1 /*
2 * Copyright (c) 1987, 1993, 1994
3 * The Regents of the University of California. All rights reserved.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
7 * are met:
8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 * notice, this list of conditions and the following disclaimer in the
12 * documentation and/or other materials provided with the distribution.
13 * 3. Neither the name of the University nor the names of its contributors
14 * may be used to endorse or promote products derived from this software
15 * without specific prior written permission.
16 *
17 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27 * SUCH DAMAGE.
28 *
29 * @(#)C.c 8.4 (Berkeley) 4/2/94
30 * $FreeBSD: head/usr.bin/ctags/C.c 299355 2016-05-10 11:11:23Z bapt $
31 */
32
33 #include <limits.h>
34 #include <stdio.h>
35 #include <string.h>
36
37 #include "ctags.h"
38
39 static int func_entry(void);
40 static void hash_entry(void);
41 static void skip_string(int);
42 static int str_entry(int);
43
44 /*
45 * c_entries --
46 * read .c and .h files and call appropriate routines
47 */
48 void
c_entries(void)49 c_entries(void)
50 {
51 int c; /* current character */
52 int level; /* brace level */
53 int token; /* if reading a token */
54 int t_def; /* if reading a typedef */
55 int t_level; /* typedef's brace level */
56 char *sp; /* buffer pointer */
57 char tok[MAXTOKEN]; /* token buffer */
58
59 lineftell = ftell(inf);
60 sp = tok; token = t_def = NO; t_level = -1; level = 0; lineno = 1;
61 while (GETC(!=, EOF)) {
62 switch (c) {
63 /*
64 * Here's where it DOESN'T handle: {
65 * foo(a)
66 * {
67 * #ifdef notdef
68 * }
69 * #endif
70 * if (a)
71 * puts("hello, world");
72 * }
73 */
74 case '{':
75 ++level;
76 goto endtok;
77 case '}':
78 /*
79 * if level goes below zero, try and fix
80 * it, even though we've already messed up
81 */
82 if (--level < 0)
83 level = 0;
84 goto endtok;
85
86 case '\n':
87 SETLINE;
88 /*
89 * the above 3 cases are similar in that they
90 * are special characters that also end tokens.
91 */
92 endtok: if (sp > tok) {
93 *sp = EOS;
94 token = YES;
95 sp = tok;
96 }
97 else
98 token = NO;
99 continue;
100
101 /*
102 * We ignore quoted strings and character constants
103 * completely.
104 */
105 case '"':
106 case '\'':
107 skip_string(c);
108 break;
109
110 /*
111 * comments can be fun; note the state is unchanged after
112 * return, in case we found:
113 * "foo() XX comment XX { int bar; }"
114 */
115 case '/':
116 if (GETC(==, '*') || c == '/') {
117 skip_comment(c);
118 continue;
119 }
120 ungetc(c, inf);
121 c = '/';
122 goto storec;
123
124 /* hash marks flag #define's. */
125 case '#':
126 if (sp == tok) {
127 hash_entry();
128 break;
129 }
130 goto storec;
131
132 /*
133 * if we have a current token, parenthesis on
134 * level zero indicates a function.
135 */
136 case '(':
137 if (!level && token) {
138 int curline;
139
140 if (sp != tok)
141 *sp = EOS;
142 /*
143 * grab the line immediately, we may
144 * already be wrong, for example,
145 * foo\n
146 * (arg1,
147 */
148 get_line();
149 curline = lineno;
150 if (func_entry()) {
151 ++level;
152 pfnote(tok, curline);
153 }
154 break;
155 }
156 goto storec;
157
158 /*
159 * semi-colons indicate the end of a typedef; if we find a
160 * typedef we search for the next semi-colon of the same
161 * level as the typedef. Ignoring "structs", they are
162 * tricky, since you can find:
163 *
164 * "typedef long time_t;"
165 * "typedef unsigned int u_int;"
166 * "typedef unsigned int u_int [10];"
167 *
168 * If looking at a typedef, we save a copy of the last token
169 * found. Then, when we find the ';' we take the current
170 * token if it starts with a valid token name, else we take
171 * the one we saved. There's probably some reasonable
172 * alternative to this...
173 */
174 case ';':
175 if (t_def && level == t_level) {
176 t_def = NO;
177 get_line();
178 if (sp != tok)
179 *sp = EOS;
180 pfnote(tok, lineno);
181 break;
182 }
183 goto storec;
184
185 /*
186 * store characters until one that can't be part of a token
187 * comes along; check the current token against certain
188 * reserved words.
189 */
190 default:
191 /* ignore whitespace */
192 if (c == ' ' || c == '\t') {
193 int save = c;
194 while (GETC(!=, EOF) && (c == ' ' || c == '\t'))
195 ;
196 if (c == EOF)
197 return;
198 ungetc(c, inf);
199 c = save;
200 }
201 storec: if (!intoken(c)) {
202 if (sp == tok)
203 break;
204 *sp = EOS;
205 if (tflag) {
206 /* no typedefs inside typedefs */
207 if (!t_def &&
208 !memcmp(tok, "typedef",8)) {
209 t_def = YES;
210 t_level = level;
211 break;
212 }
213 /* catch "typedef struct" */
214 if ((!t_def || t_level < level)
215 && (!memcmp(tok, "struct", 7)
216 || !memcmp(tok, "union", 6)
217 || !memcmp(tok, "enum", 5))) {
218 /*
219 * get line immediately;
220 * may change before '{'
221 */
222 get_line();
223 if (str_entry(c))
224 ++level;
225 break;
226 /* } */
227 }
228 }
229 sp = tok;
230 }
231 else if (sp != tok || begtoken(c)) {
232 if (sp == tok + sizeof tok - 1)
233 /* Too long -- truncate it */
234 *sp = EOS;
235 else
236 *sp++ = c;
237 token = YES;
238 }
239 continue;
240 }
241
242 sp = tok;
243 token = NO;
244 }
245 }
246
247 /*
248 * func_entry --
249 * handle a function reference
250 */
251 static int
func_entry(void)252 func_entry(void)
253 {
254 int c; /* current character */
255 int level = 0; /* for matching '()' */
256
257 /*
258 * Find the end of the assumed function declaration.
259 * Note that ANSI C functions can have type definitions so keep
260 * track of the parentheses nesting level.
261 */
262 while (GETC(!=, EOF)) {
263 switch (c) {
264 case '\'':
265 case '"':
266 /* skip strings and character constants */
267 skip_string(c);
268 break;
269 case '/':
270 /* skip comments */
271 if (GETC(==, '*') || c == '/')
272 skip_comment(c);
273 break;
274 case '(':
275 level++;
276 break;
277 case ')':
278 if (level == 0)
279 goto fnd;
280 level--;
281 break;
282 case '\n':
283 SETLINE;
284 }
285 }
286 return (NO);
287 fnd:
288 /*
289 * we assume that the character after a function's right paren
290 * is a token character if it's a function and a non-token
291 * character if it's a declaration. Comments don't count...
292 */
293 for (;;) {
294 while (GETC(!=, EOF) && iswhite(c))
295 if (c == '\n')
296 SETLINE;
297 if (intoken(c) || c == '{')
298 break;
299 if (c == '/' && (GETC(==, '*') || c == '/'))
300 skip_comment(c);
301 else { /* don't ever "read" '/' */
302 ungetc(c, inf);
303 return (NO);
304 }
305 }
306 if (c != '{')
307 skip_key('{');
308 return (YES);
309 }
310
311 /*
312 * hash_entry --
313 * handle a line starting with a '#'
314 */
315 static void
hash_entry(void)316 hash_entry(void)
317 {
318 int c; /* character read */
319 int curline; /* line started on */
320 char *sp; /* buffer pointer */
321 char tok[MAXTOKEN]; /* storage buffer */
322
323 /* ignore leading whitespace */
324 while (GETC(!=, EOF) && (c == ' ' || c == '\t'))
325 ;
326 ungetc(c, inf);
327
328 curline = lineno;
329 for (sp = tok;;) { /* get next token */
330 if (GETC(==, EOF))
331 return;
332 if (iswhite(c))
333 break;
334 if (sp == tok + sizeof tok - 1)
335 /* Too long -- truncate it */
336 *sp = EOS;
337 else
338 *sp++ = c;
339 }
340 *sp = EOS;
341 if (memcmp(tok, "define", 6)) /* only interested in #define's */
342 goto skip;
343 for (;;) { /* this doesn't handle "#define \n" */
344 if (GETC(==, EOF))
345 return;
346 if (!iswhite(c))
347 break;
348 }
349 for (sp = tok;;) { /* get next token */
350 if (sp == tok + sizeof tok - 1)
351 /* Too long -- truncate it */
352 *sp = EOS;
353 else
354 *sp++ = c;
355 if (GETC(==, EOF))
356 return;
357 /*
358 * this is where it DOESN'T handle
359 * "#define \n"
360 */
361 if (!intoken(c))
362 break;
363 }
364 *sp = EOS;
365 if (dflag || c == '(') { /* only want macros */
366 get_line();
367 pfnote(tok, curline);
368 }
369 skip: if (c == '\n') { /* get rid of rest of define */
370 SETLINE
371 if (*(sp - 1) != '\\')
372 return;
373 }
374 skip_key('\n');
375 }
376
377 /*
378 * str_entry --
379 * handle a struct, union or enum entry
380 */
381 static int
str_entry(int c)382 str_entry(int c) /* c is current character */
383 {
384 int curline; /* line started on */
385 char *sp; /* buffer pointer */
386 char tok[LINE_MAX]; /* storage buffer */
387
388 curline = lineno;
389 while (iswhite(c))
390 if (GETC(==, EOF))
391 return (NO);
392 if (c == '{') /* it was "struct {" */
393 return (YES);
394 for (sp = tok;;) { /* get next token */
395 if (sp == tok + sizeof tok - 1)
396 /* Too long -- truncate it */
397 *sp = EOS;
398 else
399 *sp++ = c;
400 if (GETC(==, EOF))
401 return (NO);
402 if (!intoken(c))
403 break;
404 }
405 switch (c) {
406 case '{': /* it was "struct foo{" */
407 --sp;
408 break;
409 case '\n': /* it was "struct foo\n" */
410 SETLINE;
411 /*FALLTHROUGH*/
412 default: /* probably "struct foo " */
413 while (GETC(!=, EOF))
414 if (!iswhite(c))
415 break;
416 if (c != '{') {
417 ungetc(c, inf);
418 return (NO);
419 }
420 }
421 *sp = EOS;
422 pfnote(tok, curline);
423 return (YES);
424 }
425
426 /*
427 * skip_comment --
428 * skip over comment
429 */
430 void
skip_comment(int t)431 skip_comment(int t) /* t is comment character */
432 {
433 int c; /* character read */
434 int star; /* '*' flag */
435
436 for (star = 0; GETC(!=, EOF);)
437 switch(c) {
438 /* comments don't nest, nor can they be escaped. */
439 case '*':
440 star = YES;
441 break;
442 case '/':
443 if (star && t == '*')
444 return;
445 break;
446 case '\n':
447 if (t == '/')
448 return;
449 SETLINE;
450 /*FALLTHROUGH*/
451 default:
452 star = NO;
453 break;
454 }
455 }
456
457 /*
458 * skip_string --
459 * skip to the end of a string or character constant.
460 */
461 void
skip_string(int key)462 skip_string(int key)
463 {
464 int c,
465 skip;
466
467 for (skip = NO; GETC(!=, EOF); )
468 switch (c) {
469 case '\\': /* a backslash escapes anything */
470 skip = !skip; /* we toggle in case it's "\\" */
471 break;
472 case '\n':
473 SETLINE;
474 /*FALLTHROUGH*/
475 default:
476 if (c == key && !skip)
477 return;
478 skip = NO;
479 }
480 }
481
482 /*
483 * skip_key --
484 * skip to next char "key"
485 */
486 int
skip_key(int key)487 skip_key(int key)
488 {
489 int c,
490 skip,
491 retval;
492
493 for (skip = retval = NO; GETC(!=, EOF);)
494 switch(c) {
495 case '\\': /* a backslash escapes anything */
496 skip = !skip; /* we toggle in case it's "\\" */
497 break;
498 case ';': /* special case for yacc; if one */
499 case '|': /* of these chars occurs, we may */
500 retval = YES; /* have moved out of the rule */
501 break; /* not used by C */
502 case '\'':
503 case '"':
504 /* skip strings and character constants */
505 skip_string(c);
506 break;
507 case '/':
508 /* skip comments */
509 if (GETC(==, '*') || c == '/') {
510 skip_comment(c);
511 break;
512 }
513 ungetc(c, inf);
514 c = '/';
515 goto norm;
516 case '\n':
517 SETLINE;
518 /*FALLTHROUGH*/
519 default:
520 norm:
521 if (c == key && !skip)
522 return (retval);
523 skip = NO;
524 }
525 return (retval);
526 }
527