1 | #ifndef INC_CharScanner_hpp__ |
---|
2 | #define INC_CharScanner_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: //depot/code/org.antlr/release/antlr-2.7.7/lib/cpp/antlr/CharScanner.hpp#2 $ |
---|
9 | */ |
---|
10 | |
---|
11 | #include <cstdio> |
---|
12 | #include <antlr/config.hpp> |
---|
13 | |
---|
14 | #include <map> |
---|
15 | |
---|
16 | #ifdef HAS_NOT_CCTYPE_H |
---|
17 | #include <ctype.h> |
---|
18 | #else |
---|
19 | #include <cctype> |
---|
20 | #endif |
---|
21 | |
---|
22 | #if ( _MSC_VER == 1200 ) |
---|
23 | // VC6 seems to need this |
---|
24 | // note that this is not a standard C++ include file. |
---|
25 | # include <stdio.h> |
---|
26 | #endif |
---|
27 | |
---|
28 | #include <antlr/TokenStream.hpp> |
---|
29 | #include <antlr/RecognitionException.hpp> |
---|
30 | #include <antlr/SemanticException.hpp> |
---|
31 | #include <antlr/MismatchedCharException.hpp> |
---|
32 | #include <antlr/InputBuffer.hpp> |
---|
33 | #include <antlr/BitSet.hpp> |
---|
34 | #include <antlr/LexerSharedInputState.hpp> |
---|
35 | |
---|
36 | #ifdef ANTLR_CXX_SUPPORTS_NAMESPACE |
---|
37 | namespace antlr { |
---|
38 | #endif |
---|
39 | |
---|
40 | class ANTLR_API CharScanner; |
---|
41 | |
---|
42 | ANTLR_C_USING(tolower) |
---|
43 | |
---|
44 | // Modification done by hand because of compilation problems in the 2009 and 2010 version of Mandriva |
---|
45 | //#ifdef ANTLR_REALLY_NO_STRCASECMP |
---|
46 | |
---|
47 | |
---|
48 | // Apparently, neither strcasecmp nor stricmp is standard, and Codewarrior |
---|
49 | // on the mac has neither... |
---|
50 | inline int strcasecmp(const char *s1, const char *s2) |
---|
51 | { |
---|
52 | while (true) |
---|
53 | { |
---|
54 | char c1 = tolower(*s1++), |
---|
55 | c2 = tolower(*s2++); |
---|
56 | if (c1 < c2) return -1; |
---|
57 | if (c1 > c2) return 1; |
---|
58 | if (c1 == 0) return 0; |
---|
59 | } |
---|
60 | } |
---|
61 | // Modification done by hand because of compilation problems in the 2009 and 2010 version of Mandriva |
---|
62 | /*#else |
---|
63 | #ifdef NO_STRCASECMP |
---|
64 | ANTLR_C_USING(stricmp) |
---|
65 | #else |
---|
66 | ANTLR_C_USING(strcasecmp) |
---|
67 | #endif |
---|
68 | #endif |
---|
69 | */ |
---|
70 | /** Functor for the literals map |
---|
71 | */ |
---|
72 | class ANTLR_API CharScannerLiteralsLess : public ANTLR_USE_NAMESPACE(std)binary_function<ANTLR_USE_NAMESPACE(std)string,ANTLR_USE_NAMESPACE(std)string,bool> { |
---|
73 | private: |
---|
74 | const CharScanner* scanner; |
---|
75 | public: |
---|
76 | #ifdef NO_TEMPLATE_PARTS |
---|
77 | CharScannerLiteralsLess() {} // not really used, definition to appease MSVC |
---|
78 | #endif |
---|
79 | CharScannerLiteralsLess(const CharScanner* theScanner) |
---|
80 | : scanner(theScanner) |
---|
81 | { |
---|
82 | } |
---|
83 | bool operator() (const ANTLR_USE_NAMESPACE(std)string& x,const ANTLR_USE_NAMESPACE(std)string& y) const; |
---|
84 | // defaults are good enough.. |
---|
85 | // CharScannerLiteralsLess(const CharScannerLiteralsLess&); |
---|
86 | // CharScannerLiteralsLess& operator=(const CharScannerLiteralsLess&); |
---|
87 | }; |
---|
88 | |
---|
89 | /** Superclass of generated lexers |
---|
90 | */ |
---|
91 | class ANTLR_API CharScanner : public TokenStream { |
---|
92 | protected: |
---|
93 | typedef RefToken (*factory_type)(); |
---|
94 | public: |
---|
95 | CharScanner(InputBuffer& cb, bool case_sensitive ); |
---|
96 | CharScanner(InputBuffer* cb, bool case_sensitive ); |
---|
97 | CharScanner(const LexerSharedInputState& state, bool case_sensitive ); |
---|
98 | |
---|
99 | virtual ~CharScanner() |
---|
100 | { |
---|
101 | } |
---|
102 | |
---|
103 | virtual int LA(unsigned int i); |
---|
104 | |
---|
105 | virtual void append(char c) |
---|
106 | { |
---|
107 | if (saveConsumedInput) |
---|
108 | { |
---|
109 | size_t l = text.length(); |
---|
110 | |
---|
111 | if ((l%256) == 0) |
---|
112 | text.reserve(l+256); |
---|
113 | |
---|
114 | text.replace(l,0,&c,1); |
---|
115 | } |
---|
116 | } |
---|
117 | |
---|
118 | virtual void append(const ANTLR_USE_NAMESPACE(std)string& s) |
---|
119 | { |
---|
120 | if( saveConsumedInput ) |
---|
121 | text += s; |
---|
122 | } |
---|
123 | |
---|
124 | virtual void commit() |
---|
125 | { |
---|
126 | inputState->getInput().commit(); |
---|
127 | } |
---|
128 | |
---|
129 | /** called by the generated lexer to do error recovery, override to |
---|
130 | * customize the behaviour. |
---|
131 | */ |
---|
132 | virtual void recover(const RecognitionException& ex, const BitSet& tokenSet) |
---|
133 | { |
---|
134 | consume(); |
---|
135 | consumeUntil(tokenSet); |
---|
136 | } |
---|
137 | |
---|
138 | virtual void consume() |
---|
139 | { |
---|
140 | if (inputState->guessing == 0) |
---|
141 | { |
---|
142 | int c = LA(1); |
---|
143 | if (caseSensitive) |
---|
144 | { |
---|
145 | append(c); |
---|
146 | } |
---|
147 | else |
---|
148 | { |
---|
149 | // use input.LA(), not LA(), to get original case |
---|
150 | // CharScanner.LA() would toLower it. |
---|
151 | append(inputState->getInput().LA(1)); |
---|
152 | } |
---|
153 | |
---|
154 | // RK: in a sense I don't like this automatic handling. |
---|
155 | if (c == '\t') |
---|
156 | tab(); |
---|
157 | else |
---|
158 | inputState->column++; |
---|
159 | } |
---|
160 | inputState->getInput().consume(); |
---|
161 | } |
---|
162 | |
---|
163 | /** Consume chars until one matches the given char */ |
---|
164 | virtual void consumeUntil(int c) |
---|
165 | { |
---|
166 | for(;;) |
---|
167 | { |
---|
168 | int la_1 = LA(1); |
---|
169 | if( la_1 == EOF_CHAR || la_1 == c ) |
---|
170 | break; |
---|
171 | consume(); |
---|
172 | } |
---|
173 | } |
---|
174 | |
---|
175 | /** Consume chars until one matches the given set */ |
---|
176 | virtual void consumeUntil(const BitSet& set) |
---|
177 | { |
---|
178 | for(;;) |
---|
179 | { |
---|
180 | int la_1 = LA(1); |
---|
181 | if( la_1 == EOF_CHAR || set.member(la_1) ) |
---|
182 | break; |
---|
183 | consume(); |
---|
184 | } |
---|
185 | } |
---|
186 | |
---|
187 | /// Mark the current position and return a id for it |
---|
188 | virtual unsigned int mark() |
---|
189 | { |
---|
190 | return inputState->getInput().mark(); |
---|
191 | } |
---|
192 | /// Rewind the scanner to a previously marked position |
---|
193 | virtual void rewind(unsigned int pos) |
---|
194 | { |
---|
195 | inputState->getInput().rewind(pos); |
---|
196 | } |
---|
197 | |
---|
198 | /// See if input contains character 'c' throw MismatchedCharException if not |
---|
199 | virtual void match(int c) |
---|
200 | { |
---|
201 | int la_1 = LA(1); |
---|
202 | if ( la_1 != c ) |
---|
203 | throw MismatchedCharException(la_1, c, false, this); |
---|
204 | consume(); |
---|
205 | } |
---|
206 | |
---|
207 | /** See if input contains element from bitset b |
---|
208 | * throw MismatchedCharException if not |
---|
209 | */ |
---|
210 | virtual void match(const BitSet& b) |
---|
211 | { |
---|
212 | int la_1 = LA(1); |
---|
213 | |
---|
214 | if ( !b.member(la_1) ) |
---|
215 | throw MismatchedCharException( la_1, b, false, this ); |
---|
216 | consume(); |
---|
217 | } |
---|
218 | |
---|
219 | /** See if input contains string 's' throw MismatchedCharException if not |
---|
220 | * @note the string cannot match EOF |
---|
221 | */ |
---|
222 | virtual void match( const char* s ) |
---|
223 | { |
---|
224 | while( *s != '\0' ) |
---|
225 | { |
---|
226 | // the & 0xFF is here to prevent sign extension lateron |
---|
227 | int la_1 = LA(1), c = (*s++ & 0xFF); |
---|
228 | |
---|
229 | if ( la_1 != c ) |
---|
230 | throw MismatchedCharException(la_1, c, false, this); |
---|
231 | |
---|
232 | consume(); |
---|
233 | } |
---|
234 | } |
---|
235 | /** See if input contains string 's' throw MismatchedCharException if not |
---|
236 | * @note the string cannot match EOF |
---|
237 | */ |
---|
238 | virtual void match(const ANTLR_USE_NAMESPACE(std)string& s) |
---|
239 | { |
---|
240 | size_t len = s.length(); |
---|
241 | |
---|
242 | for (size_t i = 0; i < len; i++) |
---|
243 | { |
---|
244 | // the & 0xFF is here to prevent sign extension lateron |
---|
245 | int la_1 = LA(1), c = (s[i] & 0xFF); |
---|
246 | |
---|
247 | if ( la_1 != c ) |
---|
248 | throw MismatchedCharException(la_1, c, false, this); |
---|
249 | |
---|
250 | consume(); |
---|
251 | } |
---|
252 | } |
---|
253 | /** See if input does not contain character 'c' |
---|
254 | * throw MismatchedCharException if not |
---|
255 | */ |
---|
256 | virtual void matchNot(int c) |
---|
257 | { |
---|
258 | int la_1 = LA(1); |
---|
259 | |
---|
260 | if ( la_1 == c ) |
---|
261 | throw MismatchedCharException(la_1, c, true, this); |
---|
262 | |
---|
263 | consume(); |
---|
264 | } |
---|
265 | /** See if input contains character in range c1-c2 |
---|
266 | * throw MismatchedCharException if not |
---|
267 | */ |
---|
268 | virtual void matchRange(int c1, int c2) |
---|
269 | { |
---|
270 | int la_1 = LA(1); |
---|
271 | |
---|
272 | if ( la_1 < c1 || la_1 > c2 ) |
---|
273 | throw MismatchedCharException(la_1, c1, c2, false, this); |
---|
274 | |
---|
275 | consume(); |
---|
276 | } |
---|
277 | |
---|
278 | virtual bool getCaseSensitive() const |
---|
279 | { |
---|
280 | return caseSensitive; |
---|
281 | } |
---|
282 | |
---|
283 | virtual void setCaseSensitive(bool t) |
---|
284 | { |
---|
285 | caseSensitive = t; |
---|
286 | } |
---|
287 | |
---|
288 | virtual bool getCaseSensitiveLiterals() const=0; |
---|
289 | |
---|
290 | /// Get the line the scanner currently is in (starts at 1) |
---|
291 | virtual int getLine() const |
---|
292 | { |
---|
293 | return inputState->line; |
---|
294 | } |
---|
295 | |
---|
296 | /// set the line number |
---|
297 | virtual void setLine(int l) |
---|
298 | { |
---|
299 | inputState->line = l; |
---|
300 | } |
---|
301 | |
---|
302 | /// Get the column the scanner currently is in (starts at 1) |
---|
303 | virtual int getColumn() const |
---|
304 | { |
---|
305 | return inputState->column; |
---|
306 | } |
---|
307 | /// set the column number |
---|
308 | virtual void setColumn(int c) |
---|
309 | { |
---|
310 | inputState->column = c; |
---|
311 | } |
---|
312 | |
---|
313 | /// get the filename for the file currently used |
---|
314 | virtual const ANTLR_USE_NAMESPACE(std)string& getFilename() const |
---|
315 | { |
---|
316 | return inputState->filename; |
---|
317 | } |
---|
318 | /// Set the filename the scanner is using (used in error messages) |
---|
319 | virtual void setFilename(const ANTLR_USE_NAMESPACE(std)string& f) |
---|
320 | { |
---|
321 | inputState->filename = f; |
---|
322 | } |
---|
323 | |
---|
324 | virtual bool getCommitToPath() const |
---|
325 | { |
---|
326 | return commitToPath; |
---|
327 | } |
---|
328 | |
---|
329 | virtual void setCommitToPath(bool commit) |
---|
330 | { |
---|
331 | commitToPath = commit; |
---|
332 | } |
---|
333 | |
---|
334 | /** return a copy of the current text buffer */ |
---|
335 | virtual const ANTLR_USE_NAMESPACE(std)string& getText() const |
---|
336 | { |
---|
337 | return text; |
---|
338 | } |
---|
339 | |
---|
340 | virtual void setText(const ANTLR_USE_NAMESPACE(std)string& s) |
---|
341 | { |
---|
342 | text = s; |
---|
343 | } |
---|
344 | |
---|
345 | virtual void resetText() |
---|
346 | { |
---|
347 | text = ""; |
---|
348 | inputState->tokenStartColumn = inputState->column; |
---|
349 | inputState->tokenStartLine = inputState->line; |
---|
350 | } |
---|
351 | |
---|
352 | virtual RefToken getTokenObject() const |
---|
353 | { |
---|
354 | return _returnToken; |
---|
355 | } |
---|
356 | |
---|
357 | /** Used to keep track of line breaks, needs to be called from |
---|
358 | * within generated lexers when a \n \r is encountered. |
---|
359 | */ |
---|
360 | virtual void newline() |
---|
361 | { |
---|
362 | ++inputState->line; |
---|
363 | inputState->column = 1; |
---|
364 | } |
---|
365 | |
---|
366 | /** Advance the current column number by an appropriate amount according |
---|
367 | * to the tabsize. This method needs to be explicitly called from the |
---|
368 | * lexer rules encountering tabs. |
---|
369 | */ |
---|
370 | virtual void tab() |
---|
371 | { |
---|
372 | int c = getColumn(); |
---|
373 | int nc = ( ((c-1)/tabsize) + 1) * tabsize + 1; // calculate tab stop |
---|
374 | setColumn( nc ); |
---|
375 | } |
---|
376 | /// set the tabsize. Returns the old tabsize |
---|
377 | int setTabsize( int size ) |
---|
378 | { |
---|
379 | int oldsize = tabsize; |
---|
380 | tabsize = size; |
---|
381 | return oldsize; |
---|
382 | } |
---|
383 | /// Return the tabsize used by the scanner |
---|
384 | int getTabSize() const |
---|
385 | { |
---|
386 | return tabsize; |
---|
387 | } |
---|
388 | |
---|
389 | /** Report exception errors caught in nextToken() */ |
---|
390 | virtual void reportError(const RecognitionException& e); |
---|
391 | |
---|
392 | /** Parser error-reporting function can be overridden in subclass */ |
---|
393 | virtual void reportError(const ANTLR_USE_NAMESPACE(std)string& s); |
---|
394 | |
---|
395 | /** Parser warning-reporting function can be overridden in subclass */ |
---|
396 | virtual void reportWarning(const ANTLR_USE_NAMESPACE(std)string& s); |
---|
397 | |
---|
398 | virtual InputBuffer& getInputBuffer() |
---|
399 | { |
---|
400 | return inputState->getInput(); |
---|
401 | } |
---|
402 | |
---|
403 | virtual LexerSharedInputState getInputState() |
---|
404 | { |
---|
405 | return inputState; |
---|
406 | } |
---|
407 | |
---|
408 | /** set the input state for the lexer. |
---|
409 | * @note state is a reference counted object, hence no reference */ |
---|
410 | virtual void setInputState(LexerSharedInputState state) |
---|
411 | { |
---|
412 | inputState = state; |
---|
413 | } |
---|
414 | |
---|
415 | /// Set the factory for created tokens |
---|
416 | virtual void setTokenObjectFactory(factory_type factory) |
---|
417 | { |
---|
418 | tokenFactory = factory; |
---|
419 | } |
---|
420 | |
---|
421 | /** Test the token text against the literals table |
---|
422 | * Override this method to perform a different literals test |
---|
423 | */ |
---|
424 | virtual int testLiteralsTable(int ttype) const |
---|
425 | { |
---|
426 | ANTLR_USE_NAMESPACE(std)map<ANTLR_USE_NAMESPACE(std)string,int,CharScannerLiteralsLess>::const_iterator i = literals.find(text); |
---|
427 | if (i != literals.end()) |
---|
428 | ttype = (*i).second; |
---|
429 | return ttype; |
---|
430 | } |
---|
431 | |
---|
432 | /** Test the text passed in against the literals table |
---|
433 | * Override this method to perform a different literals test |
---|
434 | * This is used primarily when you want to test a portion of |
---|
435 | * a token |
---|
436 | */ |
---|
437 | virtual int testLiteralsTable(const ANTLR_USE_NAMESPACE(std)string& txt,int ttype) const |
---|
438 | { |
---|
439 | ANTLR_USE_NAMESPACE(std)map<ANTLR_USE_NAMESPACE(std)string,int,CharScannerLiteralsLess>::const_iterator i = literals.find(txt); |
---|
440 | if (i != literals.end()) |
---|
441 | ttype = (*i).second; |
---|
442 | return ttype; |
---|
443 | } |
---|
444 | |
---|
445 | /// Override this method to get more specific case handling |
---|
446 | virtual int toLower(int c) const |
---|
447 | { |
---|
448 | // test on EOF_CHAR for buggy (?) STLPort tolower (or HPUX tolower?) |
---|
449 | // also VC++ 6.0 does this. (see fix 422 (is reverted by this fix) |
---|
450 | // this one is more structural. Maybe make this configurable. |
---|
451 | return (c == EOF_CHAR ? EOF_CHAR : tolower(c)); |
---|
452 | } |
---|
453 | |
---|
454 | /** This method is called by YourLexer::nextToken() when the lexer has |
---|
455 | * hit EOF condition. EOF is NOT a character. |
---|
456 | * This method is not called if EOF is reached during |
---|
457 | * syntactic predicate evaluation or during evaluation |
---|
458 | * of normal lexical rules, which presumably would be |
---|
459 | * an IOException. This traps the "normal" EOF condition. |
---|
460 | * |
---|
461 | * uponEOF() is called after the complete evaluation of |
---|
462 | * the previous token and only if your parser asks |
---|
463 | * for another token beyond that last non-EOF token. |
---|
464 | * |
---|
465 | * You might want to throw token or char stream exceptions |
---|
466 | * like: "Heh, premature eof" or a retry stream exception |
---|
467 | * ("I found the end of this file, go back to referencing file"). |
---|
468 | */ |
---|
469 | virtual void uponEOF() |
---|
470 | { |
---|
471 | } |
---|
472 | |
---|
473 | /// Methods used to change tracing behavior |
---|
474 | virtual void traceIndent(); |
---|
475 | virtual void traceIn(const char* rname); |
---|
476 | virtual void traceOut(const char* rname); |
---|
477 | |
---|
478 | #ifndef NO_STATIC_CONSTS |
---|
479 | static const int EOF_CHAR = EOF; |
---|
480 | #else |
---|
481 | enum { |
---|
482 | EOF_CHAR = EOF |
---|
483 | }; |
---|
484 | #endif |
---|
485 | protected: |
---|
486 | ANTLR_USE_NAMESPACE(std)string text; ///< Text of current token |
---|
487 | /// flag indicating wether consume saves characters |
---|
488 | bool saveConsumedInput; |
---|
489 | factory_type tokenFactory; ///< Factory for tokens |
---|
490 | bool caseSensitive; ///< Is this lexer case sensitive |
---|
491 | ANTLR_USE_NAMESPACE(std)map<ANTLR_USE_NAMESPACE(std)string,int,CharScannerLiteralsLess> literals; // set by subclass |
---|
492 | |
---|
493 | RefToken _returnToken; ///< used to return tokens w/o using return val |
---|
494 | |
---|
495 | /// Input state, gives access to input stream, shared among different lexers |
---|
496 | LexerSharedInputState inputState; |
---|
497 | |
---|
498 | /** Used during filter mode to indicate that path is desired. |
---|
499 | * A subsequent scan error will report an error as usual |
---|
500 | * if acceptPath=true; |
---|
501 | */ |
---|
502 | bool commitToPath; |
---|
503 | |
---|
504 | int tabsize; ///< tab size the scanner uses. |
---|
505 | |
---|
506 | /// Create a new RefToken of type t |
---|
507 | virtual RefToken makeToken(int t) |
---|
508 | { |
---|
509 | RefToken tok = tokenFactory(); |
---|
510 | tok->setType(t); |
---|
511 | tok->setColumn(inputState->tokenStartColumn); |
---|
512 | tok->setLine(inputState->tokenStartLine); |
---|
513 | return tok; |
---|
514 | } |
---|
515 | |
---|
516 | /** Tracer class, used when -traceLexer is passed to antlr |
---|
517 | */ |
---|
518 | class Tracer { |
---|
519 | private: |
---|
520 | CharScanner* parser; |
---|
521 | const char* text; |
---|
522 | |
---|
523 | Tracer(const Tracer& other); // undefined |
---|
524 | Tracer& operator=(const Tracer& other); // undefined |
---|
525 | public: |
---|
526 | Tracer( CharScanner* p,const char* t ) |
---|
527 | : parser(p), text(t) |
---|
528 | { |
---|
529 | parser->traceIn(text); |
---|
530 | } |
---|
531 | ~Tracer() |
---|
532 | { |
---|
533 | parser->traceOut(text); |
---|
534 | } |
---|
535 | }; |
---|
536 | |
---|
537 | int traceDepth; |
---|
538 | private: |
---|
539 | CharScanner( const CharScanner& other ); // undefined |
---|
540 | CharScanner& operator=( const CharScanner& other ); // undefined |
---|
541 | |
---|
542 | #ifndef NO_STATIC_CONSTS |
---|
543 | static const int NO_CHAR = 0; |
---|
544 | #else |
---|
545 | enum { |
---|
546 | NO_CHAR = 0 |
---|
547 | }; |
---|
548 | #endif |
---|
549 | }; |
---|
550 | |
---|
551 | inline int CharScanner::LA(unsigned int i) |
---|
552 | { |
---|
553 | int c = inputState->getInput().LA(i); |
---|
554 | |
---|
555 | if ( caseSensitive ) |
---|
556 | return c; |
---|
557 | else |
---|
558 | return toLower(c); // VC 6 tolower bug caught in toLower. |
---|
559 | } |
---|
560 | |
---|
561 | inline bool CharScannerLiteralsLess::operator() (const ANTLR_USE_NAMESPACE(std)string& x,const ANTLR_USE_NAMESPACE(std)string& y) const |
---|
562 | { |
---|
563 | if (scanner->getCaseSensitiveLiterals()) |
---|
564 | return ANTLR_USE_NAMESPACE(std)less<ANTLR_USE_NAMESPACE(std)string>()(x,y); |
---|
565 | else |
---|
566 | { |
---|
567 | #ifdef NO_STRCASECMP |
---|
568 | return (stricmp(x.c_str(),y.c_str())<0); |
---|
569 | #else |
---|
570 | return (strcasecmp(x.c_str(),y.c_str())<0); |
---|
571 | #endif |
---|
572 | } |
---|
573 | } |
---|
574 | |
---|
575 | #ifdef ANTLR_CXX_SUPPORTS_NAMESPACE |
---|
576 | } |
---|
577 | #endif |
---|
578 | |
---|
579 | #endif //INC_CharScanner_hpp__ |
---|