1 /*-
2 * SPDX-License-Identifier: BSD-3-Clause
3 *
4 * Copyright (c) 1987, 1993, 1994
5 * The Regents of the University of California. All rights reserved.
6 *
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
9 * are met:
10 * 1. Redistributions of source code must retain the above copyright
11 * notice, this list of conditions and the following disclaimer.
12 * 2. Redistributions in binary form must reproduce the above copyright
13 * notice, this list of conditions and the following disclaimer in the
14 * documentation and/or other materials provided with the distribution.
15 * 3. Neither the name of the University nor the names of its contributors
16 * may be used to endorse or promote products derived from this software
17 * without specific prior written permission.
18 *
19 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29 * SUCH DAMAGE.
30 */
31
32 #include <sys/cdefs.h>
33 #ifndef lint
34 static const char copyright[] =
35 "@(#) Copyright (c) 1987, 1993, 1994\n\
36 The Regents of the University of California. All rights reserved.\n";
37 #endif
38
39 #ifndef lint
40 static const char sccsid[] = "@(#)split.c 8.2 (Berkeley) 4/16/94";
41 #endif
42
43 #include <sys/param.h>
44 #include <sys/stat.h>
45
46 #include <ctype.h>
47 #include <err.h>
48 #include <errno.h>
49 #include <fcntl.h>
50 #include <inttypes.h>
51 #include <libutil.h>
52 #include <limits.h>
53 #include <locale.h>
54 #include <stdbool.h>
55 #include <stdint.h>
56 #include <stdio.h>
57 #include <stdlib.h>
58 #include <string.h>
59 #include <unistd.h>
60 #include <regex.h>
61 #include <sysexits.h>
62
63 #define DEFLINE 1000 /* Default num lines per file. */
64
65 static off_t bytecnt; /* Byte count to split on. */
66 static long chunks; /* Chunks count to split into. */
67 static bool clobber = true; /* Whether to overwrite existing output files. */
68 static long numlines; /* Line count to split on. */
69 static int file_open; /* If a file open. */
70 static int ifd = -1, ofd = -1; /* Input/output file descriptors. */
71 static char fname[MAXPATHLEN]; /* File name prefix. */
72 static regex_t rgx;
73 static int pflag;
74 static bool dflag;
75 static long sufflen = 2; /* File name suffix length. */
76 static bool autosfx = true; /* Whether to auto-extend the suffix length. */
77
78 static void newfile(void);
79 static void split1(void);
80 static void split2(void);
81 static void split3(void);
82 static void usage(void) __dead2;
83
84 int
main(int argc,char ** argv)85 main(int argc, char **argv)
86 {
87 char errbuf[64];
88 const char *p, *errstr;
89 int ch, error;
90
91 setlocale(LC_ALL, "");
92
93 dflag = false;
94 while ((ch = getopt(argc, argv, "0::1::2::3::4::5::6::7::8::9::a:b:cdl:n:p:")) != -1)
95 switch (ch) {
96 case '0': case '1': case '2': case '3': case '4':
97 case '5': case '6': case '7': case '8': case '9':
98 /*
99 * Undocumented kludge: split was originally designed
100 * to take a number after a dash.
101 */
102 if (numlines != 0)
103 usage();
104 numlines = ch - '0';
105 p = optarg ? optarg : "";
106 while (numlines >= 0 && *p >= '0' && *p <= '9')
107 numlines = numlines * 10 + *p++ - '0';
108 if (numlines <= 0 || *p != '\0')
109 errx(EX_USAGE, "%c%s: line count is invalid",
110 ch, optarg ? optarg : "");
111 break;
112 case 'a': /* Suffix length */
113 sufflen = strtonum(optarg, 0, INT_MAX, &errstr);
114 if (errstr != NULL) {
115 errx(EX_USAGE, "%s: suffix length is %s",
116 optarg, errstr);
117 }
118 if (sufflen == 0) {
119 sufflen = 2;
120 autosfx = true;
121 } else {
122 autosfx = false;
123 }
124 break;
125 case 'b': /* Byte count. */
126 if (expand_number(optarg, &bytecnt) != 0) {
127 errx(EX_USAGE, "%s: byte count is invalid",
128 optarg);
129 }
130 break;
131 case 'c': /* Continue, don't overwrite output files. */
132 clobber = false;
133 break;
134 case 'd': /* Decimal suffix */
135 dflag = true;
136 break;
137 case 'l': /* Line count. */
138 if (numlines != 0)
139 usage();
140 numlines = strtonum(optarg, 1, LONG_MAX, &errstr);
141 if (errstr != NULL) {
142 errx(EX_USAGE, "%s: line count is %s",
143 optarg, errstr);
144 }
145 break;
146 case 'n': /* Chunks. */
147 chunks = strtonum(optarg, 1, LONG_MAX, &errstr);
148 if (errstr != NULL) {
149 errx(EX_USAGE, "%s: number of chunks is %s",
150 optarg, errstr);
151 }
152 break;
153
154 case 'p': /* pattern matching. */
155 error = regcomp(&rgx, optarg, REG_EXTENDED|REG_NOSUB);
156 if (error != 0) {
157 regerror(error, &rgx, errbuf, sizeof(errbuf));
158 errx(EX_USAGE, "%s: regex is invalid: %s",
159 optarg, errbuf);
160 }
161 pflag = 1;
162 break;
163 default:
164 usage();
165 }
166 argv += optind;
167 argc -= optind;
168
169 if (argc > 0) { /* Input file. */
170 if (strcmp(*argv, "-") == 0)
171 ifd = STDIN_FILENO;
172 else if ((ifd = open(*argv, O_RDONLY, 0)) < 0)
173 err(EX_NOINPUT, "%s", *argv);
174 ++argv;
175 --argc;
176 }
177 if (argc > 0) { /* File name prefix. */
178 if (strlcpy(fname, *argv, sizeof(fname)) >= sizeof(fname)) {
179 errx(EX_USAGE, "%s: file name prefix is too long",
180 *argv);
181 }
182 ++argv;
183 --argc;
184 }
185 if (argc > 0)
186 usage();
187
188 if (strlen(fname) + (unsigned long)sufflen >= sizeof(fname))
189 errx(EX_USAGE, "suffix is too long");
190 if (pflag && (numlines != 0 || bytecnt != 0 || chunks != 0))
191 usage();
192
193 if (numlines == 0)
194 numlines = DEFLINE;
195 else if (bytecnt != 0 || chunks != 0)
196 usage();
197
198 if (bytecnt != 0 && chunks != 0)
199 usage();
200
201 if (ifd == -1) /* Stdin by default. */
202 ifd = 0;
203
204 if (bytecnt != 0) {
205 split1();
206 exit (0);
207 } else if (chunks != 0) {
208 split3();
209 exit (0);
210 }
211 split2();
212 if (pflag)
213 regfree(&rgx);
214 exit(0);
215 }
216
217 /*
218 * split1 --
219 * Split the input by bytes.
220 */
221 static void
split1(void)222 split1(void)
223 {
224 static char bfr[MAXBSIZE];
225 off_t bcnt;
226 char *C;
227 ssize_t dist, len;
228 int nfiles;
229
230 nfiles = 0;
231
232 for (bcnt = 0;;)
233 switch ((len = read(ifd, bfr, sizeof(bfr)))) {
234 case 0:
235 exit(0);
236 case -1:
237 err(EX_IOERR, "read");
238 /* NOTREACHED */
239 default:
240 if (!file_open) {
241 if (chunks == 0 || nfiles < chunks) {
242 newfile();
243 nfiles++;
244 }
245 }
246 if (bcnt + len >= bytecnt) {
247 dist = bytecnt - bcnt;
248 if (write(ofd, bfr, dist) != dist)
249 err(EX_IOERR, "write");
250 len -= dist;
251 for (C = bfr + dist; len >= bytecnt;
252 len -= bytecnt, C += bytecnt) {
253 if (chunks == 0 || nfiles < chunks) {
254 newfile();
255 nfiles++;
256 }
257 if (write(ofd, C, bytecnt) != bytecnt)
258 err(EX_IOERR, "write");
259 }
260 if (len != 0) {
261 if (chunks == 0 || nfiles < chunks) {
262 newfile();
263 nfiles++;
264 }
265 if (write(ofd, C, len) != len)
266 err(EX_IOERR, "write");
267 } else {
268 file_open = 0;
269 }
270 bcnt = len;
271 } else {
272 bcnt += len;
273 if (write(ofd, bfr, len) != len)
274 err(EX_IOERR, "write");
275 }
276 }
277 }
278
279 /*
280 * split2 --
281 * Split the input by lines.
282 */
283 static void
split2(void)284 split2(void)
285 {
286 char *buf;
287 size_t bufsize;
288 ssize_t len;
289 long lcnt = 0;
290 FILE *infp;
291
292 buf = NULL;
293 bufsize = 0;
294
295 /* Stick a stream on top of input file descriptor */
296 if ((infp = fdopen(ifd, "r")) == NULL)
297 err(EX_NOINPUT, "fdopen");
298
299 /* Process input one line at a time */
300 while ((errno = 0, len = getline(&buf, &bufsize, infp)) > 0) {
301 /* Check if we need to start a new file */
302 if (pflag) {
303 regmatch_t pmatch;
304
305 pmatch.rm_so = 0;
306 pmatch.rm_eo = len - 1;
307 if (regexec(&rgx, buf, 0, &pmatch, REG_STARTEND) == 0)
308 newfile();
309 } else if (lcnt++ == numlines) {
310 newfile();
311 lcnt = 1;
312 }
313
314 /* Open output file if needed */
315 if (!file_open)
316 newfile();
317
318 /* Write out line */
319 if (write(ofd, buf, len) != len)
320 err(EX_IOERR, "write");
321 }
322
323 /* EOF or error? */
324 if ((len == -1 && errno != 0) || ferror(infp))
325 err(EX_IOERR, "read");
326 else
327 exit(0);
328 }
329
330 /*
331 * split3 --
332 * Split the input into specified number of chunks
333 */
334 static void
split3(void)335 split3(void)
336 {
337 struct stat sb;
338
339 if (fstat(ifd, &sb) == -1) {
340 err(1, "stat");
341 /* NOTREACHED */
342 }
343
344 if (chunks > sb.st_size) {
345 errx(1, "can't split into more than %d files",
346 (int)sb.st_size);
347 /* NOTREACHED */
348 }
349
350 bytecnt = sb.st_size / chunks;
351 split1();
352 }
353
354
355 /*
356 * newfile --
357 * Open a new output file.
358 */
359 static void
newfile(void)360 newfile(void)
361 {
362 long i, maxfiles, tfnum;
363 static long fnum;
364 static char *fpnt;
365 char beg, end;
366 int pattlen;
367 int flags = O_WRONLY | O_CREAT | O_TRUNC;
368
369 if (!clobber)
370 flags |= O_EXCL;
371
372 if (ofd == -1) {
373 if (fname[0] == '\0') {
374 fname[0] = 'x';
375 fpnt = fname + 1;
376 } else {
377 fpnt = fname + strlen(fname);
378 }
379 } else if (close(ofd) != 0)
380 err(1, "%s", fname);
381
382 again:
383 if (dflag) {
384 beg = '0';
385 end = '9';
386 }
387 else {
388 beg = 'a';
389 end = 'z';
390 }
391 pattlen = end - beg + 1;
392
393 /*
394 * If '-a' is not specified, then we automatically expand the
395 * suffix length to accomodate splitting all input. We do this
396 * by moving the suffix pointer (fpnt) forward and incrementing
397 * sufflen by one, thereby yielding an additional two characters
398 * and allowing all output files to sort such that 'cat *' yields
399 * the input in order. I.e., the order is '... xyy xyz xzaaa
400 * xzaab ... xzyzy, xzyzz, xzzaaaa, xzzaaab' and so on.
401 */
402 if (!dflag && autosfx && (fpnt[0] == 'y') &&
403 strspn(fpnt+1, "z") == strlen(fpnt+1)) {
404 /* Ensure the generated filenames will fit into the buffer. */
405 if (strlen(fname) + 2 >= sizeof(fname))
406 errx(EX_USAGE, "combined filenames would be too long");
407
408 fpnt = fname + strlen(fname) - sufflen;
409 fpnt[sufflen + 2] = '\0';
410 fpnt[0] = end;
411 fpnt[1] = beg;
412
413 /* Basename | Suffix
414 * before:
415 * x | yz
416 * after:
417 * xz | a.. */
418 fpnt++;
419 sufflen++;
420
421 /* Reset so we start back at all 'a's in our extended suffix. */
422 fnum = 0;
423 }
424
425 /* maxfiles = pattlen^sufflen, but don't use libm. */
426 for (maxfiles = 1, i = 0; i < sufflen; i++)
427 if (LONG_MAX / pattlen < maxfiles)
428 errx(EX_USAGE, "suffix is too long (max %ld)", i);
429 else
430 maxfiles *= pattlen;
431
432 if (fnum == maxfiles)
433 errx(EX_DATAERR, "too many files");
434
435 /* Generate suffix of sufflen letters */
436 tfnum = fnum;
437 i = sufflen - 1;
438 do {
439 fpnt[i] = tfnum % pattlen + beg;
440 tfnum /= pattlen;
441 } while (i-- > 0);
442 fpnt[sufflen] = '\0';
443
444 ++fnum;
445 if ((ofd = open(fname, flags, DEFFILEMODE)) < 0) {
446 if (!clobber && errno == EEXIST)
447 goto again;
448 err(EX_IOERR, "%s", fname);
449 }
450 file_open = 1;
451 }
452
453 static void
usage(void)454 usage(void)
455 {
456 (void)fprintf(stderr,
457 "usage: split [-cd] [-l line_count] [-a suffix_length] [file [prefix]]\n"
458 " split [-cd] -b byte_count[K|k|M|m|G|g] [-a suffix_length] [file [prefix]]\n"
459 " split [-cd] -n chunk_count [-a suffix_length] [file [prefix]]\n"
460 " split [-cd] -p pattern [-a suffix_length] [file [prefix]]\n");
461 exit(EX_USAGE);
462 }
463