join.c revision 1.4
1 /*- 2 * Copyright (c) 1991 The Regents of the University of California. 3 * All rights reserved. 4 * 5 * This code is derived from software contributed to Berkeley by 6 * Steve Hayman of Indiana University, Michiro Hikida and David 7 * Goodenough. 8 * 9 * Redistribution and use in source and binary forms, with or without 10 * modification, are permitted provided that the following conditions 11 * are met: 12 * 1. Redistributions of source code must retain the above copyright 13 * notice, this list of conditions and the following disclaimer. 14 * 2. Redistributions in binary form must reproduce the above copyright 15 * notice, this list of conditions and the following disclaimer in the 16 * documentation and/or other materials provided with the distribution. 17 * 3. All advertising materials mentioning features or use of this software 18 * must display the following acknowledgement: 19 * This product includes software developed by the University of 20 * California, Berkeley and its contributors. 21 * 4. Neither the name of the University nor the names of its contributors 22 * may be used to endorse or promote products derived from this software 23 * without specific prior written permission. 24 * 25 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 26 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 27 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 28 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 29 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 30 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 31 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 32 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 33 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 34 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 35 * SUCH DAMAGE. 36 */ 37 38 #ifndef lint 39 char copyright[] = 40 "@(#) Copyright (c) 1991 The Regents of the University of California.\n\ 41 All rights reserved.\n"; 42 #endif /* not lint */ 43 44 #ifndef lint 45 /*static char sccsid[] = "from: @(#)join.c 5.1 (Berkeley) 11/18/91";*/ 46 static char rcsid[] = "$Id: join.c,v 1.4 1993/11/15 09:30:52 cgd Exp $"; 47 #endif /* not lint */ 48 49 #include <sys/types.h> 50 #include <stdio.h> 51 #include <stdlib.h> 52 #include <string.h> 53 #include <ctype.h> 54 #include <errno.h> 55 56 /* 57 * There's a structure per input file which encapsulates the state of the 58 * file. We repeatedly read lines from each file until we've read in all 59 * the consecutive lines from the file with a common join field. Then we 60 * compare the set of lines with an equivalent set from the other file. 61 */ 62 typedef struct { 63 char *line; /* line */ 64 u_long linealloc; /* line allocated count */ 65 char **fields; /* line field(s) */ 66 u_long fieldcnt; /* line field(s) count */ 67 u_long fieldalloc; /* line field(s) allocated count */ 68 } LINE; 69 70 typedef struct { 71 FILE *fp; /* file descriptor */ 72 u_long joinf; /* join field (-1, -2, -j) */ 73 int unpair; /* output unpairable lines (-a) */ 74 int number; /* 1 for file 1, 2 for file 2 */ 75 76 LINE *set; /* set of lines with same field */ 77 u_long pushback; /* line on the stack */ 78 u_long setcnt; /* set count */ 79 u_long setalloc; /* set allocated count */ 80 } INPUT; 81 INPUT input1 = { NULL, 0, 0, 1, NULL, -1, 0, 0, }, 82 input2 = { NULL, 0, 0, 1, NULL, -1, 0, 0, }; 83 84 typedef struct { 85 u_long fileno; /* file number */ 86 u_long fieldno; /* field number */ 87 } OLIST; 88 OLIST *olist; /* output field list */ 89 u_long olistcnt; /* output field list count */ 90 u_long olistalloc; /* output field allocated count */ 91 92 int joinout = 1; /* show lines with matched join fields (-v) */ 93 int needsep; /* need separator character */ 94 int showusage = 1; /* show usage for usage err() calls */ 95 int spans = 1; /* span multiple delimiters (-t) */ 96 char *empty; /* empty field replacement string (-e) */ 97 char *tabchar = " \t"; /* delimiter characters (-t) */ 98 99 int cmp __P((LINE *, u_long, LINE *, u_long)); 100 void enomem __P((void)); 101 void err __P((const char *, ...)); 102 void fieldarg __P((char *)); 103 void joinlines __P((INPUT *, INPUT *)); 104 void obsolete __P((char **)); 105 void outfield __P((LINE *, u_long)); 106 void outoneline __P((INPUT *, LINE *)); 107 void outtwoline __P((INPUT *, LINE *, INPUT *, LINE *)); 108 void slurp __P((INPUT *)); 109 void usage __P((void)); 110 111 int 112 main(argc, argv) 113 int argc; 114 char *argv[]; 115 { 116 register INPUT *F1, *F2; 117 int aflag, ch, cval, vflag; 118 char *end; 119 120 F1 = &input1; 121 F2 = &input2; 122 123 aflag = vflag = 0; 124 obsolete(argv); 125 while ((ch = getopt(argc, argv, "\01a:e:j:1:2:o:t:v:")) != EOF) { 126 switch (ch) { 127 case '\01': 128 aflag = 1; 129 F1->unpair = F2->unpair = 1; 130 break; 131 case '1': 132 if ((F1->joinf = strtol(optarg, &end, 10)) < 1) 133 err("-1 option field number less than 1"); 134 if (*end) 135 err("illegal field number -- %s", optarg); 136 --F1->joinf; 137 break; 138 case '2': 139 if ((F2->joinf = strtol(optarg, &end, 10)) < 1) 140 err("-2 option field number less than 1"); 141 if (*end) 142 err("illegal field number -- %s", optarg); 143 --F2->joinf; 144 break; 145 case 'a': 146 aflag = 1; 147 switch(strtol(optarg, &end, 10)) { 148 case 1: 149 F1->unpair = 1; 150 break; 151 case 2: 152 F2->unpair = 1; 153 break; 154 default: 155 err("-a option file number not 1 or 2"); 156 break; 157 } 158 if (*end) 159 err("illegal file number -- %s", optarg); 160 break; 161 case 'e': 162 empty = optarg; 163 break; 164 case 'j': 165 if ((F1->joinf = F2->joinf = 166 strtol(optarg, &end, 10)) < 1) 167 err("-j option field number less than 1"); 168 if (*end) 169 err("illegal field number -- %s", optarg); 170 --F1->joinf; 171 --F2->joinf; 172 break; 173 case 'o': 174 fieldarg(optarg); 175 break; 176 case 't': 177 spans = 0; 178 if (strlen(tabchar = optarg) != 1) 179 err("illegal tab character specification"); 180 break; 181 case 'v': 182 vflag = 1; 183 joinout = 0; 184 switch(strtol(optarg, &end, 10)) { 185 case 1: 186 F1->unpair = 1; 187 break; 188 case 2: 189 F2->unpair = 1; 190 break; 191 default: 192 err("-v option file number not 1 or 2"); 193 break; 194 } 195 if (*end) 196 err("illegal file number -- %s", optarg); 197 break; 198 case '?': 199 default: 200 usage(); 201 } 202 } 203 argc -= optind; 204 argv += optind; 205 206 if (aflag && vflag) 207 err("-a and -v options mutually exclusive"); 208 209 if (argc != 2) 210 usage(); 211 showusage = 0; 212 213 /* Open the files; "-" means stdin. */ 214 if (!strcmp(*argv, "-")) 215 F1->fp = stdin; 216 else if ((F1->fp = fopen(*argv, "r")) == NULL) 217 err("%s: %s", *argv, strerror(errno)); 218 ++argv; 219 if (!strcmp(*argv, "-")) 220 F2->fp = stdin; 221 else if ((F2->fp = fopen(*argv, "r")) == NULL) 222 err("%s: %s", *argv, strerror(errno)); 223 if (F1->fp == stdin && F2->fp == stdin) 224 err("only one input file may be stdin"); 225 226 slurp(F1); 227 slurp(F2); 228 while (F1->setcnt && F2->setcnt) { 229 cval = cmp(F1->set, F1->joinf, F2->set, F2->joinf); 230 if (cval == 0) { 231 /* Oh joy, oh rapture, oh beauty divine! */ 232 if (joinout) 233 joinlines(F1, F2); 234 slurp(F1); 235 slurp(F2); 236 } else if (cval < 0) { 237 /* File 1 takes the lead... */ 238 if (F1->unpair) 239 joinlines(F1, NULL); 240 slurp(F1); 241 } else { 242 /* File 2 takes the lead... */ 243 if (F2->unpair) 244 joinlines(F2, NULL); 245 slurp(F2); 246 } 247 } 248 249 /* 250 * Now that one of the files is used up, optionally output any 251 * remaining lines from the other file. 252 */ 253 if (F1->unpair) 254 while (F1->setcnt) { 255 joinlines(F1, NULL); 256 slurp(F1); 257 } 258 if (F2->unpair) 259 while (F2->setcnt) { 260 joinlines(F2, NULL); 261 slurp(F2); 262 } 263 exit(0); 264 } 265 266 void 267 slurp(F) 268 INPUT *F; 269 { 270 register LINE *lp, *lastlp; 271 LINE tmp; 272 size_t len; 273 int cnt; 274 char *bp, *fieldp, *token; 275 276 /* 277 * Read all of the lines from an input file that have the same 278 * join field. 279 */ 280 F->setcnt = 0; 281 for (lastlp = NULL;; ++F->setcnt, lastlp = lp) { 282 /* 283 * If we're out of space to hold line structures, allocate 284 * more. Initialize the structure so that we know that this 285 * is new space. 286 */ 287 if (F->setcnt == F->setalloc) { 288 cnt = F->setalloc; 289 F->setalloc += 100; 290 if ((F->set = realloc(F->set, 291 F->setalloc * sizeof(LINE))) == NULL) 292 enomem(); 293 bzero(F->set + cnt, 100 * sizeof(LINE *)); 294 } 295 296 /* 297 * Get any pushed back line, else get the next line. Allocate 298 * space as necessary. If taking the line from the stack swap 299 * the two structures so that we don't lose the allocated space. 300 * This could be avoided by doing another level of indirection, 301 * but it's probably okay as is. 302 */ 303 lp = &F->set[F->setcnt]; 304 if (F->pushback != -1) { 305 tmp = F->set[F->setcnt]; 306 F->set[F->setcnt] = F->set[F->pushback]; 307 F->set[F->pushback] = tmp; 308 F->pushback = -1; 309 continue; 310 } 311 if ((bp = fgetline(F->fp, &len)) == NULL) 312 return; 313 if (lp->linealloc <= len) { 314 if (lp->linealloc == 0) 315 lp->linealloc = 128; 316 while (lp->linealloc <= len) 317 lp->linealloc *= 2; 318 319 if ((lp->line = realloc(lp->line, 320 lp->linealloc * sizeof(char))) == NULL) 321 enomem(); 322 } 323 bcopy(bp, lp->line, len+1); 324 325 /* Split the line into fields, allocate space as necessary. */ 326 token = lp->line; 327 lp->fieldcnt = 0; 328 while ((fieldp = strsep(&token, tabchar)) != NULL) { 329 if (spans && *fieldp == '\0') 330 continue; 331 if (lp->fieldcnt == lp->fieldalloc) { 332 lp->fieldalloc += 100; 333 if ((lp->fields = realloc(lp->fields, 334 lp->fieldalloc * sizeof(char *))) == NULL) 335 enomem(); 336 } 337 lp->fields[lp->fieldcnt++] = fieldp; 338 } 339 340 /* See if the join field value has changed. */ 341 if (lastlp != NULL && cmp(lp, F->joinf, lastlp, F->joinf)) { 342 F->pushback = F->setcnt; 343 break; 344 } 345 } 346 } 347 348 int 349 cmp(lp1, fieldno1, lp2, fieldno2) 350 LINE *lp1, *lp2; 351 u_long fieldno1, fieldno2; 352 { 353 if (lp1->fieldcnt < fieldno1) 354 return (lp2->fieldcnt < fieldno2 ? 0 : 1); 355 if (lp2->fieldcnt < fieldno2) 356 return (-1); 357 return (strcmp(lp1->fields[fieldno1], lp2->fields[fieldno2])); 358 } 359 360 void 361 joinlines(F1, F2) 362 register INPUT *F1, *F2; 363 { 364 register int cnt1, cnt2; 365 366 /* 367 * Output the results of a join comparison. The output may be from 368 * either file 1 or file 2 (in which case the first argument is the 369 * file from which to output) or from both. 370 */ 371 if (F2 == NULL) { 372 for (cnt1 = 0; cnt1 < F1->setcnt; ++cnt1) 373 outoneline(F1, &F1->set[cnt1]); 374 return; 375 } 376 for (cnt1 = 0; cnt1 < F1->setcnt; ++cnt1) 377 for (cnt2 = 0; cnt2 < F2->setcnt; ++cnt2) 378 outtwoline(F1, &F1->set[cnt1], F2, &F2->set[cnt2]); 379 } 380 381 void 382 outoneline(F, lp) 383 INPUT *F; 384 register LINE *lp; 385 { 386 register int cnt; 387 388 /* 389 * Output a single line from one of the files, according to the 390 * join rules. This happens when we are writing unmatched single 391 * lines. Output empty fields in the right places. 392 */ 393 if (olist) 394 for (cnt = 0; cnt < olistcnt; ++cnt) { 395 if (olist[cnt].fileno == F->number) 396 outfield(lp, olist[cnt].fieldno); 397 } 398 else 399 for (cnt = 0; cnt < lp->fieldcnt; ++cnt) 400 outfield(lp, cnt); 401 (void)printf("\n"); 402 if (ferror(stdout)) 403 err("stdout: %s", strerror(errno)); 404 needsep = 0; 405 } 406 407 void 408 outtwoline(F1, lp1, F2, lp2) 409 register INPUT *F1, *F2; 410 register LINE *lp1, *lp2; 411 { 412 register int cnt; 413 414 /* Output a pair of lines according to the join list (if any). */ 415 if (olist) 416 for (cnt = 0; cnt < olistcnt; ++cnt) 417 if (olist[cnt].fileno == 1) 418 outfield(lp1, olist[cnt].fieldno); 419 else /* if (olist[cnt].fileno == 2) */ 420 outfield(lp2, olist[cnt].fieldno); 421 else { 422 /* 423 * Output the join field, then the remaining fields from F1 424 * and F2. 425 */ 426 outfield(lp1, F1->joinf); 427 for (cnt = 0; cnt < lp1->fieldcnt; ++cnt) 428 if (F1->joinf != cnt) 429 outfield(lp1, cnt); 430 for (cnt = 0; cnt < lp2->fieldcnt; ++cnt) 431 if (F2->joinf != cnt) 432 outfield(lp2, cnt); 433 } 434 (void)printf("\n"); 435 if (ferror(stdout)) 436 err("stdout: %s", strerror(errno)); 437 needsep = 0; 438 } 439 440 void 441 outfield(lp, fieldno) 442 LINE *lp; 443 u_long fieldno; 444 { 445 if (needsep++) 446 (void)printf("%c", *tabchar); 447 if (!ferror(stdout)) 448 if (lp->fieldcnt < fieldno) { 449 if (empty != NULL) 450 (void)printf("%s", empty); 451 } else { 452 if (*lp->fields[fieldno] == '\0') 453 return; 454 (void)printf("%s", lp->fields[fieldno]); 455 } 456 if (ferror(stdout)) 457 err("stdout: %s", strerror(errno)); 458 } 459 460 /* 461 * Convert an output list argument "2.1, 1.3, 2.4" into an array of output 462 * fields. 463 */ 464 void 465 fieldarg(option) 466 char *option; 467 { 468 u_long fieldno; 469 char *end, *token; 470 471 while ((token = strsep(&option, " \t")) != NULL) { 472 if (*token == '\0') 473 continue; 474 if (token[0] != '1' && token[0] != '2' || token[1] != '.') 475 err("malformed -o option field"); 476 fieldno = strtol(token + 2, &end, 10); 477 if (*end) 478 err("malformed -o option field"); 479 if (fieldno == 0) 480 err("field numbers are 1 based"); 481 if (olistcnt == olistalloc) { 482 olistalloc += 50; 483 if ((olist = realloc(olist, 484 olistalloc * sizeof(OLIST))) == NULL) 485 enomem(); 486 } 487 olist[olistcnt].fileno = token[0] - '0'; 488 olist[olistcnt].fieldno = fieldno - 1; 489 ++olistcnt; 490 } 491 } 492 493 void 494 obsolete(argv) 495 char **argv; 496 { 497 int len; 498 char **p, *ap, *t; 499 500 while (ap = *++argv) { 501 /* Return if "--". */ 502 if (ap[0] == '-' && ap[1] == '-') 503 return; 504 switch (ap[1]) { 505 case 'a': 506 /* 507 * The original join allowed "-a", which meant the 508 * same as -a1 plus -a2. POSIX 1003.2, Draft 11.2 509 * only specifies this as "-a 1" and "a -2", so we 510 * have to use another option flag, one that is 511 * unlikely to ever be used or accidentally entered 512 * on the command line. (Well, we could reallocate 513 * the argv array, but that hardly seems worthwhile.) 514 */ 515 if (ap[2] == '\0') 516 ap[1] = '\01'; 517 break; 518 case 'j': 519 /* 520 * The original join allowed "-j[12] arg" and "-j arg". 521 * Convert the former to "-[12] arg". Don't convert 522 * the latter since getopt(3) can handle it. 523 */ 524 switch(ap[2]) { 525 case '1': 526 if (ap[3] != '\0') 527 goto jbad; 528 ap[1] = '1'; 529 ap[2] = '\0'; 530 break; 531 case '2': 532 if (ap[3] != '\0') 533 goto jbad; 534 ap[1] = '2'; 535 ap[2] = '\0'; 536 break; 537 case '\0': 538 break; 539 default: 540 jbad: err("illegal option -- %s", ap); 541 usage(); 542 } 543 break; 544 case 'o': 545 /* 546 * The original join allowed "-o arg arg". Convert to 547 * "-o arg -o arg". 548 */ 549 if (ap[2] != '\0') 550 break; 551 for (p = argv + 2; *p; ++p) { 552 if (p[0][0] != '1' && p[0][0] != '2' || 553 p[0][1] != '.') 554 break; 555 len = strlen(*p); 556 if (len - 2 != strspn(*p + 2, "0123456789")) 557 break; 558 if ((t = malloc(len + 3)) == NULL) 559 enomem(); 560 t[0] = '-'; 561 t[1] = 'o'; 562 bcopy(*p, t + 2, len + 1); 563 *p = t; 564 } 565 argv = p - 1; 566 break; 567 } 568 } 569 } 570 571 void 572 enomem() 573 { 574 showusage = 0; 575 err("%s", strerror(errno)); 576 } 577 578 void 579 usage() 580 { 581 (void)fprintf(stderr, "%s%s\n", 582 "usage: join [-a fileno | -v fileno ] [-e string] [-1 field] ", 583 "[-2 field]\n [-o list] [-t char] file1 file2"); 584 exit(1); 585 } 586 587 #if __STDC__ 588 #include <stdarg.h> 589 #else 590 #include <varargs.h> 591 #endif 592 593 void 594 #if __STDC__ 595 err(const char *fmt, ...) 596 #else 597 err(fmt, va_alist) 598 char *fmt; 599 va_dcl 600 #endif 601 { 602 va_list ap; 603 #if __STDC__ 604 va_start(ap, fmt); 605 #else 606 va_start(ap); 607 #endif 608 (void)fprintf(stderr, "join: "); 609 (void)vfprintf(stderr, fmt, ap); 610 va_end(ap); 611 (void)fprintf(stderr, "\n"); 612 if (showusage) 613 usage(); 614 exit(1); 615 /* NOTREACHED */ 616 } 617
Indexes created Fri Sep 26 20:09:58 GMT 2025