RPM package re-classification.
[git/jnareb-git/bp-gitweb.git] / grep.c
blobc411ddd4d52e2aa68edd979733dd0b15f51d5f32
1 #include "cache.h"
2 #include <regex.h>
3 #include "grep.h"
5 void append_grep_pattern(struct grep_opt *opt, const char *pat,
6 const char *origin, int no, enum grep_pat_token t)
8 struct grep_pat *p = xcalloc(1, sizeof(*p));
9 p->pattern = pat;
10 p->origin = origin;
11 p->no = no;
12 p->token = t;
13 *opt->pattern_tail = p;
14 opt->pattern_tail = &p->next;
15 p->next = NULL;
18 static void compile_regexp(struct grep_pat *p, struct grep_opt *opt)
20 int err = regcomp(&p->regexp, p->pattern, opt->regflags);
21 if (err) {
22 char errbuf[1024];
23 char where[1024];
24 if (p->no)
25 sprintf(where, "In '%s' at %d, ",
26 p->origin, p->no);
27 else if (p->origin)
28 sprintf(where, "%s, ", p->origin);
29 else
30 where[0] = 0;
31 regerror(err, &p->regexp, errbuf, 1024);
32 regfree(&p->regexp);
33 die("%s'%s': %s", where, p->pattern, errbuf);
37 static struct grep_expr *compile_pattern_expr(struct grep_pat **);
38 static struct grep_expr *compile_pattern_atom(struct grep_pat **list)
40 struct grep_pat *p;
41 struct grep_expr *x;
43 p = *list;
44 switch (p->token) {
45 case GREP_PATTERN: /* atom */
46 case GREP_PATTERN_HEAD:
47 case GREP_PATTERN_BODY:
48 x = xcalloc(1, sizeof (struct grep_expr));
49 x->node = GREP_NODE_ATOM;
50 x->u.atom = p;
51 *list = p->next;
52 return x;
53 case GREP_OPEN_PAREN:
54 *list = p->next;
55 x = compile_pattern_expr(list);
56 if (!x)
57 return NULL;
58 if (!*list || (*list)->token != GREP_CLOSE_PAREN)
59 die("unmatched parenthesis");
60 *list = (*list)->next;
61 return x;
62 default:
63 return NULL;
67 static struct grep_expr *compile_pattern_not(struct grep_pat **list)
69 struct grep_pat *p;
70 struct grep_expr *x;
72 p = *list;
73 switch (p->token) {
74 case GREP_NOT:
75 if (!p->next)
76 die("--not not followed by pattern expression");
77 *list = p->next;
78 x = xcalloc(1, sizeof (struct grep_expr));
79 x->node = GREP_NODE_NOT;
80 x->u.unary = compile_pattern_not(list);
81 if (!x->u.unary)
82 die("--not followed by non pattern expression");
83 return x;
84 default:
85 return compile_pattern_atom(list);
89 static struct grep_expr *compile_pattern_and(struct grep_pat **list)
91 struct grep_pat *p;
92 struct grep_expr *x, *y, *z;
94 x = compile_pattern_not(list);
95 p = *list;
96 if (p && p->token == GREP_AND) {
97 if (!p->next)
98 die("--and not followed by pattern expression");
99 *list = p->next;
100 y = compile_pattern_and(list);
101 if (!y)
102 die("--and not followed by pattern expression");
103 z = xcalloc(1, sizeof (struct grep_expr));
104 z->node = GREP_NODE_AND;
105 z->u.binary.left = x;
106 z->u.binary.right = y;
107 return z;
109 return x;
112 static struct grep_expr *compile_pattern_or(struct grep_pat **list)
114 struct grep_pat *p;
115 struct grep_expr *x, *y, *z;
117 x = compile_pattern_and(list);
118 p = *list;
119 if (x && p && p->token != GREP_CLOSE_PAREN) {
120 y = compile_pattern_or(list);
121 if (!y)
122 die("not a pattern expression %s", p->pattern);
123 z = xcalloc(1, sizeof (struct grep_expr));
124 z->node = GREP_NODE_OR;
125 z->u.binary.left = x;
126 z->u.binary.right = y;
127 return z;
129 return x;
132 static struct grep_expr *compile_pattern_expr(struct grep_pat **list)
134 return compile_pattern_or(list);
137 void compile_grep_patterns(struct grep_opt *opt)
139 struct grep_pat *p;
141 for (p = opt->pattern_list; p; p = p->next) {
142 switch (p->token) {
143 case GREP_PATTERN: /* atom */
144 case GREP_PATTERN_HEAD:
145 case GREP_PATTERN_BODY:
146 if (!opt->fixed)
147 compile_regexp(p, opt);
148 break;
149 default:
150 opt->extended = 1;
151 break;
155 if (!opt->extended)
156 return;
158 /* Then bundle them up in an expression.
159 * A classic recursive descent parser would do.
161 p = opt->pattern_list;
162 opt->pattern_expression = compile_pattern_expr(&p);
163 if (p)
164 die("incomplete pattern expression: %s", p->pattern);
167 static void free_pattern_expr(struct grep_expr *x)
169 switch (x->node) {
170 case GREP_NODE_ATOM:
171 break;
172 case GREP_NODE_NOT:
173 free_pattern_expr(x->u.unary);
174 break;
175 case GREP_NODE_AND:
176 case GREP_NODE_OR:
177 free_pattern_expr(x->u.binary.left);
178 free_pattern_expr(x->u.binary.right);
179 break;
181 free(x);
184 void free_grep_patterns(struct grep_opt *opt)
186 struct grep_pat *p, *n;
188 for (p = opt->pattern_list; p; p = n) {
189 n = p->next;
190 switch (p->token) {
191 case GREP_PATTERN: /* atom */
192 case GREP_PATTERN_HEAD:
193 case GREP_PATTERN_BODY:
194 regfree(&p->regexp);
195 break;
196 default:
197 break;
199 free(p);
202 if (!opt->extended)
203 return;
204 free_pattern_expr(opt->pattern_expression);
207 static char *end_of_line(char *cp, unsigned long *left)
209 unsigned long l = *left;
210 while (l && *cp != '\n') {
211 l--;
212 cp++;
214 *left = l;
215 return cp;
218 static int word_char(char ch)
220 return isalnum(ch) || ch == '_';
223 static void show_line(struct grep_opt *opt, const char *bol, const char *eol,
224 const char *name, unsigned lno, char sign)
226 if (opt->pathname)
227 printf("%s%c", name, sign);
228 if (opt->linenum)
229 printf("%d%c", lno, sign);
230 printf("%.*s\n", (int)(eol-bol), bol);
234 * NEEDSWORK: share code with diff.c
236 #define FIRST_FEW_BYTES 8000
237 static int buffer_is_binary(const char *ptr, unsigned long size)
239 if (FIRST_FEW_BYTES < size)
240 size = FIRST_FEW_BYTES;
241 return !!memchr(ptr, 0, size);
244 static int fixmatch(const char *pattern, char *line, regmatch_t *match)
246 char *hit = strstr(line, pattern);
247 if (!hit) {
248 match->rm_so = match->rm_eo = -1;
249 return REG_NOMATCH;
251 else {
252 match->rm_so = hit - line;
253 match->rm_eo = match->rm_so + strlen(pattern);
254 return 0;
258 static int match_one_pattern(struct grep_opt *opt, struct grep_pat *p, char *bol, char *eol, enum grep_context ctx)
260 int hit = 0;
261 int at_true_bol = 1;
262 regmatch_t pmatch[10];
264 if ((p->token != GREP_PATTERN) &&
265 ((p->token == GREP_PATTERN_HEAD) != (ctx == GREP_CONTEXT_HEAD)))
266 return 0;
268 again:
269 if (!opt->fixed) {
270 regex_t *exp = &p->regexp;
271 hit = !regexec(exp, bol, ARRAY_SIZE(pmatch),
272 pmatch, 0);
274 else {
275 hit = !fixmatch(p->pattern, bol, pmatch);
278 if (hit && opt->word_regexp) {
279 if ((pmatch[0].rm_so < 0) ||
280 (eol - bol) <= pmatch[0].rm_so ||
281 (pmatch[0].rm_eo < 0) ||
282 (eol - bol) < pmatch[0].rm_eo)
283 die("regexp returned nonsense");
285 /* Match beginning must be either beginning of the
286 * line, or at word boundary (i.e. the last char must
287 * not be a word char). Similarly, match end must be
288 * either end of the line, or at word boundary
289 * (i.e. the next char must not be a word char).
291 if ( ((pmatch[0].rm_so == 0 && at_true_bol) ||
292 !word_char(bol[pmatch[0].rm_so-1])) &&
293 ((pmatch[0].rm_eo == (eol-bol)) ||
294 !word_char(bol[pmatch[0].rm_eo])) )
296 else
297 hit = 0;
299 if (!hit && pmatch[0].rm_so + bol + 1 < eol) {
300 /* There could be more than one match on the
301 * line, and the first match might not be
302 * strict word match. But later ones could be!
304 bol = pmatch[0].rm_so + bol + 1;
305 at_true_bol = 0;
306 goto again;
309 return hit;
312 static int match_expr_eval(struct grep_opt *opt,
313 struct grep_expr *x,
314 char *bol, char *eol,
315 enum grep_context ctx)
317 switch (x->node) {
318 case GREP_NODE_ATOM:
319 return match_one_pattern(opt, x->u.atom, bol, eol, ctx);
320 break;
321 case GREP_NODE_NOT:
322 return !match_expr_eval(opt, x->u.unary, bol, eol, ctx);
323 case GREP_NODE_AND:
324 return (match_expr_eval(opt, x->u.binary.left, bol, eol, ctx) &&
325 match_expr_eval(opt, x->u.binary.right, bol, eol, ctx));
326 case GREP_NODE_OR:
327 return (match_expr_eval(opt, x->u.binary.left, bol, eol, ctx) ||
328 match_expr_eval(opt, x->u.binary.right, bol, eol, ctx));
330 die("Unexpected node type (internal error) %d\n", x->node);
333 static int match_expr(struct grep_opt *opt, char *bol, char *eol,
334 enum grep_context ctx)
336 struct grep_expr *x = opt->pattern_expression;
337 return match_expr_eval(opt, x, bol, eol, ctx);
340 static int match_line(struct grep_opt *opt, char *bol, char *eol,
341 enum grep_context ctx)
343 struct grep_pat *p;
344 if (opt->extended)
345 return match_expr(opt, bol, eol, ctx);
346 for (p = opt->pattern_list; p; p = p->next) {
347 if (match_one_pattern(opt, p, bol, eol, ctx))
348 return 1;
350 return 0;
353 int grep_buffer(struct grep_opt *opt, const char *name, char *buf, unsigned long size)
355 char *bol = buf;
356 unsigned long left = size;
357 unsigned lno = 1;
358 struct pre_context_line {
359 char *bol;
360 char *eol;
361 } *prev = NULL, *pcl;
362 unsigned last_hit = 0;
363 unsigned last_shown = 0;
364 int binary_match_only = 0;
365 const char *hunk_mark = "";
366 unsigned count = 0;
367 enum grep_context ctx = GREP_CONTEXT_HEAD;
369 if (buffer_is_binary(buf, size)) {
370 switch (opt->binary) {
371 case GREP_BINARY_DEFAULT:
372 binary_match_only = 1;
373 break;
374 case GREP_BINARY_NOMATCH:
375 return 0; /* Assume unmatch */
376 break;
377 default:
378 break;
382 if (opt->pre_context)
383 prev = xcalloc(opt->pre_context, sizeof(*prev));
384 if (opt->pre_context || opt->post_context)
385 hunk_mark = "--\n";
387 while (left) {
388 char *eol, ch;
389 int hit = 0;
391 eol = end_of_line(bol, &left);
392 ch = *eol;
393 *eol = 0;
395 if ((ctx == GREP_CONTEXT_HEAD) && (eol == bol))
396 ctx = GREP_CONTEXT_BODY;
398 hit = match_line(opt, bol, eol, ctx);
399 *eol = ch;
401 /* "grep -v -e foo -e bla" should list lines
402 * that do not have either, so inversion should
403 * be done outside.
405 if (opt->invert)
406 hit = !hit;
407 if (opt->unmatch_name_only) {
408 if (hit)
409 return 0;
410 goto next_line;
412 if (hit) {
413 count++;
414 if (opt->status_only)
415 return 1;
416 if (binary_match_only) {
417 printf("Binary file %s matches\n", name);
418 return 1;
420 if (opt->name_only) {
421 printf("%s\n", name);
422 return 1;
424 /* Hit at this line. If we haven't shown the
425 * pre-context lines, we would need to show them.
426 * When asked to do "count", this still show
427 * the context which is nonsense, but the user
428 * deserves to get that ;-).
430 if (opt->pre_context) {
431 unsigned from;
432 if (opt->pre_context < lno)
433 from = lno - opt->pre_context;
434 else
435 from = 1;
436 if (from <= last_shown)
437 from = last_shown + 1;
438 if (last_shown && from != last_shown + 1)
439 printf(hunk_mark);
440 while (from < lno) {
441 pcl = &prev[lno-from-1];
442 show_line(opt, pcl->bol, pcl->eol,
443 name, from, '-');
444 from++;
446 last_shown = lno-1;
448 if (last_shown && lno != last_shown + 1)
449 printf(hunk_mark);
450 if (!opt->count)
451 show_line(opt, bol, eol, name, lno, ':');
452 last_shown = last_hit = lno;
454 else if (last_hit &&
455 lno <= last_hit + opt->post_context) {
456 /* If the last hit is within the post context,
457 * we need to show this line.
459 if (last_shown && lno != last_shown + 1)
460 printf(hunk_mark);
461 show_line(opt, bol, eol, name, lno, '-');
462 last_shown = lno;
464 if (opt->pre_context) {
465 memmove(prev+1, prev,
466 (opt->pre_context-1) * sizeof(*prev));
467 prev->bol = bol;
468 prev->eol = eol;
471 next_line:
472 bol = eol + 1;
473 if (!left)
474 break;
475 left--;
476 lno++;
479 free(prev);
481 if (opt->status_only)
482 return 0;
483 if (opt->unmatch_name_only) {
484 /* We did not see any hit, so we want to show this */
485 printf("%s\n", name);
486 return 1;
489 /* NEEDSWORK:
490 * The real "grep -c foo *.c" gives many "bar.c:0" lines,
491 * which feels mostly useless but sometimes useful. Maybe
492 * make it another option? For now suppress them.
494 if (opt->count && count)
495 printf("%s:%u\n", name, count);
496 return !!last_hit;