diff options
Diffstat (limited to 'busybox/findutils/grep.c')
-rw-r--r-- | busybox/findutils/grep.c | 397 |
1 files changed, 397 insertions, 0 deletions
diff --git a/busybox/findutils/grep.c b/busybox/findutils/grep.c new file mode 100644 index 000000000..29f4ecd4f --- /dev/null +++ b/busybox/findutils/grep.c | |||
@@ -0,0 +1,397 @@ | |||
1 | /* | ||
2 | * Mini grep implementation for busybox using libc regex. | ||
3 | * | ||
4 | * Copyright (C) 1999,2000,2001 by Lineo, inc. and Mark Whitley | ||
5 | * Copyright (C) 1999,2000,2001 by Mark Whitley <markw@codepoet.org> | ||
6 | * | ||
7 | * This program is free software; you can redistribute it and/or modify | ||
8 | * it under the terms of the GNU General Public License as published by | ||
9 | * the Free Software Foundation; either version 2 of the License, or | ||
10 | * (at your option) any later version. | ||
11 | * | ||
12 | * This program is distributed in the hope that it will be useful, | ||
13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | ||
14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | ||
15 | * General Public License for more details. | ||
16 | * | ||
17 | * You should have received a copy of the GNU General Public License | ||
18 | * along with this program; if not, write to the Free Software | ||
19 | * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA | ||
20 | * | ||
21 | */ | ||
22 | /* | ||
23 | * Apr 2004 by Vladimir Oleynik <dzo@simtreas.ru> - | ||
24 | * correction "-e pattern1 -e pattern2" logic and more optimizations. | ||
25 | */ | ||
26 | |||
27 | #include <stdio.h> | ||
28 | #include <stdlib.h> | ||
29 | #include <getopt.h> | ||
30 | #include <regex.h> | ||
31 | #include <string.h> | ||
32 | #include <errno.h> | ||
33 | #include "busybox.h" | ||
34 | |||
35 | |||
36 | /* options */ | ||
37 | #define GREP_OPTS "lnqvscFiHhe:f:L" | ||
38 | #define GREP_OPT_l (1<<0) | ||
39 | static char print_files_with_matches; | ||
40 | #define GREP_OPT_n (1<<1) | ||
41 | static char print_line_num; | ||
42 | #define GREP_OPT_q (1<<2) | ||
43 | static char be_quiet; | ||
44 | #define GREP_OPT_v (1<<3) | ||
45 | typedef char invert_search_t; | ||
46 | static invert_search_t invert_search; | ||
47 | #define GREP_OPT_s (1<<4) | ||
48 | static char suppress_err_msgs; | ||
49 | #define GREP_OPT_c (1<<5) | ||
50 | static char print_match_counts; | ||
51 | #define GREP_OPT_F (1<<6) | ||
52 | static char fgrep_flag; | ||
53 | #define GREP_OPT_i (1<<7) | ||
54 | #define GREP_OPT_H (1<<8) | ||
55 | #define GREP_OPT_h (1<<9) | ||
56 | #define GREP_OPT_e (1<<10) | ||
57 | #define GREP_OPT_f (1<<11) | ||
58 | #define GREP_OPT_L (1<<12) | ||
59 | static char print_files_without_matches; | ||
60 | #ifdef CONFIG_FEATURE_GREP_CONTEXT | ||
61 | #define GREP_OPT_CONTEXT "A:B:C" | ||
62 | #define GREP_OPT_A (1<<13) | ||
63 | #define GREP_OPT_B (1<<14) | ||
64 | #define GREP_OPT_C (1<<15) | ||
65 | #define GREP_OPT_E (1<<16) | ||
66 | #else | ||
67 | #define GREP_OPT_CONTEXT "" | ||
68 | #define GREP_OPT_E (1<<13) | ||
69 | #endif | ||
70 | #ifdef CONFIG_FEATURE_GREP_EGREP_ALIAS | ||
71 | # define OPT_EGREP "E" | ||
72 | #else | ||
73 | # define OPT_EGREP "" | ||
74 | #endif | ||
75 | |||
76 | static int reflags; | ||
77 | static int print_filename; | ||
78 | |||
79 | #ifdef CONFIG_FEATURE_GREP_CONTEXT | ||
80 | static int lines_before; | ||
81 | static int lines_after; | ||
82 | static char **before_buf; | ||
83 | static int last_line_printed; | ||
84 | #endif /* CONFIG_FEATURE_GREP_CONTEXT */ | ||
85 | |||
86 | /* globals used internally */ | ||
87 | static llist_t *pattern_head; /* growable list of patterns to match */ | ||
88 | static char *cur_file; /* the current file we are reading */ | ||
89 | |||
90 | |||
91 | static void print_line(const char *line, int linenum, char decoration) | ||
92 | { | ||
93 | #ifdef CONFIG_FEATURE_GREP_CONTEXT | ||
94 | /* possibly print the little '--' separator */ | ||
95 | if ((lines_before || lines_after) && last_line_printed && | ||
96 | last_line_printed < linenum - 1) { | ||
97 | puts("--"); | ||
98 | } | ||
99 | last_line_printed = linenum; | ||
100 | #endif | ||
101 | if (print_filename) | ||
102 | printf("%s%c", cur_file, decoration); | ||
103 | if (print_line_num) | ||
104 | printf("%i%c", linenum, decoration); | ||
105 | puts(line); | ||
106 | } | ||
107 | |||
108 | extern void xregcomp(regex_t *preg, const char *regex, int cflags); | ||
109 | |||
110 | |||
111 | static int grep_file(FILE *file) | ||
112 | { | ||
113 | char *line; | ||
114 | invert_search_t ret; | ||
115 | int linenum = 0; | ||
116 | int nmatches = 0; | ||
117 | #ifdef CONFIG_FEATURE_GREP_CONTEXT | ||
118 | int print_n_lines_after = 0; | ||
119 | int curpos = 0; /* track where we are in the circular 'before' buffer */ | ||
120 | int idx = 0; /* used for iteration through the circular buffer */ | ||
121 | #endif /* CONFIG_FEATURE_GREP_CONTEXT */ | ||
122 | |||
123 | while ((line = bb_get_chomped_line_from_file(file)) != NULL) { | ||
124 | llist_t *pattern_ptr = pattern_head; | ||
125 | |||
126 | linenum++; | ||
127 | ret = 0; | ||
128 | while (pattern_ptr) { | ||
129 | if (fgrep_flag) { | ||
130 | ret = strstr(line, pattern_ptr->data) != NULL; | ||
131 | } else { | ||
132 | /* | ||
133 | * test for a postitive-assertion match (regexec returns success (0) | ||
134 | * and the user did not specify invert search), or a negative-assertion | ||
135 | * match (regexec returns failure (REG_NOMATCH) and the user specified | ||
136 | * invert search) | ||
137 | */ | ||
138 | regex_t regex; | ||
139 | xregcomp(®ex, pattern_ptr->data, reflags); | ||
140 | ret |= regexec(®ex, line, 0, NULL, 0) == 0; | ||
141 | regfree(®ex); | ||
142 | } | ||
143 | pattern_ptr = pattern_ptr->link; | ||
144 | } /* while (pattern_ptr) */ | ||
145 | |||
146 | if ((ret ^ invert_search)) { | ||
147 | |||
148 | if (print_files_with_matches || be_quiet) | ||
149 | free(line); | ||
150 | |||
151 | /* if we found a match but were told to be quiet, stop here */ | ||
152 | if (be_quiet || print_files_without_matches) | ||
153 | return -1; | ||
154 | |||
155 | /* keep track of matches */ | ||
156 | nmatches++; | ||
157 | |||
158 | /* if we're just printing filenames, we stop after the first match */ | ||
159 | if (print_files_with_matches) | ||
160 | break; | ||
161 | |||
162 | /* print the matched line */ | ||
163 | if (print_match_counts == 0) { | ||
164 | #ifdef CONFIG_FEATURE_GREP_CONTEXT | ||
165 | int prevpos = (curpos == 0) ? lines_before - 1 : curpos - 1; | ||
166 | |||
167 | /* if we were told to print 'before' lines and there is at least | ||
168 | * one line in the circular buffer, print them */ | ||
169 | if (lines_before && before_buf[prevpos] != NULL) { | ||
170 | int first_buf_entry_line_num = linenum - lines_before; | ||
171 | |||
172 | /* advance to the first entry in the circular buffer, and | ||
173 | * figure out the line number is of the first line in the | ||
174 | * buffer */ | ||
175 | idx = curpos; | ||
176 | while (before_buf[idx] == NULL) { | ||
177 | idx = (idx + 1) % lines_before; | ||
178 | first_buf_entry_line_num++; | ||
179 | } | ||
180 | |||
181 | /* now print each line in the buffer, clearing them as we go */ | ||
182 | while (before_buf[idx] != NULL) { | ||
183 | print_line(before_buf[idx], first_buf_entry_line_num, '-'); | ||
184 | free(before_buf[idx]); | ||
185 | before_buf[idx] = NULL; | ||
186 | idx = (idx + 1) % lines_before; | ||
187 | first_buf_entry_line_num++; | ||
188 | } | ||
189 | } | ||
190 | |||
191 | /* make a note that we need to print 'after' lines */ | ||
192 | print_n_lines_after = lines_after; | ||
193 | #endif /* CONFIG_FEATURE_GREP_CONTEXT */ | ||
194 | print_line(line, linenum, ':'); | ||
195 | } | ||
196 | } | ||
197 | #ifdef CONFIG_FEATURE_GREP_CONTEXT | ||
198 | else { /* no match */ | ||
199 | /* Add the line to the circular 'before' buffer */ | ||
200 | if(lines_before) { | ||
201 | free(before_buf[curpos]); | ||
202 | before_buf[curpos] = bb_xstrdup(line); | ||
203 | curpos = (curpos + 1) % lines_before; | ||
204 | } | ||
205 | } | ||
206 | |||
207 | /* if we need to print some context lines after the last match, do so */ | ||
208 | if (print_n_lines_after && (last_line_printed != linenum)) { | ||
209 | print_line(line, linenum, '-'); | ||
210 | print_n_lines_after--; | ||
211 | } | ||
212 | #endif /* CONFIG_FEATURE_GREP_CONTEXT */ | ||
213 | free(line); | ||
214 | } | ||
215 | |||
216 | |||
217 | /* special-case file post-processing for options where we don't print line | ||
218 | * matches, just filenames and possibly match counts */ | ||
219 | |||
220 | /* grep -c: print [filename:]count, even if count is zero */ | ||
221 | if (print_match_counts) { | ||
222 | if (print_filename) | ||
223 | printf("%s:", cur_file); | ||
224 | printf("%d\n", nmatches); | ||
225 | } | ||
226 | |||
227 | /* grep -l: print just the filename, but only if we grepped the line in the file */ | ||
228 | if (print_files_with_matches && nmatches > 0) { | ||
229 | puts(cur_file); | ||
230 | } | ||
231 | |||
232 | /* grep -L: print just the filename, but only if we didn't grep the line in the file */ | ||
233 | if (print_files_without_matches && nmatches == 0) { | ||
234 | puts(cur_file); | ||
235 | } | ||
236 | |||
237 | return nmatches; | ||
238 | } | ||
239 | |||
240 | static void load_regexes_from_file(llist_t *fopt) | ||
241 | { | ||
242 | char *line; | ||
243 | FILE *f; | ||
244 | |||
245 | while(fopt) { | ||
246 | llist_t *cur = fopt; | ||
247 | char *ffile = cur->data; | ||
248 | |||
249 | fopt = cur->link; | ||
250 | free(cur); | ||
251 | f = bb_xfopen(ffile, "r"); | ||
252 | while ((line = bb_get_chomped_line_from_file(f)) != NULL) { | ||
253 | pattern_head = llist_add_to(pattern_head, line); | ||
254 | } | ||
255 | } | ||
256 | } | ||
257 | |||
258 | |||
259 | extern int grep_main(int argc, char **argv) | ||
260 | { | ||
261 | FILE *file; | ||
262 | int matched; | ||
263 | unsigned long opt; | ||
264 | llist_t *fopt = NULL; | ||
265 | |||
266 | /* do normal option parsing */ | ||
267 | #ifdef CONFIG_FEATURE_GREP_CONTEXT | ||
268 | { | ||
269 | char *junk; | ||
270 | char *slines_after; | ||
271 | char *slines_before; | ||
272 | char *Copt; | ||
273 | |||
274 | bb_opt_complementaly = "H-h:e*:f*:C-AB"; | ||
275 | opt = bb_getopt_ulflags(argc, argv, | ||
276 | GREP_OPTS GREP_OPT_CONTEXT OPT_EGREP, | ||
277 | &pattern_head, &fopt, | ||
278 | &slines_after, &slines_before, &Copt); | ||
279 | |||
280 | if(opt & GREP_OPT_C) { | ||
281 | /* C option unseted A and B options, but next -A or -B | ||
282 | may be ovewrite own option */ | ||
283 | if(!(opt & GREP_OPT_A)) /* not overwtited */ | ||
284 | slines_after = Copt; | ||
285 | if(!(opt & GREP_OPT_B)) /* not overwtited */ | ||
286 | slines_before = Copt; | ||
287 | opt |= GREP_OPT_A|GREP_OPT_B; /* set for parse now */ | ||
288 | } | ||
289 | if(opt & GREP_OPT_A) { | ||
290 | lines_after = strtoul(slines_after, &junk, 10); | ||
291 | if(*junk != '\0') | ||
292 | bb_error_msg_and_die("invalid context length argument"); | ||
293 | } | ||
294 | if(opt & GREP_OPT_B) { | ||
295 | lines_before = strtoul(slines_before, &junk, 10); | ||
296 | if(*junk != '\0') | ||
297 | bb_error_msg_and_die("invalid context length argument"); | ||
298 | } | ||
299 | /* sanity checks after parse may be invalid numbers ;-) */ | ||
300 | if ((opt & (GREP_OPT_c|GREP_OPT_q|GREP_OPT_l|GREP_OPT_L))) { | ||
301 | opt &= ~GREP_OPT_n; | ||
302 | lines_before = 0; | ||
303 | lines_after = 0; | ||
304 | } else if(lines_before > 0) | ||
305 | before_buf = (char **)xcalloc(lines_before, sizeof(char *)); | ||
306 | } | ||
307 | #else | ||
308 | /* with auto sanity checks */ | ||
309 | bb_opt_complementaly = "H-h:e*:f*:c-n:q-n:l-n"; | ||
310 | opt = bb_getopt_ulflags(argc, argv, GREP_OPTS OPT_EGREP, | ||
311 | &pattern_head, &fopt); | ||
312 | |||
313 | #endif | ||
314 | print_files_with_matches = opt & GREP_OPT_l; | ||
315 | print_files_without_matches = (opt & GREP_OPT_L) != 0; | ||
316 | print_line_num = opt & GREP_OPT_n; | ||
317 | be_quiet = opt & GREP_OPT_q; | ||
318 | invert_search = (opt & GREP_OPT_v) != 0; /* 0 | 1 */ | ||
319 | suppress_err_msgs = opt & GREP_OPT_s; | ||
320 | print_match_counts = opt & GREP_OPT_c; | ||
321 | fgrep_flag = opt & GREP_OPT_F; | ||
322 | if(opt & GREP_OPT_H) | ||
323 | print_filename++; | ||
324 | if(opt & GREP_OPT_h) | ||
325 | print_filename--; | ||
326 | if(opt & GREP_OPT_f) | ||
327 | load_regexes_from_file(fopt); | ||
328 | |||
329 | #ifdef CONFIG_FEATURE_GREP_EGREP_ALIAS | ||
330 | if(bb_applet_name[0] == 'e' || (opt & GREP_OPT_E)) | ||
331 | reflags = REG_EXTENDED | REG_NOSUB; | ||
332 | else | ||
333 | #endif | ||
334 | reflags = REG_NOSUB; | ||
335 | |||
336 | if(opt & GREP_OPT_i) | ||
337 | reflags |= REG_ICASE; | ||
338 | |||
339 | argv += optind; | ||
340 | argc -= optind; | ||
341 | |||
342 | /* if we didn't get a pattern from a -e and no command file was specified, | ||
343 | * argv[optind] should be the pattern. no pattern, no worky */ | ||
344 | if (pattern_head == NULL) { | ||
345 | if (*argv == NULL) | ||
346 | bb_show_usage(); | ||
347 | else { | ||
348 | pattern_head = llist_add_to(pattern_head, *argv++); | ||
349 | argc--; | ||
350 | } | ||
351 | } | ||
352 | |||
353 | /* argv[(optind)..(argc-1)] should be names of file to grep through. If | ||
354 | * there is more than one file to grep, we will print the filenames */ | ||
355 | if (argc > 1) { | ||
356 | print_filename++; | ||
357 | |||
358 | /* If no files were specified, or '-' was specified, take input from | ||
359 | * stdin. Otherwise, we grep through all the files specified. */ | ||
360 | } else if (argc == 0) { | ||
361 | argc++; | ||
362 | } | ||
363 | matched = 0; | ||
364 | while (argc--) { | ||
365 | cur_file = *argv++; | ||
366 | if(!cur_file || (*cur_file == '-' && !cur_file[1])) { | ||
367 | cur_file = "-"; | ||
368 | file = stdin; | ||
369 | } else { | ||
370 | file = fopen(cur_file, "r"); | ||
371 | } | ||
372 | if (file == NULL) { | ||
373 | if (!suppress_err_msgs) | ||
374 | bb_perror_msg("%s", cur_file); | ||
375 | } else { | ||
376 | matched += grep_file(file); | ||
377 | if(matched < 0) { | ||
378 | /* we found a match but were told to be quiet, stop here and | ||
379 | * return success */ | ||
380 | break; | ||
381 | } | ||
382 | fclose(file); | ||
383 | } | ||
384 | } | ||
385 | |||
386 | #ifdef CONFIG_FEATURE_CLEAN_UP | ||
387 | /* destroy all the elments in the pattern list */ | ||
388 | while (pattern_head) { | ||
389 | llist_t *pattern_head_ptr = pattern_head; | ||
390 | |||
391 | pattern_head = pattern_head->link; | ||
392 | free(pattern_head_ptr); | ||
393 | } | ||
394 | #endif | ||
395 | |||
396 | return !matched; /* invert return value 0 = success, 1 = failed */ | ||
397 | } | ||