2 * Copyright (c) 1991, 1993, 1994
3 * The Regents of the University of California. All rights reserved.
5 * This code is derived from software contributed to Berkeley by
6 * Steve Hayman of the Computer Science Department, Indiana University,
7 * Michiro Hikida and David Goodenough.
9 * Redistribution and use in source and binary forms, with or without
10 * modification, are permitted provided that the following conditions
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.
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
37 * @(#) Copyright (c) 1991, 1993, 1994 The Regents of the University of California. All rights reserved.
38 * @(#)join.c 8.6 (Berkeley) 5/4/95
39 * $FreeBSD: src/usr.bin/join/join.c,v 1.10.2.1 2002/06/18 05:14:49 jmallett Exp $
40 * $DragonFly: src/usr.bin/join/join.c,v 1.2 2003/06/17 04:29:27 dillon Exp $
43 #include <sys/param.h>
54 * There's a structure per input file which encapsulates the state of the
55 * file. We repeatedly read lines from each file until we've read in all
56 * the consecutive lines from the file with a common join field. Then we
57 * compare the set of lines with an equivalent set from the other file.
60 char *line; /* line */
61 u_long linealloc; /* line allocated count */
62 char **fields; /* line field(s) */
63 u_long fieldcnt; /* line field(s) count */
64 u_long fieldalloc; /* line field(s) allocated count */
68 FILE *fp; /* file descriptor */
69 u_long joinf; /* join field (-1, -2, -j) */
70 int unpair; /* output unpairable lines (-a) */
71 int number; /* 1 for file 1, 2 for file 2 */
73 LINE *set; /* set of lines with same field */
74 int pushbool; /* if pushback is set */
75 u_long pushback; /* line on the stack */
76 u_long setcnt; /* set count */
77 u_long setalloc; /* set allocated count */
79 INPUT input1 = { NULL, 0, 0, 1, NULL, 0, 0, 0, 0 },
80 input2 = { NULL, 0, 0, 2, NULL, 0, 0, 0, 0 };
83 u_long filenum; /* file number */
84 u_long fieldno; /* field number */
86 OLIST *olist; /* output field list */
87 u_long olistcnt; /* output field list count */
88 u_long olistalloc; /* output field allocated count */
90 int joinout = 1; /* show lines with matched join fields (-v) */
91 int needsep; /* need separator character */
92 int spans = 1; /* span multiple delimiters (-t) */
93 char *empty; /* empty field replacement string (-e) */
94 static char default_tabchar[] = " \t";
95 char *tabchar = default_tabchar;/* delimiter characters (-t) */
97 int cmp(LINE *, u_long, LINE *, u_long);
98 void fieldarg(char *);
99 void joinlines(INPUT *, INPUT *);
100 void obsolete(char **);
101 void outfield(LINE *, u_long, int);
102 void outoneline(INPUT *, LINE *);
103 void outtwoline(INPUT *, LINE *, INPUT *, LINE *);
113 int aflag, ch, cval, vflag;
116 setlocale(LC_ALL, "");
123 while ((ch = getopt(argc, argv, "\01a:e:j:1:2:o:t:v:")) != -1) {
125 case '\01': /* See comment in obsolete(). */
127 F1->unpair = F2->unpair = 1;
130 if ((F1->joinf = strtol(optarg, &end, 10)) < 1)
131 errx(1, "-1 option field number less than 1");
133 errx(1, "illegal field number -- %s", optarg);
137 if ((F2->joinf = strtol(optarg, &end, 10)) < 1)
138 errx(1, "-2 option field number less than 1");
140 errx(1, "illegal field number -- %s", optarg);
145 switch(strtol(optarg, &end, 10)) {
153 errx(1, "-a option file number not 1 or 2");
157 errx(1, "illegal file number -- %s", optarg);
163 if ((F1->joinf = F2->joinf =
164 strtol(optarg, &end, 10)) < 1)
165 errx(1, "-j option field number less than 1");
167 errx(1, "illegal field number -- %s", optarg);
176 if (strlen(tabchar = optarg) != 1)
177 errx(1, "illegal tab character specification");
182 switch (strtol(optarg, &end, 10)) {
190 errx(1, "-v option file number not 1 or 2");
194 errx(1, "illegal file number -- %s", optarg);
205 errx(1, "the -a and -v options are mutually exclusive");
210 /* Open the files; "-" means stdin. */
211 if (!strcmp(*argv, "-"))
213 else if ((F1->fp = fopen(*argv, "r")) == NULL)
216 if (!strcmp(*argv, "-"))
218 else if ((F2->fp = fopen(*argv, "r")) == NULL)
220 if (F1->fp == stdin && F2->fp == stdin)
221 errx(1, "only one input file may be stdin");
225 while (F1->setcnt && F2->setcnt) {
226 cval = cmp(F1->set, F1->joinf, F2->set, F2->joinf);
228 /* Oh joy, oh rapture, oh beauty divine! */
233 } else if (cval < 0) {
234 /* File 1 takes the lead... */
239 /* File 2 takes the lead... */
247 * Now that one of the files is used up, optionally output any
248 * remaining lines from the other file.
267 LINE *lp, *lastlp, tmp;
273 * Read all of the lines from an input file that have the same
277 for (lastlp = NULL;; ++F->setcnt) {
279 * If we're out of space to hold line structures, allocate
280 * more. Initialize the structure so that we know that this
283 if (F->setcnt == F->setalloc) {
286 if ((F->set = realloc(F->set,
287 F->setalloc * sizeof(LINE))) == NULL)
289 memset(F->set + cnt, 0, 50 * sizeof(LINE));
291 /* re-set lastlp in case it moved */
293 lastlp = &F->set[F->setcnt - 1];
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 space allocated to
300 * either structure. This could be avoided by doing another
301 * level of indirection, but it's probably okay as is.
303 lp = &F->set[F->setcnt];
305 lastlp = &F->set[F->setcnt - 1];
307 tmp = F->set[F->setcnt];
308 F->set[F->setcnt] = F->set[F->pushback];
309 F->set[F->pushback] = tmp;
313 if ((bp = fgetln(F->fp, &len)) == NULL)
315 if (lp->linealloc <= len + 1) {
316 lp->linealloc += MAX(100, len + 1 - lp->linealloc);
318 realloc(lp->line, lp->linealloc)) == NULL)
321 memmove(lp->line, bp, len);
323 /* Replace trailing newline, if it exists. */
324 if (bp[len - 1] == '\n')
325 lp->line[len - 1] = '\0';
327 lp->line[len] = '\0';
330 /* Split the line into fields, allocate space as necessary. */
332 while ((fieldp = strsep(&bp, tabchar)) != NULL) {
333 if (spans && *fieldp == '\0')
335 if (lp->fieldcnt == lp->fieldalloc) {
336 lp->fieldalloc += 50;
337 if ((lp->fields = realloc(lp->fields,
338 lp->fieldalloc * sizeof(char *))) == NULL)
341 lp->fields[lp->fieldcnt++] = fieldp;
344 /* See if the join field value has changed. */
345 if (lastlp != NULL && cmp(lp, F->joinf, lastlp, F->joinf)) {
347 F->pushback = F->setcnt;
354 cmp(lp1, fieldno1, lp2, fieldno2)
356 u_long fieldno1, fieldno2;
358 if (lp1->fieldcnt <= fieldno1)
359 return (lp2->fieldcnt <= fieldno2 ? 0 : 1);
360 if (lp2->fieldcnt <= fieldno2)
362 return (strcoll(lp1->fields[fieldno1], lp2->fields[fieldno2]));
369 unsigned int cnt1, cnt2;
372 * Output the results of a join comparison. The output may be from
373 * either file 1 or file 2 (in which case the first argument is the
374 * file from which to output) or from both.
377 for (cnt1 = 0; cnt1 < F1->setcnt; ++cnt1)
378 outoneline(F1, &F1->set[cnt1]);
381 for (cnt1 = 0; cnt1 < F1->setcnt; ++cnt1)
382 for (cnt2 = 0; cnt2 < F2->setcnt; ++cnt2)
383 outtwoline(F1, &F1->set[cnt1], F2, &F2->set[cnt2]);
394 * Output a single line from one of the files, according to the
395 * join rules. This happens when we are writing unmatched single
396 * lines. Output empty fields in the right places.
399 for (cnt = 0; cnt < olistcnt; ++cnt) {
400 if (olist[cnt].filenum == (unsigned)F->number)
401 outfield(lp, olist[cnt].fieldno, 0);
402 else if (olist[cnt].filenum == 0)
403 outfield(lp, F->joinf, 0);
408 for (cnt = 0; cnt < lp->fieldcnt; ++cnt)
409 outfield(lp, cnt, 0);
417 outtwoline(F1, lp1, F2, lp2)
423 /* Output a pair of lines according to the join list (if any). */
425 for (cnt = 0; cnt < olistcnt; ++cnt)
426 if (olist[cnt].filenum == 0) {
427 if (lp1->fieldcnt >= F1->joinf)
428 outfield(lp1, F1->joinf, 0);
430 outfield(lp2, F2->joinf, 0);
431 } else if (olist[cnt].filenum == 1)
432 outfield(lp1, olist[cnt].fieldno, 0);
433 else /* if (olist[cnt].filenum == 2) */
434 outfield(lp2, olist[cnt].fieldno, 0);
437 * Output the join field, then the remaining fields from F1
440 outfield(lp1, F1->joinf, 0);
441 for (cnt = 0; cnt < lp1->fieldcnt; ++cnt)
442 if (F1->joinf != cnt)
443 outfield(lp1, cnt, 0);
444 for (cnt = 0; cnt < lp2->fieldcnt; ++cnt)
445 if (F2->joinf != cnt)
446 outfield(lp2, cnt, 0);
455 outfield(lp, fieldno, out_empty)
461 (void)printf("%c", *tabchar);
462 if (!ferror(stdout)) {
463 if (lp->fieldcnt <= fieldno || out_empty) {
465 (void)printf("%s", empty);
467 if (*lp->fields[fieldno] == '\0')
469 (void)printf("%s", lp->fields[fieldno]);
477 * Convert an output list argument "2.1, 1.3, 2.4" into an array of output
484 u_long fieldno, filenum;
487 while ((token = strsep(&option, ", \t")) != NULL) {
491 filenum = fieldno = 0;
492 else if ((token[0] == '1' || token[0] == '2') &&
494 filenum = token[0] - '0';
495 fieldno = strtol(token + 2, &end, 10);
497 errx(1, "malformed -o option field");
499 errx(1, "field numbers are 1 based");
502 errx(1, "malformed -o option field");
503 if (olistcnt == olistalloc) {
505 if ((olist = realloc(olist,
506 olistalloc * sizeof(OLIST))) == NULL)
509 olist[olistcnt].filenum = filenum;
510 olist[olistcnt].fieldno = fieldno;
522 while ((ap = *++argv) != NULL) {
523 /* Return if "--". */
524 if (ap[0] == '-' && ap[1] == '-')
526 /* skip if not an option */
532 * The original join allowed "-a", which meant the
533 * same as -a1 plus -a2. POSIX 1003.2, Draft 11.2
534 * only specifies this as "-a 1" and "a -2", so we
535 * have to use another option flag, one that is
536 * unlikely to ever be used or accidentally entered
537 * on the command line. (Well, we could reallocate
538 * the argv array, but that hardly seems worthwhile.)
540 if (ap[2] == '\0' && (argv[1] == NULL ||
541 (strcmp(argv[1], "1") != 0 &&
542 strcmp(argv[1], "2") != 0))) {
544 warnx("-a option used without an argument; "
545 "reverting to historical behavior");
550 * The original join allowed "-j[12] arg" and "-j arg".
551 * Convert the former to "-[12] arg". Don't convert
552 * the latter since getopt(3) can handle it.
570 jbad: errx(1, "illegal option -- %s", ap);
576 * The original join allowed "-o arg arg".
577 * Convert to "-o arg -o arg".
581 for (p = argv + 2; *p; ++p) {
582 if (p[0][0] == '0' || (p[0][0] != '1' &&
583 p[0][0] != '2' || p[0][1] != '.'))
586 if (len - 2 != strspn(*p + 2, "0123456789"))
588 if ((t = malloc(len + 3)) == NULL)
592 memmove(t + 2, *p, len + 1);
604 (void)fprintf(stderr, "%s %s\n%s\n",
605 "usage: join [-a fileno | -v fileno ] [-e string] [-1 field]",
607 " [-o list] [-t char] file1 file2");