csplit.c revision 97977 
1/*-
2 * Copyright (c) 2002 Tim J. Robbins.
3 * 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 *
14 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
15 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
16 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
17 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
18 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
19 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
20 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
21 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
22 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
23 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
24 * SUCH DAMAGE.
25 */
26
27/*
28 * csplit -- split files based on context
29 *
30 * This utility splits its input into numbered output files by line number
31 * or by a regular expression. Regular expression matches have an optional
32 * offset with them, allowing the split to occur a specified number of
33 * lines before or after the match.
34 *
35 * To handle negative offsets, we stop reading when the match occurs and
36 * store the offset that the file should have been split at, then use
37 * this output file as input until all the "overflowed" lines have been read.
38 * The file is then closed and truncated to the correct length.
39 *
40 * We assume that the output files can be seeked upon (ie. they cannot be
41 * symlinks to named pipes or character devices), but make no such
42 * assumption about the input.
43 */
44
45#include <sys/cdefs.h>
46__FBSDID("$FreeBSD: head/usr.bin/csplit/csplit.c 97977 2002-06-07 01:04:24Z tjr $");
47
48#include <sys/types.h>
49
50#include <ctype.h>
51#include <err.h>
52#include <errno.h>
53#include <limits.h>
54#include <locale.h>
55#include <regex.h>
56#include <signal.h>
57#include <stdint.h>
58#include <stdio.h>
59#include <stdlib.h>
60#include <string.h>
61#include <unistd.h>
62
63void	 cleanup(void);
64void	 do_lineno(const char *);
65void	 do_rexp(const char *);
66char	*getline(void);
67void	 handlesig(int);
68FILE	*newfile(void);
69void	 toomuch(FILE *, long);
70void	 usage(void);
71
72/*
73 * Command line options
74 */
75const char *prefix;		/* File name prefix */
76long	 sufflen;		/* Number of decimal digits for suffix */
77int	 sflag;			/* Suppress output of file names */
78int	 kflag;			/* Keep output if error occurs */
79
80/*
81 * Other miscellaneous globals (XXX too many)
82 */
83long	 lineno;		/* Current line number in input file */
84long	 reps;			/* Number of repetitions for this pattern */
85long	 nfiles;		/* Number of files output so far */
86long	 maxfiles;		/* Maximum number of files we can create */
87char	 currfile[PATH_MAX];	/* Current output file */
88const char *infn;		/* Name of the input file */
89FILE	*infile;		/* Input file handle */
90FILE	*overfile;		/* Overflow file for toomuch() */
91off_t	 truncofs;		/* Offset this file should be truncated at */
92int	 doclean;		/* Should cleanup() remove output? */
93
94int
95main(int argc, char *argv[])
96{
97	long i;
98	int ch;
99	const char *expr;
100	char *ep, *p;
101	FILE *ofp;
102
103	setlocale(LC_ALL, "");
104
105	kflag = sflag = 0;
106	prefix = "xx";
107	sufflen = 2;
108	while ((ch = getopt(argc, argv, "ksf:n:")) > 0) {
109		switch (ch) {
110		case 'f':
111			prefix = optarg;
112			break;
113		case 'k':
114			kflag = 1;
115			break;
116		case 'n':
117			errno = 0;
118			sufflen = strtol(optarg, &ep, 10);
119			if (sufflen <= 0 || *ep != '\0' || errno != 0)
120				errx(1, "%s: bad suffix length", optarg);
121			break;
122		case 's':
123			sflag = 1;
124			break;
125		default:
126			usage();
127			/*NOTREACHED*/
128		}
129	}
130
131	if (sufflen + strlen(prefix) >= PATH_MAX)
132		errx(1, "name too long");
133
134	argc -= optind;
135	argv += optind;
136
137	if ((infn = *argv++) == NULL)
138		usage();
139	if (strcmp(infn, "-") == 0) {
140		infile = stdin;
141		infn = "stdin";
142	} else if ((infile = fopen(infn, "r")) == NULL)
143		err(1, "%s", infn);
144
145	if (!kflag) {
146		doclean = 1;
147		atexit(cleanup);
148		signal(SIGHUP, handlesig);
149		signal(SIGINT, handlesig);
150		signal(SIGTERM, handlesig);
151	}
152
153	lineno = 0;
154	nfiles = 0;
155	truncofs = 0;
156	overfile = NULL;
157
158	/* Ensure 10^sufflen < LONG_MAX. */
159	for (maxfiles = 1, i = 0; i < sufflen; i++) {
160		if (maxfiles > LONG_MAX / 10)
161			errx(1, "%ld: suffix too long (limit %ld)",
162			    sufflen, i);
163		maxfiles *= 10;
164	}
165
166	/* Create files based on supplied patterns. */
167	while (nfiles < maxfiles - 1 && (expr = *argv++) != NULL) {
168		/* Look ahead & see if this pattern has any repetitions. */
169		if (*argv != NULL && **argv == '{') {
170			errno = 0;
171			reps = strtol(*argv + 1, &ep, 10);
172			if (reps < 0 || *ep != '}' || errno != 0)
173				errx(1, "%s: bad repetition count", *argv + 1);
174			argv++;
175		} else
176			reps = 0;
177
178		if (*expr == '/' || *expr == '%') {
179			do
180				do_rexp(expr);
181			while (reps-- != 0 && nfiles < maxfiles - 1);
182		} else if (isdigit((unsigned char)*expr))
183			do_lineno(expr);
184		else
185			errx(1, "%s: unrecognised pattern", expr);
186	}
187
188	/* Copy the rest into a new file. */
189	if (!feof(infile)) {
190		ofp = newfile();
191		while ((p = getline()) != NULL && fputs(p, ofp) == 0)
192			;
193		if (!sflag)
194			printf("%jd\n", (intmax_t)ftello(ofp));
195		if (fclose(ofp) != 0)
196			err(1, "%s", currfile);
197	}
198
199	toomuch(NULL, 0);
200	doclean = 0;
201
202	return (0);
203}
204
205void
206usage(void)
207{
208
209	fprintf(stderr,
210"usage: csplit [-ks] [-f prefix] [-n number] file args ...\n");
211	exit(1);
212}
213
214void
215handlesig(int sig __unused)
216{
217	const char msg[] = "csplit: caught signal, cleaning up\n";
218
219	write(STDERR_FILENO, msg, sizeof(msg) - 1);
220	cleanup();
221	_exit(2);
222}
223
224/* Create a new output file. */
225FILE *
226newfile(void)
227{
228	FILE *fp;
229
230	snprintf(currfile, sizeof(currfile), "%s%0*ld", prefix, (int)sufflen,
231	    nfiles);
232	if ((fp = fopen(currfile, "w+")) == NULL)
233		err(1, "%s", currfile);
234	nfiles++;
235
236	return (fp);
237}
238
239/* Remove partial output, called before exiting. */
240void
241cleanup(void)
242{
243	char fnbuf[PATH_MAX];
244	long i;
245
246	if (!doclean)
247		return;
248
249	/*
250	 * NOTE: One cannot portably assume to be able to call snprintf()
251	 * from inside a signal handler. It does, however, appear to be safe
252	 * to do on FreeBSD. The solution to this problem is worse than the
253	 * problem itself.
254	 */
255
256	for (i = 0; i < nfiles; i++) {
257		snprintf(fnbuf, sizeof(fnbuf), "%s%0*ld", prefix,
258		    (int)sufflen, i);
259		unlink(fnbuf);
260	}
261}
262
263/* Read a line from the input into a static buffer. */
264char *
265getline(void)
266{
267	static char lbuf[LINE_MAX];
268	FILE *src;
269
270	src = overfile != NULL ? overfile : infile;
271
272again: if (fgets(lbuf, sizeof(lbuf), src) == NULL) {
273		if (src == overfile) {
274			src = infile;
275			goto again;
276		}
277		return (NULL);
278	}
279	if (ferror(src))
280		err(1, "%s", infn);
281	lineno++;
282
283	return (lbuf);
284}
285
286/* Conceptually rewind the input (as obtained by getline()) back `n' lines. */
287void
288toomuch(FILE *ofp, long n)
289{
290	char buf[BUFSIZ];
291	size_t i, nread;
292
293	if (overfile != NULL) {
294		/*
295		 * Truncate the previous file we overflowed into back to
296		 * the correct length, close it.
297		 */
298		if (fflush(overfile) != 0)
299			err(1, "overflow");
300		if (ftruncate(fileno(overfile), truncofs) != 0)
301			err(1, "overflow");
302		if (fclose(overfile) != 0)
303			err(1, "overflow");
304		overfile = NULL;
305	}
306
307	if (n == 0)
308		/* Just tidying up */
309		return;
310
311	lineno -= n;
312
313	/*
314	 * Wind the overflow file backwards to `n' lines before the
315	 * current one.
316	 */
317	do {
318		if (ftello(ofp) < (off_t)sizeof(buf))
319			rewind(ofp);
320		else
321			fseek(ofp, -(long)sizeof(buf), SEEK_CUR);
322		if (ferror(ofp))
323			errx(1, "%s: can't seek", currfile);
324		if ((nread = fread(buf, 1, sizeof(buf), ofp)) == 0)
325			errx(1, "can't read overflowed output");
326		if (fseek(ofp, -(long)nread, SEEK_CUR) != 0)
327			err(1, "%s", currfile);
328		for (i = 1; i <= nread; i++)
329			if (buf[nread - i] == '\n' && n-- == 0)
330				break;
331		if (ftello(ofp) == 0)
332			break;
333	} while (n > 0);
334	if (fseek(ofp, nread - i + 1, SEEK_CUR) != 0)
335		err(1, "%s", currfile);
336
337	/*
338	 * getline() will read from here. Next call will truncate to
339	 * truncofs in this file.
340	 */
341	overfile = ofp;
342	truncofs = ftello(overfile);
343}
344
345/* Handle splits for /regexp/ and %regexp% patterns. */
346void
347do_rexp(const char *expr)
348{
349	regex_t cre;
350	intmax_t nwritten;
351	long ofs;
352	int first;
353	char *ecopy, *ep, *p, *pofs, *re;
354	FILE *ofp;
355
356	if ((ecopy = strdup(expr)) == NULL)
357		err(1, "strdup");
358
359	re = ecopy + 1;
360	if ((pofs = strrchr(ecopy, *expr)) == NULL || pofs[-1] == '\\')
361		errx(1, "%s: missing trailing %c", expr, *expr);
362	*pofs++ = '\0';
363
364	if (*pofs != '\0') {
365		errno = 0;
366		ofs = strtol(pofs, &ep, 10);
367		if (*ep != '\0' || errno != 0)
368			errx(1, "%s: bad offset", pofs);
369	} else
370		ofs = 0;
371
372	if (regcomp(&cre, re, REG_BASIC|REG_NOSUB) != 0)
373		errx(1, "%s: bad regular expression", re);
374
375	if (*expr == '/')
376		/* /regexp/: Save results to a file. */
377		ofp = newfile();
378	else {
379		/* %regexp%: Make a temporary file for overflow. */
380		if ((ofp = tmpfile()) == NULL)
381			err(1, "tmpfile");
382	}
383
384	/* Read and output lines until we get a match. */
385	first = 1;
386	while ((p = getline()) != NULL) {
387		if (fputs(p, ofp) != 0)
388			break;
389		if (!first && regexec(&cre, p, 0, NULL, 0) == 0)
390			break;
391		first = 0;
392	}
393
394	if (p == NULL)
395		errx(1, "%s: no match", re);
396
397	if (ofs <= 0) {
398		/*
399		 * Negative (or zero) offset: throw back any lines we should
400		 * not have read yet.
401		  */
402		if (p != NULL) {
403			toomuch(ofp, -ofs + 1);
404			nwritten = (intmax_t)truncofs;
405		} else
406			nwritten = (intmax_t)ftello(ofp);
407	} else {
408		/*
409		 * Positive offset: copy the requested number of lines
410		 * after the match.
411		 */
412		while (--ofs > 0 && (p = getline()) != NULL)
413			fputs(p, ofp);
414		toomuch(NULL, 0);
415		nwritten = (intmax_t)ftello(ofp);
416		if (fclose(ofp) != 0)
417			err(1, "%s", currfile);
418	}
419
420	if (!sflag && *expr == '/')
421		printf("%jd\n", nwritten);
422
423	regfree(&cre);
424	free(ecopy);
425}
426
427/* Handle splits based on line number. */
428void
429do_lineno(const char *expr)
430{
431	long lastline, tgtline;
432	char *ep, *p;
433	FILE *ofp;
434
435	errno = 0;
436	tgtline = strtol(expr, &ep, 10);
437	if (tgtline <= 0 || errno != 0 || *ep != '\0')
438		errx(1, "%s: bad line number", expr);
439	lastline = tgtline;
440	if (lastline <= lineno)
441		errx(1, "%s: can't go backwards", expr);
442
443	while (nfiles < maxfiles - 1) {
444		ofp = newfile();
445		while (lineno + 1 != lastline) {
446			if ((p = getline()) == NULL)
447				errx(1, "%ld: out of range", lastline);
448			if (fputs(p, ofp) != 0)
449				break;
450		}
451		if (!sflag)
452			printf("%jd\n", (intmax_t)ftello(ofp));
453		if (fclose(ofp) != 0)
454			err(1, "%s", currfile);
455		if (reps-- == 0)
456			break;
457		lastline += tgtline;
458	}
459}
460