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 <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 63 void cleanup(void); 64 void do_lineno(const char *); 65 void do_rexp(const char *); 66 char *getline(void); 67 void handlesig(int); 68 FILE *newfile(void); 69 void toomuch(FILE *, long); 70 void usage(void); 71 72 /* 73 * Command line options 74 */ 75 const char *prefix; /* File name prefix */ 76 long sufflen; /* Number of decimal digits for suffix */ 77 int sflag; /* Suppress output of file names */ 78 int kflag; /* Keep output if error occurs */ 79 80 /* 81 * Other miscellaneous globals (XXX too many) 82 */ 83 long lineno; /* Current line number in input file */ 84 long reps; /* Number of repetitions for this pattern */ 85 long nfiles; /* Number of files output so far */ 86 long maxfiles; /* Maximum number of files we can create */ 87 char currfile[PATH_MAX]; /* Current output file */ 88 const char *infn; /* Name of the input file */ 89 FILE *infile; /* Input file handle */ 90 FILE *overfile; /* Overflow file for toomuch() */ 91 off_t truncofs; /* Offset this file should be truncated at */ 92 int doclean; /* Should cleanup() remove output? */ 93 94 int 95 main(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 } 204 205 void 206 usage(void) 207 { 208 209 fprintf(stderr, 210 "usage: csplit [-ks] [-f prefix] [-n number] file args ...\n"); 211 exit(1); 212 } 213 214 void 215 handlesig(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. */ 225 FILE * 226 newfile(void) 227 { 228 FILE *fp; 229 230 if (snprintf(currfile, sizeof(currfile), "%s%0*ld", prefix, 231 (int)sufflen, nfiles) >= sizeof(currfile)) { 232 errno = ENAMETOOLONG; 233 err(1, NULL); 234 } 235 if ((fp = fopen(currfile, "w+")) == NULL) 236 err(1, "%s", currfile); 237 nfiles++; 238 239 return (fp); 240 } 241 242 /* Remove partial output, called before exiting. */ 243 void 244 cleanup(void) 245 { 246 char fnbuf[PATH_MAX]; 247 long i; 248 249 if (!doclean) 250 return; 251 252 /* 253 * NOTE: One cannot portably assume to be able to call snprintf() 254 * from inside a signal handler. It does, however, appear to be safe 255 * to do on FreeBSD. The solution to this problem is worse than the 256 * problem itself. 257 */ 258 259 for (i = 0; i < nfiles; i++) { 260 snprintf(fnbuf, sizeof(fnbuf), "%s%0*ld", prefix, 261 (int)sufflen, i); 262 unlink(fnbuf); 263 } 264 } 265 266 /* Read a line from the input into a static buffer. */ 267 char * 268 getline(void) 269 { 270 static char lbuf[LINE_MAX]; 271 FILE *src; 272 273 src = overfile != NULL ? overfile : infile; 274 275 again: if (fgets(lbuf, sizeof(lbuf), src) == NULL) { 276 if (src == overfile) { 277 src = infile; 278 goto again; 279 } 280 return (NULL); 281 } 282 if (ferror(src)) 283 err(1, "%s", infn); 284 lineno++; 285 286 return (lbuf); 287 } 288 289 /* Conceptually rewind the input (as obtained by getline()) back `n' lines. */ 290 void 291 toomuch(FILE *ofp, long n) 292 { 293 char buf[BUFSIZ]; 294 size_t i, nread; 295 296 if (overfile != NULL) { 297 /* 298 * Truncate the previous file we overflowed into back to 299 * the correct length, close it. 300 */ 301 if (fflush(overfile) != 0) 302 err(1, "overflow"); 303 if (ftruncate(fileno(overfile), truncofs) != 0) 304 err(1, "overflow"); 305 if (fclose(overfile) != 0) 306 err(1, "overflow"); 307 overfile = NULL; 308 } 309 310 if (n == 0) 311 /* Just tidying up */ 312 return; 313 314 lineno -= n; 315 316 /* 317 * Wind the overflow file backwards to `n' lines before the 318 * current one. 319 */ 320 do { 321 if (ftello(ofp) < (off_t)sizeof(buf)) 322 rewind(ofp); 323 else 324 fseek(ofp, -(long)sizeof(buf), SEEK_CUR); 325 if (ferror(ofp)) 326 errx(1, "%s: can't seek", currfile); 327 if ((nread = fread(buf, 1, sizeof(buf), ofp)) == 0) 328 errx(1, "can't read overflowed output"); 329 if (fseek(ofp, -(long)nread, SEEK_CUR) != 0) 330 err(1, "%s", currfile); 331 for (i = 1; i <= nread; i++) 332 if (buf[nread - i] == '\n' && n-- == 0) 333 break; 334 if (ftello(ofp) == 0) 335 break; 336 } while (n > 0); 337 if (fseek(ofp, nread - i + 1, SEEK_CUR) != 0) 338 err(1, "%s", currfile); 339 340 /* 341 * getline() will read from here. Next call will truncate to 342 * truncofs in this file. 343 */ 344 overfile = ofp; 345 truncofs = ftello(overfile); 346 } 347 348 /* Handle splits for /regexp/ and %regexp% patterns. */ 349 void 350 do_rexp(const char *expr) 351 { 352 regex_t cre; 353 intmax_t nwritten; 354 long ofs; 355 int first; 356 char *ecopy, *ep, *p, *pofs, *re; 357 FILE *ofp; 358 359 if ((ecopy = strdup(expr)) == NULL) 360 err(1, "strdup"); 361 362 re = ecopy + 1; 363 if ((pofs = strrchr(ecopy, *expr)) == NULL || pofs[-1] == '\\') 364 errx(1, "%s: missing trailing %c", expr, *expr); 365 *pofs++ = '\0'; 366 367 if (*pofs != '\0') { 368 errno = 0; 369 ofs = strtol(pofs, &ep, 10); 370 if (*ep != '\0' || errno != 0) 371 errx(1, "%s: bad offset", pofs); 372 } else 373 ofs = 0; 374 375 if (regcomp(&cre, re, REG_BASIC|REG_NOSUB) != 0) 376 errx(1, "%s: bad regular expression", re); 377 378 if (*expr == '/') 379 /* /regexp/: Save results to a file. */ 380 ofp = newfile(); 381 else { 382 /* %regexp%: Make a temporary file for overflow. */ 383 if ((ofp = tmpfile()) == NULL) 384 err(1, "tmpfile"); 385 } 386 387 /* Read and output lines until we get a match. */ 388 first = 1; 389 while ((p = getline()) != NULL) { 390 if (fputs(p, ofp) != 0) 391 break; 392 if (!first && regexec(&cre, p, 0, NULL, 0) == 0) 393 break; 394 first = 0; 395 } 396 397 if (p == NULL) 398 errx(1, "%s: no match", re); 399 400 if (ofs <= 0) { 401 /* 402 * Negative (or zero) offset: throw back any lines we should 403 * not have read yet. 404 */ 405 if (p != NULL) { 406 toomuch(ofp, -ofs + 1); 407 nwritten = (intmax_t)truncofs; 408 } else 409 nwritten = (intmax_t)ftello(ofp); 410 } else { 411 /* 412 * Positive offset: copy the requested number of lines 413 * after the match. 414 */ 415 while (--ofs > 0 && (p = getline()) != NULL) 416 fputs(p, ofp); 417 toomuch(NULL, 0); 418 nwritten = (intmax_t)ftello(ofp); 419 if (fclose(ofp) != 0) 420 err(1, "%s", currfile); 421 } 422 423 if (!sflag && *expr == '/') 424 printf("%jd\n", nwritten); 425 426 regfree(&cre); 427 free(ecopy); 428 } 429 430 /* Handle splits based on line number. */ 431 void 432 do_lineno(const char *expr) 433 { 434 long lastline, tgtline; 435 char *ep, *p; 436 FILE *ofp; 437 438 errno = 0; 439 tgtline = strtol(expr, &ep, 10); 440 if (tgtline <= 0 || errno != 0 || *ep != '\0') 441 errx(1, "%s: bad line number", expr); 442 lastline = tgtline; 443 if (lastline <= lineno) 444 errx(1, "%s: can't go backwards", expr); 445 446 while (nfiles < maxfiles - 1) { 447 ofp = newfile(); 448 while (lineno + 1 != lastline) { 449 if ((p = getline()) == NULL) 450 errx(1, "%ld: out of range", lastline); 451 if (fputs(p, ofp) != 0) 452 break; 453 } 454 if (!sflag) 455 printf("%jd\n", (intmax_t)ftello(ofp)); 456 if (fclose(ofp) != 0) 457 err(1, "%s", currfile); 458 if (reps-- == 0) 459 break; 460 lastline += tgtline; 461 } 462 } 463