1 |
|
|
/* $OpenBSD: diff3.c,v 1.62 2016/10/18 21:06:52 millert Exp $ */ |
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 |
|
|
#include <ctype.h> |
68 |
|
|
#include <errno.h> |
69 |
|
|
#include <fcntl.h> |
70 |
|
|
#include <stdio.h> |
71 |
|
|
#include <stdlib.h> |
72 |
|
|
#include <string.h> |
73 |
|
|
#include <unistd.h> |
74 |
|
|
|
75 |
|
|
#include "atomicio.h" |
76 |
|
|
#include "cvs.h" |
77 |
|
|
#include "diff.h" |
78 |
|
|
|
79 |
|
|
/* diff3 - 3-way differential file comparison */ |
80 |
|
|
|
81 |
|
|
/* diff3 [-ex3EX] d13 d23 f1 f2 f3 [m1 m3] |
82 |
|
|
* |
83 |
|
|
* d13 = diff report on f1 vs f3 |
84 |
|
|
* d23 = diff report on f2 vs f3 |
85 |
|
|
* f1, f2, f3 the 3 files |
86 |
|
|
* if changes in f1 overlap with changes in f3, m1 and m3 are used |
87 |
|
|
* to mark the overlaps; otherwise, the file names f1 and f3 are used |
88 |
|
|
* (only for options E and X). |
89 |
|
|
*/ |
90 |
|
|
|
91 |
|
|
/* |
92 |
|
|
* "from" is first in range of changed lines; "to" is last+1 |
93 |
|
|
* from=to=line after point of insertion for added lines. |
94 |
|
|
*/ |
95 |
|
|
struct range { |
96 |
|
|
int from; |
97 |
|
|
int to; |
98 |
|
|
}; |
99 |
|
|
|
100 |
|
|
struct diff { |
101 |
|
|
struct range old; |
102 |
|
|
struct range new; |
103 |
|
|
}; |
104 |
|
|
|
105 |
|
|
static size_t szchanges; |
106 |
|
|
|
107 |
|
|
static struct diff *d13; |
108 |
|
|
static struct diff *d23; |
109 |
|
|
|
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 |
|
|
static struct diff *de; |
118 |
|
|
static char *overlap; |
119 |
|
|
static int overlapcnt = 0; |
120 |
|
|
static FILE *fp[3]; |
121 |
|
|
static int cline[3]; /* # of the last-read line in each file (0-2) */ |
122 |
|
|
|
123 |
|
|
/* |
124 |
|
|
* the latest known correspondence between line numbers of the 3 files |
125 |
|
|
* is stored in last[1-3]; |
126 |
|
|
*/ |
127 |
|
|
static int last[4]; |
128 |
|
|
static int eflag; |
129 |
|
|
static int oflag; /* indicates whether to mark overlaps (-E or -X)*/ |
130 |
|
|
static int debug = 0; |
131 |
|
|
static char f1mark[PATH_MAX], f3mark[PATH_MAX]; /* markers for -E and -X */ |
132 |
|
|
|
133 |
|
|
static int duplicate(struct range *, struct range *); |
134 |
|
|
static int edit(struct diff *, int, int); |
135 |
|
|
static char *getchange(FILE *); |
136 |
|
|
static char *get_line(FILE *, size_t *); |
137 |
|
|
static int number(char **); |
138 |
|
|
static size_t readin(int, struct diff **); |
139 |
|
|
static int skip(int, int, char *); |
140 |
|
|
static int edscript(int); |
141 |
|
|
static int merge(size_t, size_t); |
142 |
|
|
static void change(int, struct range *, int); |
143 |
|
|
static void keep(int, struct range *); |
144 |
|
|
static void prange(struct range *); |
145 |
|
|
static void repos(int); |
146 |
|
|
static void separate(const char *); |
147 |
|
|
static void increase(void); |
148 |
|
|
static int diff3_internal(int, char **, const char *, const char *); |
149 |
|
|
|
150 |
|
|
int diff3_conflicts = 0; |
151 |
|
|
RCSNUM *d3rev1 = NULL; |
152 |
|
|
RCSNUM *d3rev2 = NULL; |
153 |
|
|
|
154 |
|
|
static int fds[5]; |
155 |
|
|
|
156 |
|
|
void |
157 |
|
|
cvs_merge_file(struct cvs_file *cf, int verbose) |
158 |
|
|
{ |
159 |
|
|
int i, argc; |
160 |
|
|
char *data, *patch; |
161 |
|
|
char *argv[5], r1[CVS_REV_BUFSZ], r2[CVS_REV_BUFSZ]; |
162 |
|
|
char *dp13, *dp23, *path1, *path2, *path3; |
163 |
|
|
BUF *b1, *d1, *d2, *diffb; |
164 |
|
|
size_t dlen, plen; |
165 |
|
|
struct rcs_line *lp; |
166 |
|
|
struct rcs_lines *dlines, *plines; |
167 |
|
|
|
168 |
|
|
overlapcnt = 0; |
169 |
|
|
b1 = d1 = d2 = diffb = NULL; |
170 |
|
|
|
171 |
|
|
rcsnum_tostr(d3rev1, r1, sizeof(r1)); |
172 |
|
|
rcsnum_tostr(d3rev2, r2, sizeof(r2)); |
173 |
|
|
|
174 |
|
|
b1 = buf_load_fd(cf->fd); |
175 |
|
|
d1 = buf_alloc(128); |
176 |
|
|
d2 = buf_alloc(128); |
177 |
|
|
diffb = buf_alloc(128); |
178 |
|
|
|
179 |
|
|
(void)xasprintf(&path1, "%s/diff1.XXXXXXXXXX", cvs_tmpdir); |
180 |
|
|
(void)xasprintf(&path2, "%s/diff2.XXXXXXXXXX", cvs_tmpdir); |
181 |
|
|
(void)xasprintf(&path3, "%s/diff3.XXXXXXXXXX", cvs_tmpdir); |
182 |
|
|
|
183 |
|
|
fds[2] = buf_write_stmp(b1, path1, NULL); |
184 |
|
|
if (verbose == 1) |
185 |
|
|
cvs_printf("Retrieving revision %s\n", r1); |
186 |
|
|
fds[3] = rcs_rev_write_stmp(cf->file_rcs, d3rev1, path2, 0); |
187 |
|
|
if (verbose == 1) |
188 |
|
|
cvs_printf("Retrieving revision %s\n", r2); |
189 |
|
|
fds[4] = rcs_rev_write_stmp(cf->file_rcs, d3rev2, path3, 0); |
190 |
|
|
|
191 |
|
|
diffreg(path1, path3, fds[2], fds[4], d1, D_FORCEASCII); |
192 |
|
|
diffreg(path2, path3, fds[3], fds[4], d2, D_FORCEASCII); |
193 |
|
|
|
194 |
|
|
(void)xasprintf(&dp13, "%s/d13.XXXXXXXXXX", cvs_tmpdir); |
195 |
|
|
fds[0] = buf_write_stmp(d1, dp13, NULL); |
196 |
|
|
buf_free(d1); |
197 |
|
|
|
198 |
|
|
(void)xasprintf(&dp23, "%s/d23.XXXXXXXXXX", cvs_tmpdir); |
199 |
|
|
fds[1] = buf_write_stmp(d2, dp23, NULL); |
200 |
|
|
buf_free(d2); |
201 |
|
|
|
202 |
|
|
argc = 0; |
203 |
|
|
diffbuf = diffb; |
204 |
|
|
argv[argc++] = dp13; |
205 |
|
|
argv[argc++] = dp23; |
206 |
|
|
argv[argc++] = path1; |
207 |
|
|
argv[argc++] = path2; |
208 |
|
|
argv[argc++] = path3; |
209 |
|
|
|
210 |
|
|
if (lseek(fds[2], 0, SEEK_SET) < 0) |
211 |
|
|
fatal("cvs_merge_file: lseek fds[2]: %s", strerror(errno)); |
212 |
|
|
if (lseek(fds[3], 0, SEEK_SET) < 0) |
213 |
|
|
fatal("cvs_merge_file: lseek fds[3]: %s", strerror(errno)); |
214 |
|
|
if (lseek(fds[4], 0, SEEK_SET) < 0) |
215 |
|
|
fatal("cvs_merge_file: lseek fds[4]: %s", strerror(errno)); |
216 |
|
|
|
217 |
|
|
diff3_conflicts = diff3_internal(argc, argv, cf->file_path, r2); |
218 |
|
|
if (diff3_conflicts < 0) |
219 |
|
|
fatal("cvs_merge_file: merging failed for an unknown reason"); |
220 |
|
|
|
221 |
|
|
plen = buf_len(diffb); |
222 |
|
|
patch = buf_release(diffb); |
223 |
|
|
dlen = buf_len(b1); |
224 |
|
|
data = buf_release(b1); |
225 |
|
|
|
226 |
|
|
if (verbose == 1) |
227 |
|
|
cvs_printf("Merging differences between %s and %s into `%s'\n", |
228 |
|
|
r1, r2, cf->file_path); |
229 |
|
|
|
230 |
|
|
dlines = cvs_splitlines(data, dlen); |
231 |
|
|
plines = cvs_splitlines(patch, plen); |
232 |
|
|
|
233 |
|
|
ed_patch_lines(dlines, plines); |
234 |
|
|
cvs_freelines(plines); |
235 |
|
|
|
236 |
|
|
if (verbose == 1 && diff3_conflicts != 0) { |
237 |
|
|
cvs_log(LP_ERR, "%d conflict%s found during merge, " |
238 |
|
|
"please correct.", diff3_conflicts, |
239 |
|
|
(diff3_conflicts > 1) ? "s" : ""); |
240 |
|
|
} |
241 |
|
|
|
242 |
|
|
(void)close(cf->fd); |
243 |
|
|
cf->fd = open(cf->file_path, O_CREAT | O_RDWR | O_TRUNC, 0644); |
244 |
|
|
if (cf->fd == -1) { |
245 |
|
|
fatal("cvs_merge_file: failed to reopen fd for writing: %s", |
246 |
|
|
strerror(errno)); |
247 |
|
|
} |
248 |
|
|
|
249 |
|
|
TAILQ_FOREACH(lp, &(dlines->l_lines), l_list) { |
250 |
|
|
if (lp->l_line == NULL) |
251 |
|
|
continue; |
252 |
|
|
|
253 |
|
|
if (atomicio(vwrite, cf->fd, lp->l_line, lp->l_len) != |
254 |
|
|
lp->l_len) |
255 |
|
|
fatal("cvs_merge_file: %s", strerror(errno)); |
256 |
|
|
} |
257 |
|
|
|
258 |
|
|
cvs_freelines(dlines); |
259 |
|
|
|
260 |
|
|
free(data); |
261 |
|
|
free(patch); |
262 |
|
|
|
263 |
|
|
for (i = 0; i < 3; i++) |
264 |
|
|
fclose(fp[i]); |
265 |
|
|
|
266 |
|
|
worklist_run(&temp_files, worklist_unlink); |
267 |
|
|
|
268 |
|
|
free(path1); |
269 |
|
|
free(path2); |
270 |
|
|
free(path3); |
271 |
|
|
free(dp13); |
272 |
|
|
free(dp23); |
273 |
|
|
} |
274 |
|
|
|
275 |
|
|
static int |
276 |
|
|
diff3_internal(int argc, char **argv, const char *fmark, const char *rmark) |
277 |
|
|
{ |
278 |
|
|
size_t m, n; |
279 |
|
|
int i; |
280 |
|
|
|
281 |
|
|
eflag = 3; |
282 |
|
|
oflag = 1; |
283 |
|
|
|
284 |
|
|
if (argc < 5) |
285 |
|
|
return (-1); |
286 |
|
|
|
287 |
|
|
(void)xsnprintf(f1mark, sizeof(f1mark), "<<<<<<< %s", fmark); |
288 |
|
|
(void)xsnprintf(f3mark, sizeof(f3mark), ">>>>>>> %s", rmark); |
289 |
|
|
|
290 |
|
|
szchanges = 0; |
291 |
|
|
memset(last, 0, sizeof(last)); |
292 |
|
|
memset(cline, 0, sizeof(cline)); |
293 |
|
|
free(d13); |
294 |
|
|
free(d23); |
295 |
|
|
free(overlap); |
296 |
|
|
free(de); |
297 |
|
|
|
298 |
|
|
de = d13 = d23 = NULL; |
299 |
|
|
overlap = NULL; |
300 |
|
|
|
301 |
|
|
increase(); |
302 |
|
|
|
303 |
|
|
/* fds[0] and fds[1] are closed in readin() */ |
304 |
|
|
m = readin(fds[0], &d13); |
305 |
|
|
n = readin(fds[1], &d23); |
306 |
|
|
|
307 |
|
|
for (i = 0; i <= 2; i++) { |
308 |
|
|
if ((fp[i] = fdopen(fds[i + 2], "r")) == NULL) { |
309 |
|
|
cvs_log(LP_ERR, "%s", argv[i + 2]); |
310 |
|
|
return (-1); |
311 |
|
|
} |
312 |
|
|
} |
313 |
|
|
|
314 |
|
|
return (merge(m, n)); |
315 |
|
|
} |
316 |
|
|
|
317 |
|
|
int |
318 |
|
|
ed_patch_lines(struct rcs_lines *dlines, struct rcs_lines *plines) |
319 |
|
|
{ |
320 |
|
|
char op, *ep; |
321 |
|
|
struct rcs_line *sort, *lp, *dlp, *ndlp, *insert_after; |
322 |
|
|
int start, end, i, lineno; |
323 |
|
|
u_char tmp; |
324 |
|
|
|
325 |
|
|
dlp = TAILQ_FIRST(&(dlines->l_lines)); |
326 |
|
|
lp = TAILQ_FIRST(&(plines->l_lines)); |
327 |
|
|
|
328 |
|
|
end = 0; |
329 |
|
|
for (lp = TAILQ_NEXT(lp, l_list); lp != NULL; |
330 |
|
|
lp = TAILQ_NEXT(lp, l_list)) { |
331 |
|
|
/* Skip blank lines */ |
332 |
|
|
if (lp->l_len < 2) |
333 |
|
|
continue; |
334 |
|
|
|
335 |
|
|
/* NUL-terminate line buffer for strtol() safety. */ |
336 |
|
|
tmp = lp->l_line[lp->l_len - 1]; |
337 |
|
|
lp->l_line[lp->l_len - 1] = '\0'; |
338 |
|
|
|
339 |
|
|
op = lp->l_line[strlen(lp->l_line) - 1]; |
340 |
|
|
start = (int)strtol(lp->l_line, &ep, 10); |
341 |
|
|
|
342 |
|
|
/* Restore the last byte of the buffer */ |
343 |
|
|
lp->l_line[lp->l_len - 1] = tmp; |
344 |
|
|
|
345 |
|
|
if (op == 'a') { |
346 |
|
|
if (start > dlines->l_nblines || |
347 |
|
|
start < 0 || *ep != 'a') |
348 |
|
|
fatal("ed_patch_lines %d", start); |
349 |
|
|
} else if (op == 'c') { |
350 |
|
|
if (start > dlines->l_nblines || |
351 |
|
|
start < 0 || (*ep != ',' && *ep != 'c')) |
352 |
|
|
fatal("ed_patch_lines"); |
353 |
|
|
|
354 |
|
|
if (*ep == ',') { |
355 |
|
|
ep++; |
356 |
|
|
end = (int)strtol(ep, &ep, 10); |
357 |
|
|
if (end < 0 || *ep != 'c') |
358 |
|
|
fatal("ed_patch_lines"); |
359 |
|
|
} else { |
360 |
|
|
end = start; |
361 |
|
|
} |
362 |
|
|
} else { |
363 |
|
|
fatal("invalid op %c found while merging", op); |
364 |
|
|
} |
365 |
|
|
|
366 |
|
|
|
367 |
|
|
for (;;) { |
368 |
|
|
if (dlp == NULL) |
369 |
|
|
break; |
370 |
|
|
if (dlp->l_lineno == start) |
371 |
|
|
break; |
372 |
|
|
if (dlp->l_lineno > start) { |
373 |
|
|
dlp = TAILQ_PREV(dlp, tqh, l_list); |
374 |
|
|
} else if (dlp->l_lineno < start) { |
375 |
|
|
ndlp = TAILQ_NEXT(dlp, l_list); |
376 |
|
|
if (ndlp->l_lineno > start) |
377 |
|
|
break; |
378 |
|
|
dlp = ndlp; |
379 |
|
|
} |
380 |
|
|
} |
381 |
|
|
|
382 |
|
|
if (dlp == NULL) |
383 |
|
|
fatal("ed_patch_lines"); |
384 |
|
|
|
385 |
|
|
|
386 |
|
|
if (op == 'c') { |
387 |
|
|
insert_after = TAILQ_PREV(dlp, tqh, l_list); |
388 |
|
|
for (i = 0; i <= (end - start); i++) { |
389 |
|
|
ndlp = TAILQ_NEXT(dlp, l_list); |
390 |
|
|
TAILQ_REMOVE(&(dlines->l_lines), dlp, l_list); |
391 |
|
|
dlp = ndlp; |
392 |
|
|
} |
393 |
|
|
dlp = insert_after; |
394 |
|
|
} |
395 |
|
|
|
396 |
|
|
if (op == 'a' || op == 'c') { |
397 |
|
|
for (;;) { |
398 |
|
|
ndlp = lp; |
399 |
|
|
lp = TAILQ_NEXT(lp, l_list); |
400 |
|
|
if (lp == NULL) |
401 |
|
|
fatal("ed_patch_lines"); |
402 |
|
|
|
403 |
|
|
if (lp->l_len == 2 && |
404 |
|
|
lp->l_line[0] == '.' && |
405 |
|
|
lp->l_line[1] == '\n') |
406 |
|
|
break; |
407 |
|
|
|
408 |
|
|
TAILQ_REMOVE(&(plines->l_lines), lp, l_list); |
409 |
|
|
TAILQ_INSERT_AFTER(&(dlines->l_lines), dlp, |
410 |
|
|
lp, l_list); |
411 |
|
|
dlp = lp; |
412 |
|
|
|
413 |
|
|
lp->l_lineno = start; |
414 |
|
|
lp = ndlp; |
415 |
|
|
} |
416 |
|
|
} |
417 |
|
|
|
418 |
|
|
/* |
419 |
|
|
* always resort lines as the markers might be put at the |
420 |
|
|
* same line as we first started editing. |
421 |
|
|
*/ |
422 |
|
|
lineno = 0; |
423 |
|
|
TAILQ_FOREACH(sort, &(dlines->l_lines), l_list) |
424 |
|
|
sort->l_lineno = lineno++; |
425 |
|
|
dlines->l_nblines = lineno - 1; |
426 |
|
|
} |
427 |
|
|
|
428 |
|
|
return (0); |
429 |
|
|
} |
430 |
|
|
|
431 |
|
|
/* |
432 |
|
|
* Pick up the line numbers of all changes from one change file. |
433 |
|
|
* (This puts the numbers in a vector, which is not strictly necessary, |
434 |
|
|
* since the vector is processed in one sequential pass. |
435 |
|
|
* The vector could be optimized out of existence) |
436 |
|
|
*/ |
437 |
|
|
static size_t |
438 |
|
|
readin(int fd, struct diff **dd) |
439 |
|
|
{ |
440 |
|
|
int a, b, c, d; |
441 |
|
|
char kind, *p; |
442 |
|
|
size_t i; |
443 |
|
|
|
444 |
|
|
fp[0] = fdopen(fd, "r"); |
445 |
|
|
if (fp[0] == NULL) |
446 |
|
|
fatal("readin: fdopen: %s", strerror(errno)); |
447 |
|
|
|
448 |
|
|
for (i = 0; (p = getchange(fp[0])); i++) { |
449 |
|
|
if (i >= szchanges - 1) |
450 |
|
|
increase(); |
451 |
|
|
a = b = number(&p); |
452 |
|
|
if (*p == ',') { |
453 |
|
|
p++; |
454 |
|
|
b = number(&p); |
455 |
|
|
} |
456 |
|
|
kind = *p++; |
457 |
|
|
c = d = number(&p); |
458 |
|
|
if (*p==',') { |
459 |
|
|
p++; |
460 |
|
|
d = number(&p); |
461 |
|
|
} |
462 |
|
|
if (kind == 'a') |
463 |
|
|
a++; |
464 |
|
|
if (kind == 'd') |
465 |
|
|
c++; |
466 |
|
|
b++; |
467 |
|
|
d++; |
468 |
|
|
(*dd)[i].old.from = a; |
469 |
|
|
(*dd)[i].old.to = b; |
470 |
|
|
(*dd)[i].new.from = c; |
471 |
|
|
(*dd)[i].new.to = d; |
472 |
|
|
} |
473 |
|
|
|
474 |
|
|
if (i) { |
475 |
|
|
(*dd)[i].old.from = (*dd)[i-1].old.to; |
476 |
|
|
(*dd)[i].new.from = (*dd)[i-1].new.to; |
477 |
|
|
} |
478 |
|
|
|
479 |
|
|
(void)fclose(fp[0]); |
480 |
|
|
|
481 |
|
|
return (i); |
482 |
|
|
} |
483 |
|
|
|
484 |
|
|
static int |
485 |
|
|
number(char **lc) |
486 |
|
|
{ |
487 |
|
|
int nn; |
488 |
|
|
|
489 |
|
|
nn = 0; |
490 |
|
|
while (isdigit((unsigned char)(**lc))) |
491 |
|
|
nn = nn*10 + *(*lc)++ - '0'; |
492 |
|
|
|
493 |
|
|
return (nn); |
494 |
|
|
} |
495 |
|
|
|
496 |
|
|
static char * |
497 |
|
|
getchange(FILE *b) |
498 |
|
|
{ |
499 |
|
|
char *line; |
500 |
|
|
|
501 |
|
|
while ((line = get_line(b, NULL))) { |
502 |
|
|
if (isdigit((unsigned char)line[0])) |
503 |
|
|
return (line); |
504 |
|
|
} |
505 |
|
|
|
506 |
|
|
return (NULL); |
507 |
|
|
} |
508 |
|
|
|
509 |
|
|
static char * |
510 |
|
|
get_line(FILE *b, size_t *n) |
511 |
|
|
{ |
512 |
|
|
char *cp; |
513 |
|
|
size_t len; |
514 |
|
|
static char *buf; |
515 |
|
|
static size_t bufsize; |
516 |
|
|
|
517 |
|
|
if ((cp = fgetln(b, &len)) == NULL) |
518 |
|
|
return (NULL); |
519 |
|
|
|
520 |
|
|
if (cp[len - 1] != '\n') |
521 |
|
|
len++; |
522 |
|
|
if (len + 1 > bufsize) { |
523 |
|
|
do { |
524 |
|
|
bufsize += 1024; |
525 |
|
|
} while (len + 1 > bufsize); |
526 |
|
|
buf = xreallocarray(buf, 1, bufsize); |
527 |
|
|
} |
528 |
|
|
memcpy(buf, cp, len - 1); |
529 |
|
|
buf[len - 1] = '\n'; |
530 |
|
|
buf[len] = '\0'; |
531 |
|
|
if (n != NULL) |
532 |
|
|
*n = len; |
533 |
|
|
|
534 |
|
|
return (buf); |
535 |
|
|
} |
536 |
|
|
|
537 |
|
|
static int |
538 |
|
|
merge(size_t m1, size_t m2) |
539 |
|
|
{ |
540 |
|
|
struct diff *d1, *d2, *d3; |
541 |
|
|
int dpl, j, t1, t2; |
542 |
|
|
|
543 |
|
|
d1 = d13; |
544 |
|
|
d2 = d23; |
545 |
|
|
j = 0; |
546 |
|
|
for (;;) { |
547 |
|
|
t1 = (d1 < d13 + m1); |
548 |
|
|
t2 = (d2 < d23 + m2); |
549 |
|
|
if (!t1 && !t2) |
550 |
|
|
break; |
551 |
|
|
|
552 |
|
|
if (debug) { |
553 |
|
|
printf("%d,%d=%d,%d %d,%d=%d,%d\n", |
554 |
|
|
d1->old.from, d1->old.to, |
555 |
|
|
d1->new.from, d1->new.to, |
556 |
|
|
d2->old.from, d2->old.to, |
557 |
|
|
d2->new.from, d2->new.to); |
558 |
|
|
} |
559 |
|
|
|
560 |
|
|
/* first file is different from others */ |
561 |
|
|
if (!t2 || (t1 && d1->new.to < d2->new.from)) { |
562 |
|
|
/* stuff peculiar to 1st file */ |
563 |
|
|
if (eflag==0) { |
564 |
|
|
separate("1"); |
565 |
|
|
change(1, &d1->old, 0); |
566 |
|
|
keep(2, &d1->new); |
567 |
|
|
change(3, &d1->new, 0); |
568 |
|
|
} |
569 |
|
|
d1++; |
570 |
|
|
continue; |
571 |
|
|
} |
572 |
|
|
|
573 |
|
|
/* second file is different from others */ |
574 |
|
|
if (!t1 || (t2 && d2->new.to < d1->new.from)) { |
575 |
|
|
if (eflag==0) { |
576 |
|
|
separate("2"); |
577 |
|
|
keep(1, &d2->new); |
578 |
|
|
change(2, &d2->old, 0); |
579 |
|
|
change(3, &d2->new, 0); |
580 |
|
|
} |
581 |
|
|
d2++; |
582 |
|
|
continue; |
583 |
|
|
} |
584 |
|
|
|
585 |
|
|
/* |
586 |
|
|
* Merge overlapping changes in first file |
587 |
|
|
* this happens after extension (see below). |
588 |
|
|
*/ |
589 |
|
|
if (d1 + 1 < d13 + m1 && d1->new.to >= d1[1].new.from) { |
590 |
|
|
d1[1].old.from = d1->old.from; |
591 |
|
|
d1[1].new.from = d1->new.from; |
592 |
|
|
d1++; |
593 |
|
|
continue; |
594 |
|
|
} |
595 |
|
|
|
596 |
|
|
/* merge overlapping changes in second */ |
597 |
|
|
if (d2 + 1 < d23 + m2 && d2->new.to >= d2[1].new.from) { |
598 |
|
|
d2[1].old.from = d2->old.from; |
599 |
|
|
d2[1].new.from = d2->new.from; |
600 |
|
|
d2++; |
601 |
|
|
continue; |
602 |
|
|
} |
603 |
|
|
/* stuff peculiar to third file or different in all */ |
604 |
|
|
if (d1->new.from == d2->new.from && d1->new.to == d2->new.to) { |
605 |
|
|
dpl = duplicate(&d1->old,&d2->old); |
606 |
|
|
if (dpl == -1) |
607 |
|
|
return (-1); |
608 |
|
|
|
609 |
|
|
/* |
610 |
|
|
* dpl = 0 means all files differ |
611 |
|
|
* dpl = 1 means files 1 and 2 identical |
612 |
|
|
*/ |
613 |
|
|
if (eflag==0) { |
614 |
|
|
separate(dpl ? "3" : ""); |
615 |
|
|
change(1, &d1->old, dpl); |
616 |
|
|
change(2, &d2->old, 0); |
617 |
|
|
d3 = d1->old.to > d1->old.from ? d1 : d2; |
618 |
|
|
change(3, &d3->new, 0); |
619 |
|
|
} else |
620 |
|
|
j = edit(d1, dpl, j); |
621 |
|
|
d1++; |
622 |
|
|
d2++; |
623 |
|
|
continue; |
624 |
|
|
} |
625 |
|
|
|
626 |
|
|
/* |
627 |
|
|
* Overlapping changes from file 1 and 2; extend changes |
628 |
|
|
* appropriately to make them coincide. |
629 |
|
|
*/ |
630 |
|
|
if (d1->new.from < d2->new.from) { |
631 |
|
|
d2->old.from -= d2->new.from-d1->new.from; |
632 |
|
|
d2->new.from = d1->new.from; |
633 |
|
|
} else if (d2->new.from < d1->new.from) { |
634 |
|
|
d1->old.from -= d1->new.from-d2->new.from; |
635 |
|
|
d1->new.from = d2->new.from; |
636 |
|
|
} |
637 |
|
|
if (d1->new.to > d2->new.to) { |
638 |
|
|
d2->old.to += d1->new.to - d2->new.to; |
639 |
|
|
d2->new.to = d1->new.to; |
640 |
|
|
} else if (d2->new.to > d1->new.to) { |
641 |
|
|
d1->old.to += d2->new.to - d1->new.to; |
642 |
|
|
d1->new.to = d2->new.to; |
643 |
|
|
} |
644 |
|
|
} |
645 |
|
|
|
646 |
|
|
return (edscript(j)); |
647 |
|
|
} |
648 |
|
|
|
649 |
|
|
static void |
650 |
|
|
separate(const char *s) |
651 |
|
|
{ |
652 |
|
|
diff_output("====%s\n", s); |
653 |
|
|
} |
654 |
|
|
|
655 |
|
|
/* |
656 |
|
|
* The range of lines rold.from thru rold.to in file i is to be changed. |
657 |
|
|
* It is to be printed only if it does not duplicate something to be |
658 |
|
|
* printed later. |
659 |
|
|
*/ |
660 |
|
|
static void |
661 |
|
|
change(int i, struct range *rold, int fdup) |
662 |
|
|
{ |
663 |
|
|
diff_output("%d:", i); |
664 |
|
|
last[i] = rold->to; |
665 |
|
|
prange(rold); |
666 |
|
|
if (fdup || debug) |
667 |
|
|
return; |
668 |
|
|
i--; |
669 |
|
|
(void)skip(i, rold->from, NULL); |
670 |
|
|
(void)skip(i, rold->to, " "); |
671 |
|
|
} |
672 |
|
|
|
673 |
|
|
/* |
674 |
|
|
* print the range of line numbers, rold.from thru rold.to, as n1,n2 or n1 |
675 |
|
|
*/ |
676 |
|
|
static void |
677 |
|
|
prange(struct range *rold) |
678 |
|
|
{ |
679 |
|
|
if (rold->to <= rold->from) |
680 |
|
|
diff_output("%da\n", rold->from - 1); |
681 |
|
|
else { |
682 |
|
|
diff_output("%d", rold->from); |
683 |
|
|
if (rold->to > rold->from+1) |
684 |
|
|
diff_output(",%d", rold->to - 1); |
685 |
|
|
diff_output("c\n"); |
686 |
|
|
} |
687 |
|
|
} |
688 |
|
|
|
689 |
|
|
/* |
690 |
|
|
* No difference was reported by diff between file 1 (or 2) and file 3, |
691 |
|
|
* and an artificial dummy difference (trange) must be ginned up to |
692 |
|
|
* correspond to the change reported in the other file. |
693 |
|
|
*/ |
694 |
|
|
static void |
695 |
|
|
keep(int i, struct range *rnew) |
696 |
|
|
{ |
697 |
|
|
int delta; |
698 |
|
|
struct range trange; |
699 |
|
|
|
700 |
|
|
delta = last[3] - last[i]; |
701 |
|
|
trange.from = rnew->from - delta; |
702 |
|
|
trange.to = rnew->to - delta; |
703 |
|
|
change(i, &trange, 1); |
704 |
|
|
} |
705 |
|
|
|
706 |
|
|
/* |
707 |
|
|
* skip to just before line number from in file "i". If "pr" is non-NULL, |
708 |
|
|
* print all skipped stuff with string pr as a prefix. |
709 |
|
|
*/ |
710 |
|
|
static int |
711 |
|
|
skip(int i, int from, char *pr) |
712 |
|
|
{ |
713 |
|
|
size_t j, n; |
714 |
|
|
char *line; |
715 |
|
|
|
716 |
|
|
for (n = 0; cline[i] < from - 1; n += j) { |
717 |
|
|
if ((line = get_line(fp[i], &j)) == NULL) |
718 |
|
|
return (-1); |
719 |
|
|
if (pr != NULL) |
720 |
|
|
diff_output("%s%s", pr, line); |
721 |
|
|
cline[i]++; |
722 |
|
|
} |
723 |
|
|
return ((int) n); |
724 |
|
|
} |
725 |
|
|
|
726 |
|
|
/* |
727 |
|
|
* Return 1 or 0 according as the old range (in file 1) contains exactly |
728 |
|
|
* the same data as the new range (in file 2). |
729 |
|
|
*/ |
730 |
|
|
static int |
731 |
|
|
duplicate(struct range *r1, struct range *r2) |
732 |
|
|
{ |
733 |
|
|
int c,d; |
734 |
|
|
int nchar; |
735 |
|
|
int nline; |
736 |
|
|
|
737 |
|
|
if (r1->to-r1->from != r2->to-r2->from) |
738 |
|
|
return (0); |
739 |
|
|
(void)skip(0, r1->from, NULL); |
740 |
|
|
(void)skip(1, r2->from, NULL); |
741 |
|
|
nchar = 0; |
742 |
|
|
for (nline=0; nline < r1->to - r1->from; nline++) { |
743 |
|
|
do { |
744 |
|
|
c = getc(fp[0]); |
745 |
|
|
d = getc(fp[1]); |
746 |
|
|
if (c == -1 || d== -1) |
747 |
|
|
return (-1); |
748 |
|
|
nchar++; |
749 |
|
|
if (c != d) { |
750 |
|
|
repos(nchar); |
751 |
|
|
return (0); |
752 |
|
|
} |
753 |
|
|
} while (c != '\n'); |
754 |
|
|
} |
755 |
|
|
repos(nchar); |
756 |
|
|
return (1); |
757 |
|
|
} |
758 |
|
|
|
759 |
|
|
static void |
760 |
|
|
repos(int nchar) |
761 |
|
|
{ |
762 |
|
|
int i; |
763 |
|
|
|
764 |
|
|
for (i = 0; i < 2; i++) |
765 |
|
|
(void)fseek(fp[i], (long)-nchar, SEEK_CUR); |
766 |
|
|
} |
767 |
|
|
|
768 |
|
|
/* |
769 |
|
|
* collect an editing script for later regurgitation |
770 |
|
|
*/ |
771 |
|
|
static int |
772 |
|
|
edit(struct diff *diff, int fdup, int j) |
773 |
|
|
{ |
774 |
|
|
if (((fdup + 1) & eflag) == 0) |
775 |
|
|
return (j); |
776 |
|
|
j++; |
777 |
|
|
overlap[j] = !fdup; |
778 |
|
|
if (!fdup) |
779 |
|
|
overlapcnt++; |
780 |
|
|
de[j].old.from = diff->old.from; |
781 |
|
|
de[j].old.to = diff->old.to; |
782 |
|
|
de[j].new.from = de[j-1].new.to + skip(2, diff->new.from, NULL); |
783 |
|
|
de[j].new.to = de[j].new.from + skip(2, diff->new.to, NULL); |
784 |
|
|
return (j); |
785 |
|
|
} |
786 |
|
|
|
787 |
|
|
/* regurgitate */ |
788 |
|
|
static int |
789 |
|
|
edscript(int n) |
790 |
|
|
{ |
791 |
|
|
int j, k; |
792 |
|
|
char block[BUFSIZ+1]; |
793 |
|
|
|
794 |
|
|
for (; n > 0; n--) { |
795 |
|
|
if (!oflag || !overlap[n]) |
796 |
|
|
prange(&de[n].old); |
797 |
|
|
else |
798 |
|
|
diff_output("%da\n=======\n", de[n].old.to -1); |
799 |
|
|
(void)fseek(fp[2], (long)de[n].new.from, SEEK_SET); |
800 |
|
|
for (k = de[n].new.to-de[n].new.from; k > 0; k-= j) { |
801 |
|
|
j = k > BUFSIZ ? BUFSIZ : k; |
802 |
|
|
if (fread(block, 1, j, fp[2]) != (size_t)j) |
803 |
|
|
return (-1); |
804 |
|
|
block[j] = '\0'; |
805 |
|
|
diff_output("%s", block); |
806 |
|
|
} |
807 |
|
|
|
808 |
|
|
if (!oflag || !overlap[n]) |
809 |
|
|
diff_output(".\n"); |
810 |
|
|
else { |
811 |
|
|
diff_output("%s\n.\n", f3mark); |
812 |
|
|
diff_output("%da\n%s\n.\n", de[n].old.from - 1, f1mark); |
813 |
|
|
} |
814 |
|
|
} |
815 |
|
|
|
816 |
|
|
return (overlapcnt); |
817 |
|
|
} |
818 |
|
|
|
819 |
|
|
static void |
820 |
|
|
increase(void) |
821 |
|
|
{ |
822 |
|
|
size_t newsz, incr; |
823 |
|
|
|
824 |
|
|
/* are the memset(3) calls needed? */ |
825 |
|
|
newsz = szchanges == 0 ? 64 : 2 * szchanges; |
826 |
|
|
incr = newsz - szchanges; |
827 |
|
|
|
828 |
|
|
d13 = xreallocarray(d13, newsz, sizeof(*d13)); |
829 |
|
|
memset(d13 + szchanges, 0, incr * sizeof(*d13)); |
830 |
|
|
d23 = xreallocarray(d23, newsz, sizeof(*d23)); |
831 |
|
|
memset(d23 + szchanges, 0, incr * sizeof(*d23)); |
832 |
|
|
de = xreallocarray(de, newsz, sizeof(*de)); |
833 |
|
|
memset(de + szchanges, 0, incr * sizeof(*de)); |
834 |
|
|
overlap = xreallocarray(overlap, newsz, sizeof(*overlap)); |
835 |
|
|
memset(overlap + szchanges, 0, incr * sizeof(*overlap)); |
836 |
|
|
szchanges = newsz; |
837 |
|
|
} |