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("$FreeBSD$"); 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 <regex.h> 55 #include <signal.h> 56 #include <stdint.h> 57 #include <stdio.h> 58 #include <stdlib.h> 59 #include <string.h> 60 #include <unistd.h> 61 62 void cleanup(void); 63 void do_lineno(const char *); 64 void do_rexp(const char *); 65 char *getline(void); 66 void handlesig(int); 67 FILE *newfile(void); 68 void toomuch(FILE *, long); 69 void usage(void); 70 71 /* 72 * Command line options 73 */ 74 const char *prefix; /* File name prefix */ 75 long sufflen; /* Number of decimal digits for suffix */ 76 int sflag; /* Suppress output of file names */ 77 int kflag; /* Keep output if error occurs */ 78 79 /* 80 * Other miscellaneous globals (XXX too many) 81 */ 82 long lineno; /* Current line number in input file */ 83 long reps; /* Number of repetitions for this pattern */ 84 long nfiles; /* Number of files output so far */ 85 long maxfiles; /* Maximum number of files we can create */ 86 char currfile[PATH_MAX]; /* Current output file */ 87 const char *infn; /* Name of the input file */ 88 FILE *infile; /* Input file handle */ 89 FILE *overfile; /* Overflow file for toomuch() */ 90 off_t truncofs; /* Offset this file should be truncated at */ 91 int doclean; /* Should cleanup() remove output? */ 92 93 int 94 main(int argc, char *argv[]) 95 { 96 long i; 97 int ch; 98 const char *expr; 99 char *ep, *p; 100 FILE *ofp; 101 102 kflag = sflag = 0; 103 prefix = "xx"; 104 sufflen = 2; 105 while ((ch = getopt(argc, argv, "ksf:n:")) > 0) { 106 switch (ch) { 107 case 'f': 108 prefix = optarg; 109 break; 110 case 'k': 111 kflag = 1; 112 break; 113 case 'n': 114 errno = 0; 115 sufflen = strtol(optarg, &ep, 10); 116 if (sufflen <= 0 || *ep != '\0' || errno != 0) 117 errx(1, "%s: bad suffix length", optarg); 118 break; 119 case 's': 120 sflag = 1; 121 break; 122 default: 123 usage(); 124 /*NOTREACHED*/ 125 } 126 } 127 128 if (sufflen + strlen(prefix) >= PATH_MAX) 129 errx(1, "name too long"); 130 131 argc -= optind; 132 argv += optind; 133 134 if ((infn = *argv++) == NULL) 135 usage(); 136 if (strcmp(infn, "-") == 0) { 137 infile = stdin; 138 infn = "stdin"; 139 } else if ((infile = fopen(infn, "r")) == NULL) 140 err(1, "%s", infn); 141 142 if (!kflag) { 143 doclean = 1; 144 atexit(cleanup); 145 signal(SIGHUP, handlesig); 146 signal(SIGINT, handlesig); 147 signal(SIGTERM, handlesig); 148 } 149 150 lineno = 0; 151 nfiles = 0; 152 truncofs = 0; 153 overfile = NULL; 154 155 /* Ensure 10^sufflen < LONG_MAX. */ 156 for (maxfiles = 1, i = 0; i < sufflen; i++) { 157 if (maxfiles > LONG_MAX / 10) 158 errx(1, "%ld: suffix too long (limit %ld)", 159 sufflen, i); 160 maxfiles *= 10; 161 } 162 163 /* Create files based on supplied patterns. */ 164 while (nfiles < maxfiles - 1 && (expr = *argv++) != NULL) { 165 /* Look ahead & see if this pattern has any repetitions. */ 166 if (*argv != NULL && **argv == '{') { 167 errno = 0; 168 reps = strtol(*argv + 1, &ep, 10); 169 if (reps < 0 || *ep != '}' || errno != 0) 170 errx(1, "%s: bad repetition count", *argv + 1); 171 argv++; 172 } else 173 reps = 0; 174 175 if (*expr == '/' || *expr == '%') { 176 do 177 do_rexp(expr); 178 while (reps-- != 0 && nfiles < maxfiles - 1); 179 } else if (isdigit((unsigned char)*expr)) 180 do_lineno(expr); 181 else 182 errx(1, "%s: unrecognised pattern", expr); 183 } 184 185 /* Copy the rest into a new file. */ 186 if (!feof(infile)) { 187 ofp = newfile(); 188 while ((p = getline()) != NULL && fputs(p, ofp) == 0) 189 ; 190 if (!sflag) 191 printf("%jd\n", (intmax_t)ftello(ofp)); 192 if (fclose(ofp) != 0) 193 err(1, "%s", currfile); 194 } 195 196 toomuch(NULL, 0); 197 doclean = 0; 198 199 return (0); 200 } 201 202 void 203 usage(void) 204 { 205 206 fprintf(stderr, 207 "usage: csplit [-ks] [-f prefix] [-n number] file [args ...]\n"); 208 exit(1); 209 } 210 211 void 212 handlesig(int sig __unused) 213 { 214 const char msg[] = "csplit: caught signal, cleaning up\n"; 215 216 write(STDERR_FILENO, msg, sizeof(msg) - 1); 217 cleanup(); 218 _exit(2); 219 } 220 221 /* Create a new output file. */ 222 FILE * 223 newfile(void) 224 { 225 FILE *fp; 226 227 snprintf(currfile, sizeof(currfile), "%s%0*ld", prefix, (int)sufflen, 228 nfiles); 229 if ((fp = fopen(currfile, "w+")) == NULL) 230 err(1, "%s", currfile); 231 nfiles++; 232 233 return (fp); 234 } 235 236 /* Remove partial output, called before exiting. */ 237 void 238 cleanup(void) 239 { 240 char fnbuf[PATH_MAX]; 241 long i; 242 243 if (!doclean) 244 return; 245 246 /* 247 * NOTE: One cannot portably assume to be able to call snprintf() 248 * from inside a signal handler. It does, however, appear to be safe 249 * to do on FreeBSD. The solution to this problem is worse than the 250 * problem itself. 251 */ 252 253 for (i = 0; i < nfiles; i++) { 254 snprintf(fnbuf, sizeof(fnbuf), "%s%0*ld", prefix, 255 (int)sufflen, i); 256 unlink(fnbuf); 257 } 258 } 259 260 /* Read a line from the input into a static buffer. */ 261 char * 262 getline(void) 263 { 264 static char lbuf[LINE_MAX]; 265 FILE *src; 266 267 src = overfile != NULL ? overfile : infile; 268 269 again: if (fgets(lbuf, sizeof(lbuf), src) == NULL) { 270 if (src == overfile) { 271 src = infile; 272 goto again; 273 } 274 return (NULL); 275 } 276 if (ferror(src)) 277 err(1, "%s", infn); 278 lineno++; 279 280 return (lbuf); 281 } 282 283 /* Conceptually rewind the input (as obtained by getline()) back `n' lines. */ 284 void 285 toomuch(FILE *ofp, long n) 286 { 287 char buf[BUFSIZ]; 288 size_t i, nread; 289 290 if (overfile != NULL) { 291 /* 292 * Truncate the previous file we overflowed into back to 293 * the correct length, close it. 294 */ 295 if (fflush(overfile) != 0) 296 err(1, "overflow"); 297 if (ftruncate(fileno(overfile), truncofs) != 0) 298 err(1, "overflow"); 299 if (fclose(overfile) != 0) 300 err(1, "overflow"); 301 overfile = NULL; 302 } 303 304 if (n == 0) 305 /* Just tidying up */ 306 return; 307 308 lineno -= n; 309 310 /* 311 * Wind the overflow file backwards to `n' lines before the 312 * current one. 313 */ 314 do { 315 if (ftello(ofp) < (off_t)sizeof(buf)) 316 rewind(ofp); 317 else 318 fseek(ofp, -(long)sizeof(buf), SEEK_CUR); 319 if (ferror(ofp)) 320 errx(1, "%s: can't seek", currfile); 321 if ((nread = fread(buf, 1, sizeof(buf), ofp)) == 0) 322 errx(1, "can't read overflowed output"); 323 if (fseek(ofp, -(long)nread, SEEK_CUR) != 0) 324 err(1, "%s", currfile); 325 for (i = 1; i <= nread; i++) 326 if (buf[nread - i] == '\n' && n-- == 0) 327 break; 328 } while (n > 0); 329 if (fseek(ofp, nread - i + 1, SEEK_CUR) != 0) 330 err(1, "%s", currfile); 331 332 /* 333 * getline() will read from here. Next call will truncate to 334 * truncofs in this file. 335 */ 336 overfile = ofp; 337 truncofs = ftello(overfile); 338 } 339 340 /* Handle splits for /regexp/ and %regexp% patterns. */ 341 void 342 do_rexp(const char *expr) 343 { 344 regex_t cre; 345 intmax_t nwritten; 346 long ofs; 347 int first; 348 char *ecopy, *ep, *p, *pofs, *re; 349 FILE *ofp; 350 351 if ((ecopy = strdup(expr)) == NULL) 352 err(1, "strdup"); 353 354 re = ecopy + 1; 355 if ((pofs = strrchr(ecopy, *expr)) == NULL || pofs[-1] == '\\') 356 errx(1, "%s: missing trailing %c", expr, *expr); 357 *pofs++ = '\0'; 358 359 if (*pofs != '\0') { 360 errno = 0; 361 ofs = strtol(pofs, &ep, 10); 362 if (*ep != '\0' || errno != 0) 363 errx(1, "%s: bad offset", pofs); 364 } else 365 ofs = 0; 366 367 if (regcomp(&cre, re, REG_BASIC|REG_NOSUB) != 0) 368 errx(1, "%s: bad regular expression", re); 369 370 if (*expr == '/') 371 /* /regexp/: Save results to a file. */ 372 ofp = newfile(); 373 else { 374 /* %regexp%: Make a temporary file for overflow. */ 375 if ((ofp = tmpfile()) == NULL) 376 err(1, "tmpfile"); 377 } 378 379 /* Read and output lines until we get a match. */ 380 first = 1; 381 while ((p = getline()) != NULL) { 382 if (fputs(p, ofp) != 0) 383 break; 384 if (!first && regexec(&cre, p, 0, NULL, 0) == 0) 385 break; 386 first = 0; 387 } 388 389 if (p == NULL) 390 errx(1, "%s: no match", re); 391 392 if (ofs <= 0) { 393 /* 394 * Negative (or zero) offset: throw back any lines we should 395 * not have read yet. 396 */ 397 if (p != NULL) { 398 toomuch(ofp, -ofs + 1); 399 nwritten = (intmax_t)truncofs; 400 } else 401 nwritten = (intmax_t)ftello(ofp); 402 } else { 403 /* 404 * Positive offset: copy the requested number of lines 405 * after the match. 406 */ 407 while (--ofs > 0 && (p = getline()) != NULL) 408 fputs(p, ofp); 409 toomuch(NULL, 0); 410 nwritten = (intmax_t)ftello(ofp); 411 if (fclose(ofp) != 0) 412 err(1, "%s", currfile); 413 } 414 415 if (!sflag && *expr == '/') 416 printf("%jd\n", nwritten); 417 418 regfree(&cre); 419 free(ecopy); 420 } 421 422 /* Handle splits based on line number. */ 423 void 424 do_lineno(const char *expr) 425 { 426 long lastline, tgtline; 427 char *ep, *p; 428 FILE *ofp; 429 430 errno = 0; 431 tgtline = strtol(expr, &ep, 10); 432 if (tgtline <= 0 || errno != 0 || *ep != '\0') 433 errx(1, "%s: bad line number", expr); 434 lastline = tgtline; 435 if (lastline <= lineno) 436 errx(1, "%s: can't go backwards", expr); 437 438 while (nfiles < maxfiles - 1) { 439 ofp = newfile(); 440 while (lineno + 1 != lastline) { 441 if ((p = getline()) == NULL) 442 errx(1, "%ld: out of range", lastline); 443 if (fputs(p, ofp) != 0) 444 break; 445 } 446 if (!sflag) 447 printf("%jd\n", (intmax_t)ftello(ofp)); 448 if (fclose(ofp) != 0) 449 err(1, "%s", currfile); 450 if (reps-- == 0) 451 break; 452 lastline += tgtline; 453 } 454 } 455