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
39 static char copyright[] =
40 "@(#) Copyright (c) 1991, 1993, 1994\n\
41 The Regents of the University of California. All rights reserved.\n";
45 static char sccsid[] = "@(#)join.c 8.3 (Berkeley) 4/16/94";
48 #include <sys/param.h>
58 * There's a structure per input file which encapsulates the state of the
59 * file. We repeatedly read lines from each file until we've read in all
60 * the consecutive lines from the file with a common join field. Then we
61 * compare the set of lines with an equivalent set from the other file.
64 char *line; /* line */
65 u_long linealloc; /* line allocated count */
66 char **fields; /* line field(s) */
67 u_long fieldcnt; /* line field(s) count */
68 u_long fieldalloc; /* line field(s) allocated count */
72 FILE *fp; /* file descriptor */
73 u_long joinf; /* join field (-1, -2, -j) */
74 int unpair; /* output unpairable lines (-a) */
75 int number; /* 1 for file 1, 2 for file 2 */
77 LINE *set; /* set of lines with same field */
78 int pushbool; /* if pushback is set */
79 u_long pushback; /* line on the stack */
80 u_long setcnt; /* set count */
81 u_long setalloc; /* set allocated count */
83 INPUT input1 = { NULL, 0, 0, 1, NULL, 0, 0, 0, },
84 input2 = { NULL, 0, 0, 2, NULL, 0, 0, 0, };
87 u_long filenum; /* file number */
88 u_long fieldno; /* field number */
90 OLIST *olist; /* output field list */
91 u_long olistcnt; /* output field list count */
92 u_long olistalloc; /* output field allocated count */
94 int joinout = 1; /* show lines with matched join fields (-v) */
95 int needsep; /* need separator character */
96 int spans = 1; /* span multiple delimiters (-t) */
97 char *empty; /* empty field replacement string (-e) */
98 char *tabchar = " \t"; /* delimiter characters (-t) */
100 int cmp __P((LINE *, u_long, LINE *, u_long));
101 void fieldarg __P((char *));
102 void joinlines __P((INPUT *, INPUT *));
103 void obsolete __P((char **));
104 void outfield __P((LINE *, u_long, int));
105 void outoneline __P((INPUT *, LINE *));
106 void outtwoline __P((INPUT *, LINE *, INPUT *, LINE *));
107 void slurp __P((INPUT *));
108 void usage __P((void));
116 int aflag, ch, cval, vflag;
124 while ((ch = getopt(argc, argv, "\01a:e:j:1:2:o:t:v:")) != EOF) {
126 case '\01': /* See comment in obsolete(). */
128 F1->unpair = F2->unpair = 1;
131 if ((F1->joinf = strtol(optarg, &end, 10)) < 1)
132 errx(1, "-1 option field number less than 1");
134 errx(1, "illegal field number -- %s", optarg);
138 if ((F2->joinf = strtol(optarg, &end, 10)) < 1)
139 errx(1, "-2 option field number less than 1");
141 errx(1, "illegal field number -- %s", optarg);
146 switch(strtol(optarg, &end, 10)) {
154 errx(1, "-a option file number not 1 or 2");
158 errx(1, "illegal file number -- %s", optarg);
164 if ((F1->joinf = F2->joinf =
165 strtol(optarg, &end, 10)) < 1)
166 errx(1, "-j option field number less than 1");
168 errx(1, "illegal field number -- %s", optarg);
177 if (strlen(tabchar = optarg) != 1)
178 errx(1, "illegal tab character specification");
183 switch (strtol(optarg, &end, 10)) {
191 errx(1, "-v option file number not 1 or 2");
195 errx(1, "illegal file number -- %s", optarg);
206 errx(1, "the -a and -v options are mutually exclusive");
211 /* Open the files; "-" means stdin. */
212 if (!strcmp(*argv, "-"))
214 else if ((F1->fp = fopen(*argv, "r")) == NULL)
217 if (!strcmp(*argv, "-"))
219 else if ((F2->fp = fopen(*argv, "r")) == NULL)
221 if (F1->fp == stdin && F2->fp == stdin)
222 errx(1, "only one input file may be stdin");
226 while (F1->setcnt && F2->setcnt) {
227 cval = cmp(F1->set, F1->joinf, F2->set, F2->joinf);
229 /* Oh joy, oh rapture, oh beauty divine! */
234 } else if (cval < 0) {
235 /* File 1 takes the lead... */
240 /* File 2 takes the lead... */
248 * Now that one of the files is used up, optionally output any
249 * remaining lines from the other file.
268 LINE *lp, *lastlp, tmp;
274 * Read all of the lines from an input file that have the same
278 for (lastlp = NULL;; ++F->setcnt, lastlp = lp) {
280 * If we're out of space to hold line structures, allocate
281 * more. Initialize the structure so that we know that this
284 if (F->setcnt == F->setalloc) {
287 if ((F->set = realloc(F->set,
288 F->setalloc * sizeof(LINE))) == NULL)
290 memset(F->set + cnt, 0, 50 * sizeof(LINE));
294 * Get any pushed back line, else get the next line. Allocate
295 * space as necessary. If taking the line from the stack swap
296 * the two structures so that we don't lose space allocated to
297 * either structure. This could be avoided by doing another
298 * level of indirection, but it's probably okay as is.
299 * but it's probably okay as is.
301 lp = &F->set[F->setcnt];
303 tmp = F->set[F->setcnt];
304 F->set[F->setcnt] = F->set[F->pushback];
305 F->set[F->pushback] = tmp;
309 if ((bp = fgetln(F->fp, &len)) == NULL)
311 if (lp->linealloc <= len + 1) {
312 lp->linealloc += MAX(100, len + 1);
314 realloc(lp->line, lp->linealloc)) == NULL)
317 memmove(lp->line, bp, len);
319 /* Replace trailing newline, if it exists. */
320 if (bp[len - 1] == '\n')
321 lp->line[len - 1] = '\0';
323 lp->line[len] = '\0';
326 /* Split the line into fields, allocate space as necessary. */
328 while ((fieldp = strsep(&bp, tabchar)) != NULL) {
329 if (spans && *fieldp == '\0')
331 if (lp->fieldcnt == lp->fieldalloc) {
332 lp->fieldalloc += 50;
333 if ((lp->fields = realloc(lp->fields,
334 lp->fieldalloc * sizeof(char *))) == NULL)
337 lp->fields[lp->fieldcnt++] = fieldp;
340 /* See if the join field value has changed. */
341 if (lastlp != NULL && cmp(lp, F->joinf, lastlp, F->joinf)) {
343 F->pushback = F->setcnt;
350 cmp(lp1, fieldno1, lp2, fieldno2)
352 u_long fieldno1, fieldno2;
354 if (lp1->fieldcnt < fieldno1)
355 return (lp2->fieldcnt < fieldno2 ? 0 : 1);
356 if (lp2->fieldcnt < fieldno2)
358 return (strcmp(lp1->fields[fieldno1], lp2->fields[fieldno2]));
368 * Output the results of a join comparison. The output may be from
369 * either file 1 or file 2 (in which case the first argument is the
370 * file from which to output) or from both.
373 for (cnt1 = 0; cnt1 < F1->setcnt; ++cnt1)
374 outoneline(F1, &F1->set[cnt1]);
377 for (cnt1 = 0; cnt1 < F1->setcnt; ++cnt1)
378 for (cnt2 = 0; cnt2 < F2->setcnt; ++cnt2)
379 outtwoline(F1, &F1->set[cnt1], F2, &F2->set[cnt2]);
390 * Output a single line from one of the files, according to the
391 * join rules. This happens when we are writing unmatched single
392 * lines. Output empty fields in the right places.
395 for (cnt = 0; cnt < olistcnt; ++cnt) {
396 if (olist[cnt].filenum == F->number)
397 outfield(lp, olist[cnt].fieldno, 0);
402 for (cnt = 0; cnt < lp->fieldcnt; ++cnt)
403 outfield(lp, cnt, 0);
411 outtwoline(F1, lp1, F2, lp2)
417 /* Output a pair of lines according to the join list (if any). */
419 for (cnt = 0; cnt < olistcnt; ++cnt)
420 if (olist[cnt].filenum == 1)
421 outfield(lp1, olist[cnt].fieldno, 0);
422 else /* if (olist[cnt].filenum == 2) */
423 outfield(lp2, olist[cnt].fieldno, 0);
426 * Output the join field, then the remaining fields from F1
429 outfield(lp1, F1->joinf, 0);
430 for (cnt = 0; cnt < lp1->fieldcnt; ++cnt)
431 if (F1->joinf != cnt)
432 outfield(lp1, cnt, 0);
433 for (cnt = 0; cnt < lp2->fieldcnt; ++cnt)
434 if (F2->joinf != cnt)
435 outfield(lp2, cnt, 0);
444 outfield(lp, fieldno, out_empty)
450 (void)printf("%c", *tabchar);
452 if (lp->fieldcnt < fieldno || out_empty) {
454 (void)printf("%s", empty);
456 if (*lp->fields[fieldno] == '\0')
458 (void)printf("%s", lp->fields[fieldno]);
465 * Convert an output list argument "2.1, 1.3, 2.4" into an array of output
475 while ((token = strsep(&option, " \t")) != NULL) {
478 if (token[0] != '1' && token[0] != '2' || token[1] != '.')
479 errx(1, "malformed -o option field");
480 fieldno = strtol(token + 2, &end, 10);
482 errx(1, "malformed -o option field");
484 errx(1, "field numbers are 1 based");
485 if (olistcnt == olistalloc) {
487 if ((olist = realloc(olist,
488 olistalloc * sizeof(OLIST))) == NULL)
491 olist[olistcnt].filenum = token[0] - '0';
492 olist[olistcnt].fieldno = fieldno - 1;
504 while ((ap = *++argv) != NULL) {
505 /* Return if "--". */
506 if (ap[0] == '-' && ap[1] == '-')
511 * The original join allowed "-a", which meant the
512 * same as -a1 plus -a2. POSIX 1003.2, Draft 11.2
513 * only specifies this as "-a 1" and "a -2", so we
514 * have to use another option flag, one that is
515 * unlikely to ever be used or accidentally entered
516 * on the command line. (Well, we could reallocate
517 * the argv array, but that hardly seems worthwhile.)
524 * The original join allowed "-j[12] arg" and "-j arg".
525 * Convert the former to "-[12] arg". Don't convert
526 * the latter since getopt(3) can handle it.
544 jbad: errx(1, "illegal option -- %s", ap);
550 * The original join allowed "-o arg arg".
551 * Convert to "-o arg -o arg".
555 for (p = argv + 2; *p; ++p) {
556 if (p[0][0] != '1' &&
557 p[0][0] != '2' || p[0][1] != '.')
560 if (len - 2 != strspn(*p + 2, "0123456789"))
562 if ((t = malloc(len + 3)) == NULL)
566 memmove(t + 2, *p, len + 1);
578 (void)fprintf(stderr, "%s%s\n",
579 "usage: join [-a fileno | -v fileno ] [-e string] [-1 field] ",
580 "[-2 field]\n [-o list] [-t char] file1 file2");