xref: /linux/scripts/genksyms/lex.l (revision f3a8b6645dc2e60d11f20c1c23afd964ff4e55ae)
1 /* Lexical analysis for genksyms.
2    Copyright 1996, 1997 Linux International.
3 
4    New implementation contributed by Richard Henderson <rth@tamu.edu>
5    Based on original work by Bjorn Ekwall <bj0rn@blox.se>
6 
7    Taken from Linux modutils 2.4.22.
8 
9    This program is free software; you can redistribute it and/or modify it
10    under the terms of the GNU General Public License as published by the
11    Free Software Foundation; either version 2 of the License, or (at your
12    option) any later version.
13 
14    This program is distributed in the hope that it will be useful, but
15    WITHOUT ANY WARRANTY; without even the implied warranty of
16    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
17    General Public License for more details.
18 
19    You should have received a copy of the GNU General Public License
20    along with this program; if not, write to the Free Software Foundation,
21    Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
22 
23 
24 %{
25 
26 #include <limits.h>
27 #include <stdlib.h>
28 #include <string.h>
29 #include <ctype.h>
30 
31 #include "genksyms.h"
32 #include "parse.tab.h"
33 
34 /* We've got a two-level lexer here.  We let flex do basic tokenization
35    and then we categorize those basic tokens in the second stage.  */
36 #define YY_DECL		static int yylex1(void)
37 
38 %}
39 
40 IDENT			[A-Za-z_\$][A-Za-z0-9_\$]*
41 
42 O_INT			0[0-7]*
43 D_INT			[1-9][0-9]*
44 X_INT			0[Xx][0-9A-Fa-f]+
45 I_SUF			[Uu]|[Ll]|[Uu][Ll]|[Ll][Uu]
46 INT			({O_INT}|{D_INT}|{X_INT}){I_SUF}?
47 
48 FRAC			([0-9]*\.[0-9]+)|([0-9]+\.)
49 EXP			[Ee][+-]?[0-9]+
50 F_SUF			[FfLl]
51 REAL			({FRAC}{EXP}?{F_SUF}?)|([0-9]+{EXP}{F_SUF}?)
52 
53 STRING			L?\"([^\\\"]*\\.)*[^\\\"]*\"
54 CHAR			L?\'([^\\\']*\\.)*[^\\\']*\'
55 
56 MC_TOKEN		([~%^&*+=|<>/-]=)|(&&)|("||")|(->)|(<<)|(>>)
57 
58 /* We don't do multiple input files.  */
59 %option noyywrap
60 
61 %option noinput
62 
63 %%
64 
65 
66  /* Keep track of our location in the original source files.  */
67 ^#[ \t]+{INT}[ \t]+\"[^\"\n]+\".*\n	return FILENAME;
68 ^#.*\n					cur_line++;
69 \n					cur_line++;
70 
71  /* Ignore all other whitespace.  */
72 [ \t\f\v\r]+				;
73 
74 
75 {STRING}				return STRING;
76 {CHAR}					return CHAR;
77 {IDENT}					return IDENT;
78 
79  /* The Pedant requires that the other C multi-character tokens be
80     recognized as tokens.  We don't actually use them since we don't
81     parse expressions, but we do want whitespace to be arranged
82     around them properly.  */
83 {MC_TOKEN}				return OTHER;
84 {INT}					return INT;
85 {REAL}					return REAL;
86 
87 "..."					return DOTS;
88 
89  /* All other tokens are single characters.  */
90 .					return yytext[0];
91 
92 
93 %%
94 
95 /* Bring in the keyword recognizer.  */
96 
97 #include "keywords.hash.c"
98 
99 
100 /* Macros to append to our phrase collection list.  */
101 
102 /*
103  * We mark any token, that that equals to a known enumerator, as
104  * SYM_ENUM_CONST. The parser will change this for struct and union tags later,
105  * the only problem is struct and union members:
106  *    enum e { a, b }; struct s { int a, b; }
107  * but in this case, the only effect will be, that the ABI checksums become
108  * more volatile, which is acceptable. Also, such collisions are quite rare,
109  * so far it was only observed in include/linux/telephony.h.
110  */
111 #define _APP(T,L)	do {						   \
112 			  cur_node = next_node;				   \
113 			  next_node = xmalloc(sizeof(*next_node));	   \
114 			  next_node->next = cur_node;			   \
115 			  cur_node->string = memcpy(xmalloc(L+1), T, L+1); \
116 			  cur_node->tag =				   \
117 			    find_symbol(cur_node->string, SYM_ENUM_CONST, 1)?\
118 			    SYM_ENUM_CONST : SYM_NORMAL ;		   \
119 			  cur_node->in_source_file = in_source_file;       \
120 			} while (0)
121 
122 #define APP		_APP(yytext, yyleng)
123 
124 
125 /* The second stage lexer.  Here we incorporate knowledge of the state
126    of the parser to tailor the tokens that are returned.  */
127 
128 int
129 yylex(void)
130 {
131   static enum {
132     ST_NOTSTARTED, ST_NORMAL, ST_ATTRIBUTE, ST_ASM, ST_TYPEOF, ST_TYPEOF_1,
133     ST_BRACKET, ST_BRACE, ST_EXPRESSION,
134     ST_TABLE_1, ST_TABLE_2, ST_TABLE_3, ST_TABLE_4,
135     ST_TABLE_5, ST_TABLE_6
136   } lexstate = ST_NOTSTARTED;
137 
138   static int suppress_type_lookup, dont_want_brace_phrase;
139   static struct string_list *next_node;
140 
141   int token, count = 0;
142   struct string_list *cur_node;
143 
144   if (lexstate == ST_NOTSTARTED)
145     {
146       next_node = xmalloc(sizeof(*next_node));
147       next_node->next = NULL;
148       lexstate = ST_NORMAL;
149     }
150 
151 repeat:
152   token = yylex1();
153 
154   if (token == 0)
155     return 0;
156   else if (token == FILENAME)
157     {
158       char *file, *e;
159 
160       /* Save the filename and line number for later error messages.  */
161 
162       if (cur_filename)
163 	free(cur_filename);
164 
165       file = strchr(yytext, '\"')+1;
166       e = strchr(file, '\"');
167       *e = '\0';
168       cur_filename = memcpy(xmalloc(e-file+1), file, e-file+1);
169       cur_line = atoi(yytext+2);
170 
171       if (!source_file) {
172         source_file = xstrdup(cur_filename);
173         in_source_file = 1;
174       } else {
175         in_source_file = (strcmp(cur_filename, source_file) == 0);
176       }
177 
178       goto repeat;
179     }
180 
181   switch (lexstate)
182     {
183     case ST_NORMAL:
184       switch (token)
185 	{
186 	case IDENT:
187 	  APP;
188 	  {
189 	    const struct resword *r = is_reserved_word(yytext, yyleng);
190 	    if (r)
191 	      {
192 		switch (token = r->token)
193 		  {
194 		  case ATTRIBUTE_KEYW:
195 		    lexstate = ST_ATTRIBUTE;
196 		    count = 0;
197 		    goto repeat;
198 		  case ASM_KEYW:
199 		    lexstate = ST_ASM;
200 		    count = 0;
201 		    goto repeat;
202 		  case TYPEOF_KEYW:
203 		    lexstate = ST_TYPEOF;
204 		    count = 0;
205 		    goto repeat;
206 
207 		  case STRUCT_KEYW:
208 		  case UNION_KEYW:
209 		  case ENUM_KEYW:
210 		    dont_want_brace_phrase = 3;
211 		    suppress_type_lookup = 2;
212 		    goto fini;
213 
214 		  case EXPORT_SYMBOL_KEYW:
215 		      goto fini;
216 		  }
217 	      }
218 	    if (!suppress_type_lookup)
219 	      {
220 		if (find_symbol(yytext, SYM_TYPEDEF, 1))
221 		  token = TYPE;
222 	      }
223 	  }
224 	  break;
225 
226 	case '[':
227 	  APP;
228 	  lexstate = ST_BRACKET;
229 	  count = 1;
230 	  goto repeat;
231 
232 	case '{':
233 	  APP;
234 	  if (dont_want_brace_phrase)
235 	    break;
236 	  lexstate = ST_BRACE;
237 	  count = 1;
238 	  goto repeat;
239 
240 	case '=': case ':':
241 	  APP;
242 	  lexstate = ST_EXPRESSION;
243 	  break;
244 
245 	case DOTS:
246 	default:
247 	  APP;
248 	  break;
249 	}
250       break;
251 
252     case ST_ATTRIBUTE:
253       APP;
254       switch (token)
255 	{
256 	case '(':
257 	  ++count;
258 	  goto repeat;
259 	case ')':
260 	  if (--count == 0)
261 	    {
262 	      lexstate = ST_NORMAL;
263 	      token = ATTRIBUTE_PHRASE;
264 	      break;
265 	    }
266 	  goto repeat;
267 	default:
268 	  goto repeat;
269 	}
270       break;
271 
272     case ST_ASM:
273       APP;
274       switch (token)
275 	{
276 	case '(':
277 	  ++count;
278 	  goto repeat;
279 	case ')':
280 	  if (--count == 0)
281 	    {
282 	      lexstate = ST_NORMAL;
283 	      token = ASM_PHRASE;
284 	      break;
285 	    }
286 	  goto repeat;
287 	default:
288 	  goto repeat;
289 	}
290       break;
291 
292     case ST_TYPEOF_1:
293       if (token == IDENT)
294 	{
295 	  if (is_reserved_word(yytext, yyleng)
296 	      || find_symbol(yytext, SYM_TYPEDEF, 1))
297 	    {
298 	      yyless(0);
299 	      unput('(');
300 	      lexstate = ST_NORMAL;
301 	      token = TYPEOF_KEYW;
302 	      break;
303 	    }
304 	  _APP("(", 1);
305 	}
306 	lexstate = ST_TYPEOF;
307 	/* FALLTHRU */
308 
309     case ST_TYPEOF:
310       switch (token)
311 	{
312 	case '(':
313 	  if ( ++count == 1 )
314 	    lexstate = ST_TYPEOF_1;
315 	  else
316 	    APP;
317 	  goto repeat;
318 	case ')':
319 	  APP;
320 	  if (--count == 0)
321 	    {
322 	      lexstate = ST_NORMAL;
323 	      token = TYPEOF_PHRASE;
324 	      break;
325 	    }
326 	  goto repeat;
327 	default:
328 	  APP;
329 	  goto repeat;
330 	}
331       break;
332 
333     case ST_BRACKET:
334       APP;
335       switch (token)
336 	{
337 	case '[':
338 	  ++count;
339 	  goto repeat;
340 	case ']':
341 	  if (--count == 0)
342 	    {
343 	      lexstate = ST_NORMAL;
344 	      token = BRACKET_PHRASE;
345 	      break;
346 	    }
347 	  goto repeat;
348 	default:
349 	  goto repeat;
350 	}
351       break;
352 
353     case ST_BRACE:
354       APP;
355       switch (token)
356 	{
357 	case '{':
358 	  ++count;
359 	  goto repeat;
360 	case '}':
361 	  if (--count == 0)
362 	    {
363 	      lexstate = ST_NORMAL;
364 	      token = BRACE_PHRASE;
365 	      break;
366 	    }
367 	  goto repeat;
368 	default:
369 	  goto repeat;
370 	}
371       break;
372 
373     case ST_EXPRESSION:
374       switch (token)
375 	{
376 	case '(': case '[': case '{':
377 	  ++count;
378 	  APP;
379 	  goto repeat;
380 	case '}':
381 	  /* is this the last line of an enum declaration? */
382 	  if (count == 0)
383 	    {
384 	      /* Put back the token we just read so's we can find it again
385 		 after registering the expression.  */
386 	      unput(token);
387 
388 	      lexstate = ST_NORMAL;
389 	      token = EXPRESSION_PHRASE;
390 	      break;
391 	    }
392 	  /* FALLTHRU */
393 	case ')': case ']':
394 	  --count;
395 	  APP;
396 	  goto repeat;
397 	case ',': case ';':
398 	  if (count == 0)
399 	    {
400 	      /* Put back the token we just read so's we can find it again
401 		 after registering the expression.  */
402 	      unput(token);
403 
404 	      lexstate = ST_NORMAL;
405 	      token = EXPRESSION_PHRASE;
406 	      break;
407 	    }
408 	  APP;
409 	  goto repeat;
410 	default:
411 	  APP;
412 	  goto repeat;
413 	}
414       break;
415 
416     case ST_TABLE_1:
417       goto repeat;
418 
419     case ST_TABLE_2:
420       if (token == IDENT && yyleng == 1 && yytext[0] == 'X')
421 	{
422 	  token = EXPORT_SYMBOL_KEYW;
423 	  lexstate = ST_TABLE_5;
424 	  APP;
425 	  break;
426 	}
427       lexstate = ST_TABLE_6;
428       /* FALLTHRU */
429 
430     case ST_TABLE_6:
431       switch (token)
432 	{
433 	case '{': case '[': case '(':
434 	  ++count;
435 	  break;
436 	case '}': case ']': case ')':
437 	  --count;
438 	  break;
439 	case ',':
440 	  if (count == 0)
441 	    lexstate = ST_TABLE_2;
442 	  break;
443 	};
444       goto repeat;
445 
446     case ST_TABLE_3:
447       goto repeat;
448 
449     case ST_TABLE_4:
450       if (token == ';')
451 	lexstate = ST_NORMAL;
452       goto repeat;
453 
454     case ST_TABLE_5:
455       switch (token)
456 	{
457 	case ',':
458 	  token = ';';
459 	  lexstate = ST_TABLE_2;
460 	  APP;
461 	  break;
462 	default:
463 	  APP;
464 	  break;
465 	}
466       break;
467 
468     default:
469       exit(1);
470     }
471 fini:
472 
473   if (suppress_type_lookup > 0)
474     --suppress_type_lookup;
475   if (dont_want_brace_phrase > 0)
476     --dont_want_brace_phrase;
477 
478   yylval = &next_node->next;
479 
480   return token;
481 }
482