1 | /* vi: set sw=8 ts=8: */
|
---|
2 | /*
|
---|
3 | * cut.c - minimalist version of cut
|
---|
4 | *
|
---|
5 | * Copyright (C) 1999,2000,2001 by Lineo, inc.
|
---|
6 | * Written by Mark Whitley <markw@codepoet.org>
|
---|
7 | *
|
---|
8 | * This program is free software; you can redistribute it and/or modify
|
---|
9 | * it under the terms of the GNU General Public License as published by
|
---|
10 | * the Free Software Foundation; either version 2 of the License, or
|
---|
11 | * (at your option) any later version.
|
---|
12 | *
|
---|
13 | * This program is distributed in the hope that it will be useful,
|
---|
14 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
15 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
---|
16 | * General Public License for more details.
|
---|
17 | *
|
---|
18 | * You should have received a copy of the GNU General Public License
|
---|
19 | * along with this program; if not, write to the Free Software
|
---|
20 | * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
---|
21 | *
|
---|
22 | */
|
---|
23 |
|
---|
24 | #include <stdio.h>
|
---|
25 | #include <stdlib.h>
|
---|
26 | #include <unistd.h>
|
---|
27 | #include <string.h>
|
---|
28 | #include <limits.h>
|
---|
29 | #include "busybox.h"
|
---|
30 |
|
---|
31 |
|
---|
32 | /* option vars */
|
---|
33 | static const char optstring[] = "b:c:f:d:sn";
|
---|
34 | #define OPT_BYTE_FLGS 1
|
---|
35 | #define OPT_CHAR_FLGS 2
|
---|
36 | #define OPT_FIELDS_FLGS 4
|
---|
37 | #define OPT_DELIM_FLGS 8
|
---|
38 | #define OPT_SUPRESS_FLGS 16
|
---|
39 | static char part; /* (b)yte, (c)har, (f)ields */
|
---|
40 | static unsigned int supress_non_delimited_lines;
|
---|
41 | static char delim = '\t'; /* delimiter, default is tab */
|
---|
42 |
|
---|
43 | struct cut_list {
|
---|
44 | int startpos;
|
---|
45 | int endpos;
|
---|
46 | };
|
---|
47 |
|
---|
48 | enum {
|
---|
49 | BOL = 0,
|
---|
50 | EOL = INT_MAX,
|
---|
51 | NON_RANGE = -1
|
---|
52 | };
|
---|
53 |
|
---|
54 | static struct cut_list *cut_lists = NULL; /* growable array holding a series of lists */
|
---|
55 | static unsigned int nlists = 0; /* number of elements in above list */
|
---|
56 |
|
---|
57 |
|
---|
58 | static int cmpfunc(const void *a, const void *b)
|
---|
59 | {
|
---|
60 | struct cut_list *la = (struct cut_list *)a;
|
---|
61 | struct cut_list *lb = (struct cut_list *)b;
|
---|
62 |
|
---|
63 | if (la->startpos > lb->startpos)
|
---|
64 | return 1;
|
---|
65 | if (la->startpos < lb->startpos)
|
---|
66 | return -1;
|
---|
67 | return 0;
|
---|
68 | }
|
---|
69 |
|
---|
70 |
|
---|
71 | /*
|
---|
72 | * parse_lists() - parses a list and puts values into startpos and endpos.
|
---|
73 | * valid list formats: N, N-, N-M, -M
|
---|
74 | * more than one list can be separated by commas
|
---|
75 | */
|
---|
76 | static void parse_lists(char *lists)
|
---|
77 | {
|
---|
78 | char *ltok = NULL;
|
---|
79 | char *ntok = NULL;
|
---|
80 | char *junk;
|
---|
81 | int s = 0, e = 0;
|
---|
82 |
|
---|
83 | /* take apart the lists, one by one (they are separated with commas */
|
---|
84 | while ((ltok = strsep(&lists, ",")) != NULL) {
|
---|
85 |
|
---|
86 | /* it's actually legal to pass an empty list */
|
---|
87 | if (strlen(ltok) == 0)
|
---|
88 | continue;
|
---|
89 |
|
---|
90 | /* get the start pos */
|
---|
91 | ntok = strsep(<ok, "-");
|
---|
92 | if (ntok == NULL) {
|
---|
93 | fprintf(stderr, "Help ntok is null for starting position! What do I do?\n");
|
---|
94 | } else if (strlen(ntok) == 0) {
|
---|
95 | s = BOL;
|
---|
96 | } else {
|
---|
97 | s = strtoul(ntok, &junk, 10);
|
---|
98 | if(*junk != '\0' || s < 0)
|
---|
99 | bb_error_msg_and_die("invalid byte or field list");
|
---|
100 |
|
---|
101 | /* account for the fact that arrays are zero based, while the user
|
---|
102 | * expects the first char on the line to be char # 1 */
|
---|
103 | if (s != 0)
|
---|
104 | s--;
|
---|
105 | }
|
---|
106 |
|
---|
107 | /* get the end pos */
|
---|
108 | ntok = strsep(<ok, "-");
|
---|
109 | if (ntok == NULL) {
|
---|
110 | e = NON_RANGE;
|
---|
111 | } else if (strlen(ntok) == 0) {
|
---|
112 | e = EOL;
|
---|
113 | } else {
|
---|
114 | e = strtoul(ntok, &junk, 10);
|
---|
115 | if(*junk != '\0' || e < 0)
|
---|
116 | bb_error_msg_and_die("invalid byte or field list");
|
---|
117 | /* if the user specified and end position of 0, that means "til the
|
---|
118 | * end of the line */
|
---|
119 | if (e == 0)
|
---|
120 | e = INT_MAX;
|
---|
121 | e--; /* again, arrays are zero based, lines are 1 based */
|
---|
122 | if (e == s)
|
---|
123 | e = NON_RANGE;
|
---|
124 | }
|
---|
125 |
|
---|
126 | /* if there's something left to tokenize, the user past an invalid list */
|
---|
127 | if (ltok)
|
---|
128 | bb_error_msg_and_die("invalid byte or field list");
|
---|
129 |
|
---|
130 | /* add the new list */
|
---|
131 | cut_lists = xrealloc(cut_lists, sizeof(struct cut_list) * (++nlists));
|
---|
132 | cut_lists[nlists-1].startpos = s;
|
---|
133 | cut_lists[nlists-1].endpos = e;
|
---|
134 | }
|
---|
135 |
|
---|
136 | /* make sure we got some cut positions out of all that */
|
---|
137 | if (nlists == 0)
|
---|
138 | bb_error_msg_and_die("missing list of positions");
|
---|
139 |
|
---|
140 | /* now that the lists are parsed, we need to sort them to make life easier
|
---|
141 | * on us when it comes time to print the chars / fields / lines */
|
---|
142 | qsort(cut_lists, nlists, sizeof(struct cut_list), cmpfunc);
|
---|
143 |
|
---|
144 | }
|
---|
145 |
|
---|
146 |
|
---|
147 | static void cut_line_by_chars(const char *line)
|
---|
148 | {
|
---|
149 | int c, l;
|
---|
150 | /* set up a list so we can keep track of what's been printed */
|
---|
151 | char *printed = xcalloc(strlen(line), sizeof(char));
|
---|
152 |
|
---|
153 | /* print the chars specified in each cut list */
|
---|
154 | for (c = 0; c < nlists; c++) {
|
---|
155 | l = cut_lists[c].startpos;
|
---|
156 | while (l < strlen(line)) {
|
---|
157 | if (!printed[l]) {
|
---|
158 | putchar(line[l]);
|
---|
159 | printed[l] = 'X';
|
---|
160 | }
|
---|
161 | l++;
|
---|
162 | if (cut_lists[c].endpos == NON_RANGE || l > cut_lists[c].endpos)
|
---|
163 | break;
|
---|
164 | }
|
---|
165 | }
|
---|
166 | putchar('\n'); /* cuz we were handed a chomped line */
|
---|
167 | free(printed);
|
---|
168 | }
|
---|
169 |
|
---|
170 |
|
---|
171 | static void cut_line_by_fields(char *line)
|
---|
172 | {
|
---|
173 | int c, f;
|
---|
174 | int ndelim = -1; /* zero-based / one-based problem */
|
---|
175 | int nfields_printed = 0;
|
---|
176 | char *field = NULL;
|
---|
177 | char d[2] = { delim, 0 };
|
---|
178 | char *printed;
|
---|
179 |
|
---|
180 | /* test the easy case first: does this line contain any delimiters? */
|
---|
181 | if (strchr(line, delim) == NULL) {
|
---|
182 | if (!supress_non_delimited_lines)
|
---|
183 | puts(line);
|
---|
184 | return;
|
---|
185 | }
|
---|
186 |
|
---|
187 | /* set up a list so we can keep track of what's been printed */
|
---|
188 | printed = xcalloc(strlen(line), sizeof(char));
|
---|
189 |
|
---|
190 | /* process each list on this line, for as long as we've got a line to process */
|
---|
191 | for (c = 0; c < nlists && line; c++) {
|
---|
192 | f = cut_lists[c].startpos;
|
---|
193 | do {
|
---|
194 |
|
---|
195 | /* find the field we're looking for */
|
---|
196 | while (line && ndelim < f) {
|
---|
197 | field = strsep(&line, d);
|
---|
198 | ndelim++;
|
---|
199 | }
|
---|
200 |
|
---|
201 | /* we found it, and it hasn't been printed yet */
|
---|
202 | if (field && ndelim == f && !printed[ndelim]) {
|
---|
203 | /* if this isn't our first time through, we need to print the
|
---|
204 | * delimiter after the last field that was printed */
|
---|
205 | if (nfields_printed > 0)
|
---|
206 | putchar(delim);
|
---|
207 | fputs(field, stdout);
|
---|
208 | printed[ndelim] = 'X';
|
---|
209 | nfields_printed++;
|
---|
210 | }
|
---|
211 |
|
---|
212 | f++;
|
---|
213 |
|
---|
214 | /* keep going as long as we have a line to work with, this is a
|
---|
215 | * list, and we're not at the end of that list */
|
---|
216 | } while (line && cut_lists[c].endpos != NON_RANGE && f <= cut_lists[c].endpos);
|
---|
217 | }
|
---|
218 |
|
---|
219 | /* if we printed anything at all, we need to finish it with a newline cuz
|
---|
220 | * we were handed a chomped line */
|
---|
221 | putchar('\n');
|
---|
222 |
|
---|
223 | free(printed);
|
---|
224 | }
|
---|
225 |
|
---|
226 |
|
---|
227 | static void cut_file_by_lines(const char *line, unsigned int linenum)
|
---|
228 | {
|
---|
229 | static int c = 0;
|
---|
230 | static int l = -1;
|
---|
231 |
|
---|
232 | /* I can't initialize this above cuz the "initializer isn't
|
---|
233 | * constant" *sigh* */
|
---|
234 | if (l == -1)
|
---|
235 | l = cut_lists[c].startpos;
|
---|
236 |
|
---|
237 | /* get out if we have no more lists to process or if the lines are lower
|
---|
238 | * than what we're interested in */
|
---|
239 | if (c >= nlists || linenum < l)
|
---|
240 | return;
|
---|
241 |
|
---|
242 | /* if the line we're looking for is lower than the one we were passed, it
|
---|
243 | * means we displayed it already, so move on */
|
---|
244 | while (l < linenum) {
|
---|
245 | l++;
|
---|
246 | /* move on to the next list if we're at the end of this one */
|
---|
247 | if (cut_lists[c].endpos == NON_RANGE || l > cut_lists[c].endpos) {
|
---|
248 | c++;
|
---|
249 | /* get out if there's no more lists to process */
|
---|
250 | if (c >= nlists)
|
---|
251 | return;
|
---|
252 | l = cut_lists[c].startpos;
|
---|
253 | /* get out if the current line is lower than the one we just became
|
---|
254 | * interested in */
|
---|
255 | if (linenum < l)
|
---|
256 | return;
|
---|
257 | }
|
---|
258 | }
|
---|
259 |
|
---|
260 | /* If we made it here, it means we've found the line we're looking for, so print it */
|
---|
261 | puts(line);
|
---|
262 | }
|
---|
263 |
|
---|
264 |
|
---|
265 | /*
|
---|
266 | * snippy-snip
|
---|
267 | */
|
---|
268 | static void cut_file(FILE *file)
|
---|
269 | {
|
---|
270 | char *line = NULL;
|
---|
271 | unsigned int linenum = 0; /* keep these zero-based to be consistent */
|
---|
272 |
|
---|
273 | /* go through every line in the file */
|
---|
274 | while ((line = bb_get_chomped_line_from_file(file)) != NULL) {
|
---|
275 |
|
---|
276 | /* cut based on chars/bytes XXX: only works when sizeof(char) == byte */
|
---|
277 | if ((part & (OPT_CHAR_FLGS | OPT_BYTE_FLGS)))
|
---|
278 | cut_line_by_chars(line);
|
---|
279 |
|
---|
280 | /* cut based on fields */
|
---|
281 | else {
|
---|
282 | if (delim == '\n')
|
---|
283 | cut_file_by_lines(line, linenum);
|
---|
284 | else
|
---|
285 | cut_line_by_fields(line);
|
---|
286 | }
|
---|
287 |
|
---|
288 | linenum++;
|
---|
289 | free(line);
|
---|
290 | }
|
---|
291 | }
|
---|
292 |
|
---|
293 |
|
---|
294 | int cut_main(int argc, char **argv)
|
---|
295 | {
|
---|
296 | unsigned long opt;
|
---|
297 | char *sopt, *sdopt;
|
---|
298 |
|
---|
299 | bb_opt_complementally = "b--bcf:c--bcf:f--bcf";
|
---|
300 | opt = bb_getopt_ulflags(argc, argv, optstring, &sopt, &sopt, &sopt, &sdopt);
|
---|
301 | part = opt & (OPT_BYTE_FLGS|OPT_CHAR_FLGS|OPT_FIELDS_FLGS);
|
---|
302 | if(part == 0)
|
---|
303 | bb_error_msg_and_die("you must specify a list of bytes, characters, or fields");
|
---|
304 | if(opt & BB_GETOPT_ERROR)
|
---|
305 | bb_error_msg_and_die("only one type of list may be specified");
|
---|
306 | parse_lists(sopt);
|
---|
307 | if((opt & (OPT_DELIM_FLGS))) {
|
---|
308 | if (strlen(sdopt) > 1) {
|
---|
309 | bb_error_msg_and_die("the delimiter must be a single character");
|
---|
310 | }
|
---|
311 | delim = sdopt[0];
|
---|
312 | }
|
---|
313 | supress_non_delimited_lines = opt & OPT_SUPRESS_FLGS;
|
---|
314 |
|
---|
315 | /* non-field (char or byte) cutting has some special handling */
|
---|
316 | if (part != OPT_FIELDS_FLGS) {
|
---|
317 | if (supress_non_delimited_lines) {
|
---|
318 | bb_error_msg_and_die("suppressing non-delimited lines makes sense"
|
---|
319 | " only when operating on fields");
|
---|
320 | }
|
---|
321 | if (delim != '\t') {
|
---|
322 | bb_error_msg_and_die("a delimiter may be specified only when operating on fields");
|
---|
323 | }
|
---|
324 | }
|
---|
325 |
|
---|
326 | /* argv[(optind)..(argc-1)] should be names of file to process. If no
|
---|
327 | * files were specified or '-' was specified, take input from stdin.
|
---|
328 | * Otherwise, we process all the files specified. */
|
---|
329 | if (argv[optind] == NULL || (strcmp(argv[optind], "-") == 0)) {
|
---|
330 | cut_file(stdin);
|
---|
331 | }
|
---|
332 | else {
|
---|
333 | int i;
|
---|
334 | FILE *file;
|
---|
335 | for (i = optind; i < argc; i++) {
|
---|
336 | file = bb_wfopen(argv[i], "r");
|
---|
337 | if(file) {
|
---|
338 | cut_file(file);
|
---|
339 | fclose(file);
|
---|
340 | }
|
---|
341 | }
|
---|
342 | }
|
---|
343 |
|
---|
344 | return EXIT_SUCCESS;
|
---|
345 | }
|
---|