Annotation of src/usr.bin/diff3/diff3prog.c, Revision 1.3
1.3 ! avsm 1: /* $OpenBSD: diff3prog.c,v 1.2 2003/07/14 17:43:52 deraadt Exp $ */
1.1 millert 2:
3: /*
4: * Copyright (C) Caldera International Inc. 2001-2002.
5: * All rights reserved.
6: *
7: * Redistribution and use in source and binary forms, with or without
8: * modification, are permitted provided that the following conditions
9: * are met:
10: * 1. Redistributions of source code and documentation must retain the above
11: * copyright notice, this list of conditions and the following disclaimer.
12: * 2. Redistributions in binary form must reproduce the above copyright
13: * notice, this list of conditions and the following disclaimer in the
14: * documentation and/or other materials provided with the distribution.
15: * 3. All advertising materials mentioning features or use of this software
16: * must display the following acknowledgement:
17: * This product includes software developed or owned by Caldera
18: * International, Inc.
19: * 4. Neither the name of Caldera International, Inc. nor the names of other
20: * contributors may be used to endorse or promote products derived from
21: * this software without specific prior written permission.
22: *
23: * USE OF THE SOFTWARE PROVIDED FOR UNDER THIS LICENSE BY CALDERA
24: * INTERNATIONAL, INC. AND CONTRIBUTORS ``AS IS'' AND ANY EXPRESS OR
25: * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
26: * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
27: * IN NO EVENT SHALL CALDERA INTERNATIONAL, INC. BE LIABLE FOR ANY DIRECT,
28: * INDIRECT INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
29: * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
30: * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
31: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
32: * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING
33: * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
34: * POSSIBILITY OF SUCH DAMAGE.
35: */
36: /*-
37: * Copyright (c) 1991, 1993
38: * The Regents of the University of California. All rights reserved.
39: *
40: * Redistribution and use in source and binary forms, with or without
41: * modification, are permitted provided that the following conditions
42: * are met:
43: * 1. Redistributions of source code must retain the above copyright
44: * notice, this list of conditions and the following disclaimer.
45: * 2. Redistributions in binary form must reproduce the above copyright
46: * notice, this list of conditions and the following disclaimer in the
47: * documentation and/or other materials provided with the distribution.
48: * 3. Neither the name of the University nor the names of its contributors
49: * may be used to endorse or promote products derived from this software
50: * without specific prior written permission.
51: *
52: * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
53: * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
54: * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
55: * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
56: * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
57: * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
58: * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
59: * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
60: * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
61: * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
62: * SUCH DAMAGE.
63: *
64: * @(#)diff3.c 8.1 (Berkeley) 6/6/93
65: */
66:
67: #ifndef lint
68: static const char copyright[] =
69: "@(#) Copyright (c) 1991, 1993\n\
70: The Regents of the University of California. All rights reserved.\n";
71: #endif /* not lint */
72:
73: #ifndef lint
1.3 ! avsm 74: static const char rcsid[] = "$OpenBSD: diff3prog.c,v 1.2 2003/07/14 17:43:52 deraadt Exp $";
1.1 millert 75: #endif /* not lint */
76:
77: #include <stdio.h>
78: #include <stdlib.h>
79: #include <ctype.h>
80: #include <err.h>
81:
82: /* diff3 - 3-way differential file comparison */
83:
84: /* diff3 [-ex3EX] d13 d23 f1 f2 f3 [m1 m3]
85: *
86: * d13 = diff report on f1 vs f3
87: * d23 = diff report on f2 vs f3
88: * f1, f2, f3 the 3 files
89: * if changes in f1 overlap with changes in f3, m1 and m3 are used
90: * to mark the overlaps; otherwise, the file names f1 and f3 are used
91: * (only for options E and X).
92: */
93:
94: /*
95: * "from" is first in range of changed lines; "to" is last+1
96: * from=to=line after point of insertion for added lines.
97: */
98: struct range {
99: int from;
100: int to;
101: };
102: struct diff {
103: struct range old;
104: struct range new;
105: };
106:
107: #define NC 200
108: struct diff d13[NC];
109: struct diff d23[NC];
110: /*
111: * "de" is used to gather editing scripts. These are later spewed out in
112: * reverse order. Its first element must be all zero, the "new" component
113: * of "de" contains line positions or byte positions depending on when you
114: * look (!?). Array overlap indicates which sections in "de" correspond to
115: * lines that are different in all three files.
116: */
117: struct diff de[NC];
118: char overlap[NC];
119: int overlapcnt;
120: char line[256];
121: FILE *fp[3];
122: int cline[3]; /* # of the last-read line in each file (0-2) */
123: /*
124: * the latest known correspondence between line numbers of the 3 files
125: * is stored in last[1-3];
126: */
127: int last[4];
128: int eflag;
1.2 deraadt 129: int oflag; /* indicates whether to mark overlaps (-E or -X)*/
1.1 millert 130: int debug = 0;
131: char f1mark[40], f3mark[40]; /* markers for -E and -X */
132:
133: int duplicate(struct range *, struct range *);
134: int edit(struct diff *, int, int);
135: int getchange(FILE *);
136: int getline(FILE *);
137: int number(char **);
138: int readin(char *, struct diff *);
139: int skip(int, int, char *);
140: void change(int, struct range *, int);
141: void keep(int, struct range *);
142: void merge(int, int);
143: void prange(struct range *);
144: void repos(int);
145: void separate(const char *);
146: __dead void edscript(int);
147: __dead void trouble(void);
148: __dead void usage(void);
149:
150: int
151: main(int argc, char **argv)
152: {
153: int ch, i, m, n;
154:
1.2 deraadt 155: eflag = 0;
1.1 millert 156: oflag = 0;
157: while ((ch = getopt(argc, argv, "EeXx3")) != -1) {
158: switch (ch) {
1.2 deraadt 159: case 'E':
160: eflag = 3;
161: oflag = 1;
162: break;
1.1 millert 163: case 'e':
164: eflag = 3;
165: break;
1.2 deraadt 166: case 'X':
167: oflag = eflag = 1;
168: break;
1.1 millert 169: case 'x':
170: eflag = 1;
1.2 deraadt 171: break;
1.1 millert 172: case '3':
173: eflag = 2;
174: break;
175: }
176: }
177: argc -= optind;
178: argv += optind;
179: /* XXX - argc usage seems wrong here */
180: if (argc < 5)
181: usage();
182:
1.2 deraadt 183: if (oflag) {
184: (void)snprintf(f1mark, sizeof(f1mark), "<<<<<<< %s",
1.1 millert 185: argc >= 6 ? argv[5] : argv[2]);
1.2 deraadt 186: (void)snprintf(f3mark, sizeof(f3mark), ">>>>>>> %s",
1.1 millert 187: argc >= 7 ? argv[6] : argv[4]);
1.2 deraadt 188: }
1.1 millert 189:
190: m = readin(argv[0], d13);
191: n = readin(argv[1], d23);
192: for (i = 0; i <= 2; i++) {
193: if ((fp[i] = fopen(argv[i + 2], "r")) == NULL) {
194: printf("diff3: can't open %s\n", argv[i + 2]);
195: exit(EXIT_FAILURE);
196: }
197: }
198: merge(m, n);
199: exit(EXIT_SUCCESS);
200: }
201:
202: /*
203: * Pick up the line numbers of allcahnges from one change file.
204: * (This puts the numbers in a vector, which is not strictly necessary,
205: * since the vector is processed in one sequential pass.
206: * The vector could be optimized out of existence)
207: */
208: int
209: readin(char *name, struct diff *dd)
210: {
211: int i;
212: int a,b,c,d;
213: char kind;
214: char *p;
215: fp[0] = fopen(name, "r");
216: for (i=0; getchange(fp[0]); i++) {
217: if (i >= NC)
218: err(EXIT_FAILURE, "too many changes");
219: p = line;
220: a = b = number(&p);
221: if (*p == ',') {
222: p++;
223: b = number(&p);
224: }
225: kind = *p++;
226: c = d = number(&p);
227: if (*p==',') {
228: p++;
229: d = number(&p);
230: }
231: if (kind == 'a')
232: a++;
233: if (kind == 'd')
234: c++;
235: b++;
236: d++;
237: dd[i].old.from = a;
238: dd[i].old.to = b;
239: dd[i].new.from = c;
240: dd[i].new.to = d;
241: }
242: dd[i].old.from = dd[i-1].old.to;
243: dd[i].new.from = dd[i-1].new.to;
244: (void)fclose(fp[0]);
245: return (i);
246: }
247:
248: int
249: number(char **lc)
250: {
251: int nn;
252: nn = 0;
253: while (isdigit((unsigned char)(**lc)))
254: nn = nn*10 + *(*lc)++ - '0';
255: return (nn);
256: }
257:
258: int
259: getchange(FILE *b)
260: {
261: while (getline(b)) {
262: if (isdigit((unsigned char)line[0]))
263: return (1);
264: }
265: return (0);
266: }
267:
268: int
269: getline(FILE *b)
270: {
271: int i, c;
272:
273: for (i = 0; i < sizeof(line) - 1; i++) {
274: c = getc(b);
275: if (c == EOF)
276: break;
277: line[i] = c;
278: if (c == '\n') {
279: line[++i] = 0;
280: return (i);
281: }
282: }
283: return (0);
284: }
285:
286: void
287: merge(int m1, int m2)
288: {
289: struct diff *d1, *d2, *d3;
290: int dup, j, t1, t2;
291:
292: d1 = d13;
293: d2 = d23;
294: j = 0;
295: while ((t1 = d1 < d13 + m1) | (t2 = d2 < d23 + m2)) {
296: if (debug) {
297: printf("%d,%d=%d,%d %d,%d=%d,%d\n",
298: d1->old.from,d1->old.to,
299: d1->new.from,d1->new.to,
300: d2->old.from,d2->old.to,
301: d2->new.from,d2->new.to);
302: }
303: /* first file is different from others */
304: if ((!t2 || t1) && d1->new.to < d2->new.from) {
305: /* stuff peculiar to 1st file */
306: if (eflag==0) {
307: separate("1");
308: change(1, &d1->old, 0);
309: keep(2, &d1->new);
310: change(3, &d1->new, 0);
311: }
312: d1++;
313: continue;
314: }
315: /* second file is different from others */
316: if ((!t1 || t2) && d2->new.to < d1->new.from) {
317: if (eflag==0) {
318: separate("2");
319: keep(1, &d2->new);
320: change(2, &d2->old, 0);
321: change(3, &d2->new, 0);
322: }
323: d2++;
324: continue;
325: }
326: /*
327: * Merge overlapping changes in first file
328: * this happens after extension (see below).
329: */
330: if (d1 + 1 < d13 + m1 && d1->new.to >= d1[1].new.from) {
331: d1[1].old.from = d1->old.from;
332: d1[1].new.from = d1->new.from;
333: d1++;
334: continue;
335: }
336:
337: /* merge overlapping changes in second */
338: if (d2 + 1 < d23 + m2 && d2->new.to >= d2[1].new.from) {
339: d2[1].old.from = d2->old.from;
340: d2[1].new.from = d2->new.from;
341: d2++;
342: continue;
343: }
344: /* stuff peculiar to third file or different in all */
345: if (d1->new.from == d2->new.from && d1->new.to == d2->new.to) {
346: dup = duplicate(&d1->old,&d2->old);
347: /*
348: * dup = 0 means all files differ
349: * dup = 1 meands files 1 and 2 identical
350: */
351: if (eflag==0) {
352: separate(dup ? "3" : "");
353: change(1, &d1->old, dup);
354: change(2, &d2->old, 0);
355: d3 = d1->old.to > d1->old.from ? d1 : d2;
356: change(3, &d3->new, 0);
357: } else
358: j = edit(d1, dup, j);
359: d1++;
360: d2++;
361: continue;
362: }
363: /*
364: * Overlapping changes from file 1 and 2; extend changes
365: * appropriately to make them coincide.
366: */
1.2 deraadt 367: if (d1->new.from < d2->new.from) {
1.1 millert 368: d2->old.from -= d2->new.from-d1->new.from;
369: d2->new.from = d1->new.from;
370: } else if (d2->new.from < d1->new.from) {
371: d1->old.from -= d1->new.from-d2->new.from;
372: d1->new.from = d2->new.from;
373: }
374: if (d1->new.to > d2->new.to) {
375: d2->old.to += d1->new.to - d2->new.to;
376: d2->new.to = d1->new.to;
377: } else if (d2->new.to > d1->new.to) {
378: d1->old.to += d2->new.to - d1->new.to;
379: d1->new.to = d2->new.to;
380: }
381: }
382: if (eflag)
383: edscript(j);
384: }
385:
386: void
387: separate(const char *s)
388: {
389: printf("====%s\n", s);
390: }
391:
392: /*
393: * The range of lines rold.from thru rold.to in file i is to be changed.
394: * It is to be printed only if it does not duplicate something to be
395: * printed later.
396: */
397: void
398: change(int i, struct range *rold, int dup)
399: {
400: printf("%d:", i);
401: last[i] = rold->to;
402: prange(rold);
403: if (dup || debug)
404: return;
405: i--;
406: (void)skip(i, rold->from, NULL);
407: (void)skip(i, rold->to, " ");
408: }
409:
410: /*
411: * print the range of line numbers, rold.from thru rold.to, as n1,n2 or n1
412: */
413: void
414: prange(struct range *rold)
415: {
416: if (rold->to <= rold->from)
417: printf("%da\n", rold->from - 1);
418: else {
419: printf("%d", rold->from);
420: if (rold->to > rold->from+1)
421: printf(",%d", rold->to - 1);
422: printf("c\n");
423: }
424: }
425:
426: /*
427: * No difference was reported by diff between file 1 (or 2) and file 3,
428: * and an artificial dummy difference (trange) must be ginned up to
429: * correspond to the change reported in the other file.
430: */
431: void
432: keep(int i, struct range *rnew)
433: {
434: int delta;
435: struct range trange;
436:
437: delta = last[3] - last[i];
438: trange.from = rnew->from - delta;
439: trange.to = rnew->to - delta;
440: change(i, &trange, 1);
441: }
442:
443: /*
444: * skip to just befor line number from in file "i". If "pr" is non-NULL,
445: * print all skipped stuff with string pr as a prefix.
446: */
447: int
448: skip(int i, int from, char *pr)
449: {
450: int j, n;
451:
452: for (n = 0; cline[i] < from - 1; n += j) {
1.3 ! avsm 453: if ((j = getline(fp[i])) == 0)
1.1 millert 454: trouble();
455: if (pr != NULL)
456: printf("%s%s", pr, line);
457: cline[i]++;
458: }
459: return (n);
460: }
461:
462: /*
463: * Return 1 or 0 according as the old range (in file 1) contains exactly
464: * the same data as the new range (in file 2).
465: */
466: int
467: duplicate(struct range *r1, struct range *r2)
468: {
469: int c,d;
470: int nchar;
471: int nline;
472:
473: if (r1->to-r1->from != r2->to-r2->from)
474: return (0);
475: (void)skip(0, r1->from, NULL);
476: (void)skip(1, r2->from, NULL);
477: nchar = 0;
478: for (nline=0; nline < r1->to - r1->from; nline++) {
479: do {
480: c = getc(fp[0]);
481: d = getc(fp[1]);
482: if (c == -1 || d== -1)
483: trouble();
484: nchar++;
485: if (c != d) {
486: repos(nchar);
487: return (0);
488: }
489: } while (c != '\n');
490: }
491: repos(nchar);
492: return (1);
493: }
494:
495: void
496: repos(int nchar)
497: {
498: int i;
499:
1.2 deraadt 500: for (i = 0; i < 2; i++)
1.1 millert 501: (void)fseek(fp[i], (long)-nchar, 1);
502: }
503:
504: __dead void
505: trouble(void)
506: {
507: errx(EXIT_FAILURE, "logic error");
508: }
509:
510: /*
511: * collect an editing script for later regurgitation
512: */
513: int
514: edit(struct diff *diff, int dup, int j)
515: {
516: if (((dup + 1) & eflag) == 0)
517: return (j);
518: j++;
1.2 deraadt 519: overlap[j] = !dup;
520: if (!dup)
1.1 millert 521: overlapcnt++;
522: de[j].old.from = diff->old.from;
523: de[j].old.to = diff->old.to;
524: de[j].new.from = de[j-1].new.to + skip(2, diff->new.from, NULL);
525: de[j].new.to = de[j].new.from + skip(2, diff->new.to, NULL);
526: return (j);
527: }
528:
529: /* regurgitate */
530: __dead void
531: edscript(int n)
532: {
533: int j,k;
534: char block[BUFSIZ];
535:
536: for (n = n; n > 0; n--) {
1.2 deraadt 537: if (!oflag || !overlap[n])
538: prange(&de[n].old);
539: else
540: printf("%da\n=======\n", de[n].old.to -1);
1.1 millert 541: (void)fseek(fp[2], (long)de[n].new.from, 0);
542: for (k = de[n].new.to-de[n].new.from; k > 0; k-= j) {
543: j = k > BUFSIZ ? BUFSIZ : k;
544: if (fread(block, 1, j, fp[2]) != j)
545: trouble();
546: (void)fwrite(block, 1, j, stdout);
547: }
1.2 deraadt 548: if (!oflag || !overlap[n])
549: printf(".\n");
550: else {
551: printf("%s\n.\n", f3mark);
552: printf("%da\n%s\n.\n", de[n].old.from - 1, f1mark);
553: }
1.1 millert 554: }
1.2 deraadt 555: exit(overlapcnt);
1.1 millert 556: }
557:
558: __dead void
559: usage(void)
560: {
561: extern char *__progname;
562:
563: fprintf(stderr, "usage: %s [-exEX3] /tmp/d3a.?????????? "
564: "/tmp/d3b.?????????? file1 file2 file3\n", __progname);
565: exit(EXIT_FAILURE);
566: }