LL1Parser
 * @see antlr.LLkParser
 *
 * @todo add constructors with ASTFactory.
 */
class ANTLR_API Parser {
protected:
        Parser(TokenBuffer& input)
        : inputState(new ParserInputState(input)), astFactory(0), traceDepth(0)
        {
        }
        Parser(TokenBuffer* input)
        : inputState(new ParserInputState(input)), astFactory(0), traceDepth(0)
        {
        }
        Parser(const ParserSharedInputState& state)
        : inputState(state), astFactory(0), traceDepth(0)
        {
        }
public:
        virtual ~Parser()
        {
        }

        /** Return the token type of the ith token of lookahead where i=1
         * is the current token being examined by the parser (i.e., it
         * has not been matched yet).
         */
        virtual int LA(unsigned int i)=0;

        /// Return the i-th token of lookahead
        virtual RefToken LT(unsigned int i)=0;

        /** DEPRECATED! Specify the factory to be used during tree building. (Compulsory)
         * Setting the factory is nowadays compulsory.
         * @see setASTFactory
         */
        virtual void setASTNodeFactory( ASTFactory *factory )
        {
                astFactory = factory;
        }
        /** Specify the factory to be used during tree building. (Compulsory)
         * Setting the factory is nowadays compulsory.
         */
        virtual void setASTFactory( ASTFactory *factory )
        {
                astFactory = factory;
        }
        /** Return a pointer to the ASTFactory used.
         * So you might use it in subsequent treewalkers or to reload AST's
         * from disk.
         */
        virtual ASTFactory* getASTFactory()
        {
                return astFactory;
        }
        /** Get the root AST node of the generated AST. When using a custom AST type
         * or heterogenous AST's, you'll have to convert it to the right type
         * yourself.
         */
        virtual RefAST getAST() = 0;

        /// Return the filename of the input file.
        virtual inline ANTLR_USE_NAMESPACE(std)string getFilename() const
        {
                return inputState->filename;
        }
        /// Set the filename of the input file (used for error reporting).
        virtual void setFilename(const ANTLR_USE_NAMESPACE(std)string& f)
        {
                inputState->filename = f;
        }

        virtual void setInputState(ParserSharedInputState state)
        {
                inputState = state;
        }
        virtual inline ParserSharedInputState getInputState() const
        {
                return inputState;
        }

        /// Get another token object from the token stream
        virtual void consume()=0;
        /// Consume tokens until one matches the given token
        virtual void consumeUntil(int tokenType)
        {
                while (LA(1) != Token::EOF_TYPE && LA(1) != tokenType)
                        consume();
        }

        /// Consume tokens until one matches the given token set
        virtual void consumeUntil(const BitSet& set)
        {
                while (LA(1) != Token::EOF_TYPE && !set.member(LA(1)))
                        consume();
        }

        /** Make sure current lookahead symbol matches token type <tt>t</tt>.
         * Throw an exception upon mismatch, which is catch by either the
         * error handler or by the syntactic predicate.
         */
        virtual void match(int t)
        {
                if ( DEBUG_PARSER )
                {
                        traceIndent();
                        ANTLR_USE_NAMESPACE(std)cout << "enter match(" << t << ") with LA(1)=" << LA(1) << ANTLR_USE_NAMESPACE(std)endl;
                }
                if ( LA(1) != t )
                {
                        if ( DEBUG_PARSER )
                        {
                                traceIndent();
                                ANTLR_USE_NAMESPACE(std)cout << "token mismatch: " << LA(1) << "!=" << t << ANTLR_USE_NAMESPACE(std)endl;
                        }
                        throw MismatchedTokenException(getTokenNames(), getNumTokens(), LT(1), t, false, getFilename());
                }
                else
                {
                        // mark token as consumed -- fetch next token deferred until LA/LT
                        consume();
                }
        }

        virtual void matchNot(int t)
        {
                if ( LA(1)==t )
                {
                        // Throws inverted-sense exception
                        throw MismatchedTokenException(getTokenNames(), getNumTokens(), LT(1), t, true, getFilename());
                }
                else
                {
                        // mark token as consumed -- fetch next token deferred until LA/LT
                        consume();
                }
        }

        /** Make sure current lookahead symbol matches the given set
         * Throw an exception upon mismatch, which is catch by either the
         * error handler or by the syntactic predicate.
         */
        virtual void match(const BitSet& b)
        {
                if ( DEBUG_PARSER )
                {
                        traceIndent();
                        ANTLR_USE_NAMESPACE(std)cout << "enter match(" << "bitset" /*b.toString()*/
                                << ") with LA(1)=" << LA(1) << ANTLR_USE_NAMESPACE(std)endl;
                }
                if ( !b.member(LA(1)) )
                {
                        if ( DEBUG_PARSER )
                        {
                                traceIndent();
                                ANTLR_USE_NAMESPACE(std)cout << "token mismatch: " << LA(1) << " not member of "
                                        << "bitset" /*b.toString()*/ << ANTLR_USE_NAMESPACE(std)endl;
                        }
                        throw MismatchedTokenException(getTokenNames(), getNumTokens(), LT(1), b, false, getFilename());
                }
                else
                {
                        // mark token as consumed -- fetch next token deferred until LA/LT
                        consume();
                }
        }

        /** Mark a spot in the input and return the position.
         * Forwarded to TokenBuffer.
         */
        virtual inline unsigned int mark()
        {
                return inputState->getInput().mark();
        }
        /// rewind to a previously marked position
        virtual inline void rewind(unsigned int pos)
        {
                inputState->getInput().rewind(pos);
        }
        /** called by the generated parser to do error recovery, override to
         * customize the behaviour.
         */
        virtual void recover(const RecognitionException& ex, const BitSet& tokenSet)
        {
                consume();
                consumeUntil(tokenSet);
        }

        /// Parser error-reporting function can be overridden in subclass
        virtual void reportError(const RecognitionException& ex);
        /// Parser error-reporting function can be overridden in subclass
        virtual void reportError(const ANTLR_USE_NAMESPACE(std)string& s);
        /// Parser warning-reporting function can be overridden in subclass
        virtual void reportWarning(const ANTLR_USE_NAMESPACE(std)string& s);

        /// get the token name for the token number 'num'
        virtual const char* getTokenName(int num) const = 0;
        /// get a vector with all token names
        virtual const char* const* getTokenNames() const = 0;
        /** Get the number of tokens defined.
         * This one should be overridden in subclasses.
         */
        virtual int getNumTokens(void) const = 0;

        /** Set or change the input token buffer */
//      void setTokenBuffer(TokenBuffer<Token>* t);

        virtual void traceIndent();
        virtual void traceIn(const char* rname);
        virtual void traceOut(const char* rname);
protected:
//      void setTokenNames(const char** tokenNames_);

        ParserSharedInputState inputState;

//      /// AST return value for a rule is squirreled away here
//      RefAST returnAST;

        /// AST support code; parser and treeparser delegate to this object
        ASTFactory *astFactory;

        // used to keep track of the indentation for the trace
        int traceDepth;

        /** Utility class which allows tracing to work even when exceptions are
         * thrown.
         */
        class Tracer { /*{{{*/
        private:
                Parser* parser;
                const char* text;
        public:
                Tracer(Parser* p,const char * t)
                : parser(p), text(t)
                {
                        parser->traceIn(text);
                }
                ~Tracer()
                {
#ifdef ANTLR_CXX_SUPPORTS_UNCAUGHT_EXCEPTION
                        // Only give trace if there's no uncaught exception..
                        if(!ANTLR_USE_NAMESPACE(std)uncaught_exception())
#endif
                                parser->traceOut(text);
                }
        private:
                Tracer(const Tracer&);                                                  // undefined
                const Tracer& operator=(const Tracer&); // undefined
                /*}}}*/
        };
private:
        Parser(const Parser&);                                                          // undefined
        const Parser& operator=(const Parser&);         // undefined
};

#ifdef ANTLR_CXX_SUPPORTS_NAMESPACE
}
#endif

#endif //INC_Parser_hpp__
Revision:	1558
Committed:	Wed May 11 16:32:48 2011 UTC (14 years, 5 months ago) by gezelter
File size:	8802 byte(s)
Log Message:	Updated antlr, some minor formatting changes
#	Content
1	#ifndef INC_Parser_hpp__
2	#define INC_Parser_hpp__
3
4	/* ANTLR Translator Generator
5	* Project led by Terence Parr at http://www.jGuru.com
6	* Software rights: http://www.antlr.org/license.html
7	*
8	* $Id$
9	*/
10
11	#include <antlr/config.hpp>
12
13	#include <iostream>
14	#include <exception>
15
16	#include <antlr/BitSet.hpp>
17	#include <antlr/TokenBuffer.hpp>
18	#include <antlr/RecognitionException.hpp>
19	#include <antlr/MismatchedTokenException.hpp>
20	#include <antlr/ASTFactory.hpp>
21	#include <antlr/ParserSharedInputState.hpp>
22
23	#ifdef ANTLR_CXX_SUPPORTS_NAMESPACE
24	namespace antlr {
25	#endif
26
27	extern bool DEBUG_PARSER;
28
29	/** A generic ANTLR parser (LL(k) for k>=1) containing a bunch of
30	* utility routines useful at any lookahead depth. We distinguish between
31	* the LL(1) and LL(k) parsers because of efficiency. This may not be
32	* necessary in the near future.
33	*
34	* Each parser object contains the state of the parse including a lookahead
35	* cache (the form of which is determined by the subclass), whether or
36	* not the parser is in guess mode, where tokens come from, etc...
37	*
38	* <p>
39	* During <b>guess</b> mode, the current lookahead token(s) and token type(s)
40	* cache must be saved because the token stream may not have been informed
41	* to save the token (via <tt>mark</tt>) before the <tt>try</tt> block.
42	* Guessing is started by:
43	* <ol>
44	* <li>saving the lookahead cache.
45	* <li>marking the current position in the TokenBuffer.
46	* <li>increasing the guessing level.
47	* </ol>
48	*
49	* After guessing, the parser state is restored by:
50	* <ol>
51	* <li>restoring the lookahead cache.
52	* <li>rewinding the TokenBuffer.
53	* <li>decreasing the guessing level.
54	* </ol>
55	*
56	* @see antlr.Token
57	* @see antlr.TokenBuffer
58	* @see antlr.TokenStream
59	* @see antlr.LL1Parser
60	* @see antlr.LLkParser
61	*
62	* @todo add constructors with ASTFactory.
63	*/
64	class ANTLR_API Parser {
65	protected:
66	Parser(TokenBuffer& input)
67	: inputState(new ParserInputState(input)), astFactory(0), traceDepth(0)
68	{
69	}
70	Parser(TokenBuffer* input)
71	: inputState(new ParserInputState(input)), astFactory(0), traceDepth(0)
72	{
73	}
74	Parser(const ParserSharedInputState& state)
75	: inputState(state), astFactory(0), traceDepth(0)
76	{
77	}
78	public:
79	virtual ~Parser()
80	{
81	}
82
83	/** Return the token type of the ith token of lookahead where i=1
84	* is the current token being examined by the parser (i.e., it
85	* has not been matched yet).
86	*/
87	virtual int LA(unsigned int i)=0;
88
89	/// Return the i-th token of lookahead
90	virtual RefToken LT(unsigned int i)=0;
91
92	/** DEPRECATED! Specify the factory to be used during tree building. (Compulsory)
93	* Setting the factory is nowadays compulsory.
94	* @see setASTFactory
95	*/
96	virtual void setASTNodeFactory( ASTFactory *factory )
97	{
98	astFactory = factory;
99	}
100	/** Specify the factory to be used during tree building. (Compulsory)
101	* Setting the factory is nowadays compulsory.
102	*/
103	virtual void setASTFactory( ASTFactory *factory )
104	{
105	astFactory = factory;
106	}
107	/** Return a pointer to the ASTFactory used.
108	* So you might use it in subsequent treewalkers or to reload AST's
109	* from disk.
110	*/
111	virtual ASTFactory* getASTFactory()
112	{
113	return astFactory;
114	}
115	/** Get the root AST node of the generated AST. When using a custom AST type
116	* or heterogenous AST's, you'll have to convert it to the right type
117	* yourself.
118	*/
119	virtual RefAST getAST() = 0;
120
121	/// Return the filename of the input file.
122	virtual inline ANTLR_USE_NAMESPACE(std)string getFilename() const
123	{
124	return inputState->filename;
125	}
126	/// Set the filename of the input file (used for error reporting).
127	virtual void setFilename(const ANTLR_USE_NAMESPACE(std)string& f)
128	{
129	inputState->filename = f;
130	}
131
132	virtual void setInputState(ParserSharedInputState state)
133	{
134	inputState = state;
135	}
136	virtual inline ParserSharedInputState getInputState() const
137	{
138	return inputState;
139	}
140
141	/// Get another token object from the token stream
142	virtual void consume()=0;
143	/// Consume tokens until one matches the given token
144	virtual void consumeUntil(int tokenType)
145	{
146	while (LA(1) != Token::EOF_TYPE && LA(1) != tokenType)
147	consume();
148	}
149
150	/// Consume tokens until one matches the given token set
151	virtual void consumeUntil(const BitSet& set)
152	{
153	while (LA(1) != Token::EOF_TYPE && !set.member(LA(1)))
154	consume();
155	}
156
157	/** Make sure current lookahead symbol matches token type <tt>t</tt>.
158	* Throw an exception upon mismatch, which is catch by either the
159	* error handler or by the syntactic predicate.
160	*/
161	virtual void match(int t)
162	{
163	if ( DEBUG_PARSER )
164	{
165	traceIndent();
166	ANTLR_USE_NAMESPACE(std)cout << "enter match(" << t << ") with LA(1)=" << LA(1) << ANTLR_USE_NAMESPACE(std)endl;
167	}
168	if ( LA(1) != t )
169	{
170	if ( DEBUG_PARSER )
171	{
172	traceIndent();
173	ANTLR_USE_NAMESPACE(std)cout << "token mismatch: " << LA(1) << "!=" << t << ANTLR_USE_NAMESPACE(std)endl;
174	}
175	throw MismatchedTokenException(getTokenNames(), getNumTokens(), LT(1), t, false, getFilename());
176	}
177	else
178	{
179	// mark token as consumed -- fetch next token deferred until LA/LT
180	consume();
181	}
182	}
183
184	virtual void matchNot(int t)
185	{
186	if ( LA(1)==t )
187	{
188	// Throws inverted-sense exception
189	throw MismatchedTokenException(getTokenNames(), getNumTokens(), LT(1), t, true, getFilename());
190	}
191	else
192	{
193	// mark token as consumed -- fetch next token deferred until LA/LT
194	consume();
195	}
196	}
197
198	/** Make sure current lookahead symbol matches the given set
199	* Throw an exception upon mismatch, which is catch by either the
200	* error handler or by the syntactic predicate.
201	*/
202	virtual void match(const BitSet& b)
203	{
204	if ( DEBUG_PARSER )
205	{
206	traceIndent();
207	ANTLR_USE_NAMESPACE(std)cout << "enter match(" << "bitset" /b.toString()/
208	<< ") with LA(1)=" << LA(1) << ANTLR_USE_NAMESPACE(std)endl;
209	}
210	if ( !b.member(LA(1)) )
211	{
212	if ( DEBUG_PARSER )
213	{
214	traceIndent();
215	ANTLR_USE_NAMESPACE(std)cout << "token mismatch: " << LA(1) << " not member of "
216	<< "bitset" /b.toString()/ << ANTLR_USE_NAMESPACE(std)endl;
217	}
218	throw MismatchedTokenException(getTokenNames(), getNumTokens(), LT(1), b, false, getFilename());
219	}
220	else
221	{
222	// mark token as consumed -- fetch next token deferred until LA/LT
223	consume();
224	}
225	}
226
227	/** Mark a spot in the input and return the position.
228	* Forwarded to TokenBuffer.
229	*/
230	virtual inline unsigned int mark()
231	{
232	return inputState->getInput().mark();
233	}
234	/// rewind to a previously marked position
235	virtual inline void rewind(unsigned int pos)
236	{
237	inputState->getInput().rewind(pos);
238	}
239	/** called by the generated parser to do error recovery, override to
240	* customize the behaviour.
241	*/
242	virtual void recover(const RecognitionException& ex, const BitSet& tokenSet)
243	{
244	consume();
245	consumeUntil(tokenSet);
246	}
247
248	/// Parser error-reporting function can be overridden in subclass
249	virtual void reportError(const RecognitionException& ex);
250	/// Parser error-reporting function can be overridden in subclass
251	virtual void reportError(const ANTLR_USE_NAMESPACE(std)string& s);
252	/// Parser warning-reporting function can be overridden in subclass
253	virtual void reportWarning(const ANTLR_USE_NAMESPACE(std)string& s);
254
255	/// get the token name for the token number 'num'
256	virtual const char* getTokenName(int num) const = 0;
257	/// get a vector with all token names
258	virtual const char* const* getTokenNames() const = 0;
259	/** Get the number of tokens defined.
260	* This one should be overridden in subclasses.
261	*/
262	virtual int getNumTokens(void) const = 0;
263
264	/** Set or change the input token buffer */
265	// void setTokenBuffer(TokenBuffer<Token>* t);
266
267	virtual void traceIndent();
268	virtual void traceIn(const char* rname);
269	virtual void traceOut(const char* rname);
270	protected:
271	// void setTokenNames(const char** tokenNames_);
272
273	ParserSharedInputState inputState;
274
275	// /// AST return value for a rule is squirreled away here
276	// RefAST returnAST;
277
278	/// AST support code; parser and treeparser delegate to this object
279	ASTFactory *astFactory;
280
281	// used to keep track of the indentation for the trace
282	int traceDepth;
283
284	/** Utility class which allows tracing to work even when exceptions are
285	* thrown.
286	*/
287	class Tracer { /{{{/
288	private:
289	Parser* parser;
290	const char* text;
291	public:
292	Tracer(Parser* p,const char * t)
293	: parser(p), text(t)
294	{
295	parser->traceIn(text);
296	}
297	~Tracer()
298	{
299	#ifdef ANTLR_CXX_SUPPORTS_UNCAUGHT_EXCEPTION
300	// Only give trace if there's no uncaught exception..
301	if(!ANTLR_USE_NAMESPACE(std)uncaught_exception())
302	#endif
303	parser->traceOut(text);
304	}
305	private:
306	Tracer(const Tracer&); // undefined
307	const Tracer& operator=(const Tracer&); // undefined
308	/}}}/
309	};
310	private:
311	Parser(const Parser&); // undefined
312	const Parser& operator=(const Parser&); // undefined
313	};
314
315	#ifdef ANTLR_CXX_SUPPORTS_NAMESPACE
316	}
317	#endif
318
319	#endif //INC_Parser_hpp__