xdiff/xmerge.c

   1 /*
   2  *  LibXDiff by Davide Libenzi ( File Differential Library )
   3  *  Copyright (C) 2003-2006 Davide Libenzi, Johannes E. Schindelin
   4  *
   5  *  This library is free software; you can redistribute it and/or
   6  *  modify it under the terms of the GNU Lesser General Public
   7  *  License as published by the Free Software Foundation; either
   8  *  version 2.1 of the License, or (at your option) any later version.
   9  *
  10  *  This library is distributed in the hope that it will be useful,
  11  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13  *  Lesser General Public License for more details.
  14  *
  15  *  You should have received a copy of the GNU Lesser General Public
  16  *  License along with this library; if not, write to the Free Software
  17  *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  18  *
  19  *  Davide Libenzi <davidel@xmailserver.org>
  20  *
  21  */
  22
  23 #include "xinclude.h"
  24
  25 typedef struct s_xdmerge {
  26         struct s_xdmerge *next;
  27         /*
  28          * 0 = conflict,
  29          * 1 = no conflict, take first,
  30          * 2 = no conflict, take second.
  31          * 3 = no conflict, take both.
  32          */
  33         int mode;
  34         /*
  35          * These point at the respective postimages.  E.g. <i1,chg1> is
  36          * how side #1 wants to change the common ancestor; if there is no
  37          * overlap, lines before i1 in the postimage of side #1 appear
  38          * in the merge result as a region touched by neither side.
  39          */
  40         long i1, i2;
  41         long chg1, chg2;
  42         /*
  43          * These point at the preimage; of course there is just one
  44          * preimage, that is from the shared common ancestor.
  45          */
  46         long i0;
  47         long chg0;
  48 } xdmerge_t;
  49
  50 static int xdl_append_merge(xdmerge_t **merge, int mode,
  51                             long i0, long chg0,
  52                             long i1, long chg1,
  53                             long i2, long chg2)
  54 {
  55         xdmerge_t *m = *merge;
  56         if (m && (i1 <= m->i1 + m->chg1 || i2 <= m->i2 + m->chg2)) {
  57                 if (mode != m->mode)
  58                         m->mode = 0;
  59                 m->chg0 = i0 + chg0 - m->i0;
  60                 m->chg1 = i1 + chg1 - m->i1;
  61                 m->chg2 = i2 + chg2 - m->i2;
  62         } else {
  63                 m = xdl_malloc(sizeof(xdmerge_t));
  64                 if (!m)
  65                         return -1;
  66                 m->next = NULL;
  67                 m->mode = mode;
  68                 m->i0 = i0;
  69                 m->chg0 = chg0;
  70                 m->i1 = i1;
  71                 m->chg1 = chg1;
  72                 m->i2 = i2;
  73                 m->chg2 = chg2;
  74                 if (*merge)
  75                         (*merge)->next = m;
  76                 *merge = m;
  77         }
  78         return 0;
  79 }
  80
  81 static int xdl_cleanup_merge(xdmerge_t *c)
  82 {
  83         int count = 0;
  84         xdmerge_t *next_c;
  85
  86         /* were there conflicts? */
  87         for (; c; c = next_c) {
  88                 if (c->mode == 0)
  89                         count++;
  90                 next_c = c->next;
  91                 free(c);
  92         }
  93         return count;
  94 }
  95
  96 static int xdl_merge_cmp_lines(xdfenv_t *xe1, int i1, xdfenv_t *xe2, int i2,
  97                 int line_count, long flags)
  98 {
  99         int i;
 100         xrecord_t **rec1 = xe1->xdf2.recs + i1;
 101         xrecord_t **rec2 = xe2->xdf2.recs + i2;
 102
 103         for (i = 0; i < line_count; i++) {
 104                 int result = xdl_recmatch(rec1[i]->ptr, rec1[i]->size,
 105                         rec2[i]->ptr, rec2[i]->size, flags);
 106                 if (!result)
 107                         return -1;
 108         }
 109         return 0;
 110 }
 111
 112 static int xdl_recs_copy_0(int use_orig, xdfenv_t *xe, int i, int count, int add_nl, char *dest)
 113 {
 114         xrecord_t **recs;
 115         int size = 0;
 116
 117         recs = (use_orig ? xe->xdf1.recs : xe->xdf2.recs) + i;
 118
 119         if (count < 1)
 120                 return 0;
 121
 122         for (i = 0; i < count; size += recs[i++]->size)
 123                 if (dest)
 124                         memcpy(dest + size, recs[i]->ptr, recs[i]->size);
 125         if (add_nl) {
 126                 i = recs[count - 1]->size;
 127                 if (i == 0 || recs[count - 1]->ptr[i - 1] != '\n') {
 128                         if (dest)
 129                                 dest[size] = '\n';
 130                         size++;
 131                 }
 132         }
 133         return size;
 134 }
 135
 136 static int xdl_recs_copy(xdfenv_t *xe, int i, int count, int add_nl, char *dest)
 137 {
 138         return xdl_recs_copy_0(0, xe, i, count, add_nl, dest);
 139 }
 140
 141 static int xdl_orig_copy(xdfenv_t *xe, int i, int count, int add_nl, char *dest)
 142 {
 143         return xdl_recs_copy_0(1, xe, i, count, add_nl, dest);
 144 }
 145
 146 static int fill_conflict_hunk(xdfenv_t *xe1, const char *name1,
 147                               xdfenv_t *xe2, const char *name2,
 148                               const char *name3,
 149                               int size, int i, int style,
 150                               xdmerge_t *m, char *dest, int marker_size)
 151 {
 152         int marker1_size = (name1 ? strlen(name1) + 1 : 0);
 153         int marker2_size = (name2 ? strlen(name2) + 1 : 0);
 154         int marker3_size = (name3 ? strlen(name3) + 1 : 0);
 155
 156         if (marker_size <= 0)
 157                 marker_size = DEFAULT_CONFLICT_MARKER_SIZE;
 158
 159         /* Before conflicting part */
 160         size += xdl_recs_copy(xe1, i, m->i1 - i, 0,
 161                               dest ? dest + size : NULL);
 162
 163         if (!dest) {
 164                 size += marker_size + 1 + marker1_size;
 165         } else {
 166                 memset(dest + size, '<', marker_size);
 167                 size += marker_size;
 168                 if (marker1_size) {
 169                         dest[size] = ' ';
 170                         memcpy(dest + size + 1, name1, marker1_size - 1);
 171                         size += marker1_size;
 172                 }
 173                 dest[size++] = '\n';
 174         }
 175
 176         /* Postimage from side #1 */
 177         size += xdl_recs_copy(xe1, m->i1, m->chg1, 1,
 178                               dest ? dest + size : NULL);
 179
 180         if (style == XDL_MERGE_DIFF3) {
 181                 /* Shared preimage */
 182                 if (!dest) {
 183                         size += marker_size + 1 + marker3_size;
 184                 } else {
 185                         memset(dest + size, '|', marker_size);
 186                         size += marker_size;
 187                         if (marker3_size) {
 188                                 dest[size] = ' ';
 189                                 memcpy(dest + size + 1, name3, marker3_size - 1);
 190                                 size += marker3_size;
 191                         }
 192                         dest[size++] = '\n';
 193                 }
 194                 size += xdl_orig_copy(xe1, m->i0, m->chg0, 1,
 195                                       dest ? dest + size : NULL);
 196         }
 197
 198         if (!dest) {
 199                 size += marker_size + 1;
 200         } else {
 201                 memset(dest + size, '=', marker_size);
 202                 size += marker_size;
 203                 dest[size++] = '\n';
 204         }
 205
 206         /* Postimage from side #2 */
 207         size += xdl_recs_copy(xe2, m->i2, m->chg2, 1,
 208                               dest ? dest + size : NULL);
 209         if (!dest) {
 210                 size += marker_size + 1 + marker2_size;
 211         } else {
 212                 memset(dest + size, '>', marker_size);
 213                 size += marker_size;
 214                 if (marker2_size) {
 215                         dest[size] = ' ';
 216                         memcpy(dest + size + 1, name2, marker2_size - 1);
 217                         size += marker2_size;
 218                 }
 219                 dest[size++] = '\n';
 220         }
 221         return size;
 222 }
 223
 224 static int xdl_fill_merge_buffer(xdfenv_t *xe1, const char *name1,
 225                                  xdfenv_t *xe2, const char *name2,
 226                                  const char *ancestor_name,
 227                                  int favor,
 228                                  xdmerge_t *m, char *dest, int style,
 229                                  int marker_size)
 230 {
 231         int size, i;
 232
 233         for (size = i = 0; m; m = m->next) {
 234                 if (favor && !m->mode)
 235                         m->mode = favor;
 236
 237                 if (m->mode == 0)
 238                         size = fill_conflict_hunk(xe1, name1, xe2, name2,
 239                                                   ancestor_name,
 240                                                   size, i, style, m, dest,
 241                                                   marker_size);
 242                 else if (m->mode & 3) {
 243                         /* Before conflicting part */
 244                         size += xdl_recs_copy(xe1, i, m->i1 - i, 0,
 245                                               dest ? dest + size : NULL);
 246                         /* Postimage from side #1 */
 247                         if (m->mode & 1)
 248                                 size += xdl_recs_copy(xe1, m->i1, m->chg1, 1,
 249                                                       dest ? dest + size : NULL);
 250                         /* Postimage from side #2 */
 251                         if (m->mode & 2)
 252                                 size += xdl_recs_copy(xe2, m->i2, m->chg2, 1,
 253                                                       dest ? dest + size : NULL);
 254                 } else
 255                         continue;
 256                 i = m->i1 + m->chg1;
 257         }
 258         size += xdl_recs_copy(xe1, i, xe1->xdf2.nrec - i, 0,
 259                               dest ? dest + size : NULL);
 260         return size;
 261 }
 262
 263 /*
 264  * Sometimes, changes are not quite identical, but differ in only a few
 265  * lines. Try hard to show only these few lines as conflicting.
 266  */
 267 static int xdl_refine_conflicts(xdfenv_t *xe1, xdfenv_t *xe2, xdmerge_t *m,
 268                 xpparam_t const *xpp)
 269 {
 270         for (; m; m = m->next) {
 271                 mmfile_t t1, t2;
 272                 xdfenv_t xe;
 273                 xdchange_t *xscr, *x;
 274                 int i1 = m->i1, i2 = m->i2;
 275
 276                 /* let's handle just the conflicts */
 277                 if (m->mode)
 278                         continue;
 279
 280                 /* no sense refining a conflict when one side is empty */
 281                 if (m->chg1 == 0 || m->chg2 == 0)
 282                         continue;
 283
 284                 /*
 285                  * This probably does not work outside git, since
 286                  * we have a very simple mmfile structure.
 287                  */
 288                 t1.ptr = (char *)xe1->xdf2.recs[m->i1]->ptr;
 289                 t1.size = xe1->xdf2.recs[m->i1 + m->chg1 - 1]->ptr
 290                         + xe1->xdf2.recs[m->i1 + m->chg1 - 1]->size - t1.ptr;
 291                 t2.ptr = (char *)xe2->xdf2.recs[m->i2]->ptr;
 292                 t2.size = xe2->xdf2.recs[m->i2 + m->chg2 - 1]->ptr
 293                         + xe2->xdf2.recs[m->i2 + m->chg2 - 1]->size - t2.ptr;
 294                 if (xdl_do_diff(&t1, &t2, xpp, &xe) < 0)
 295                         return -1;
 296                 if (xdl_change_compact(&xe.xdf1, &xe.xdf2, xpp->flags) < 0 ||
 297                     xdl_change_compact(&xe.xdf2, &xe.xdf1, xpp->flags) < 0 ||
 298                     xdl_build_script(&xe, &xscr) < 0) {
 299                         xdl_free_env(&xe);
 300                         return -1;
 301                 }
 302                 if (!xscr) {
 303                         /* If this happens, the changes are identical. */
 304                         xdl_free_env(&xe);
 305                         m->mode = 4;
 306                         continue;
 307                 }
 308                 x = xscr;
 309                 m->i1 = xscr->i1 + i1;
 310                 m->chg1 = xscr->chg1;
 311                 m->i2 = xscr->i2 + i2;
 312                 m->chg2 = xscr->chg2;
 313                 while (xscr->next) {
 314                         xdmerge_t *m2 = xdl_malloc(sizeof(xdmerge_t));
 315                         if (!m2) {
 316                                 xdl_free_env(&xe);
 317                                 xdl_free_script(x);
 318                                 return -1;
 319                         }
 320                         xscr = xscr->next;
 321                         m2->next = m->next;
 322                         m->next = m2;
 323                         m = m2;
 324                         m->mode = 0;
 325                         m->i1 = xscr->i1 + i1;
 326                         m->chg1 = xscr->chg1;
 327                         m->i2 = xscr->i2 + i2;
 328                         m->chg2 = xscr->chg2;
 329                 }
 330                 xdl_free_env(&xe);
 331                 xdl_free_script(x);
 332         }
 333         return 0;
 334 }
 335
 336 static int line_contains_alnum(const char *ptr, long size)
 337 {
 338         while (size--)
 339                 if (isalnum(*(ptr++)))
 340                         return 1;
 341         return 0;
 342 }
 343
 344 static int lines_contain_alnum(xdfenv_t *xe, int i, int chg)
 345 {
 346         for (; chg; chg--, i++)
 347                 if (line_contains_alnum(xe->xdf2.recs[i]->ptr,
 348                                 xe->xdf2.recs[i]->size))
 349                         return 1;
 350         return 0;
 351 }
 352
 353 /*
 354  * This function merges m and m->next, marking everything between those hunks
 355  * as conflicting, too.
 356  */
 357 static void xdl_merge_two_conflicts(xdmerge_t *m)
 358 {
 359         xdmerge_t *next_m = m->next;
 360         m->chg1 = next_m->i1 + next_m->chg1 - m->i1;
 361         m->chg2 = next_m->i2 + next_m->chg2 - m->i2;
 362         m->next = next_m->next;
 363         free(next_m);
 364 }
 365
 366 /*
 367  * If there are less than 3 non-conflicting lines between conflicts,
 368  * it appears simpler -- because it takes up less (or as many) lines --
 369  * if the lines are moved into the conflicts.
 370  */
 371 static int xdl_simplify_non_conflicts(xdfenv_t *xe1, xdmerge_t *m,
 372                                       int simplify_if_no_alnum)
 373 {
 374         int result = 0;
 375
 376         if (!m)
 377                 return result;
 378         for (;;) {
 379                 xdmerge_t *next_m = m->next;
 380                 int begin, end;
 381
 382                 if (!next_m)
 383                         return result;
 384
 385                 begin = m->i1 + m->chg1;
 386                 end = next_m->i1;
 387
 388                 if (m->mode != 0 || next_m->mode != 0 ||
 389                     (end - begin > 3 &&
 390                      (!simplify_if_no_alnum ||
 391                       lines_contain_alnum(xe1, begin, end - begin)))) {
 392                         m = next_m;
 393                 } else {
 394                         result++;
 395                         xdl_merge_two_conflicts(m);
 396                 }
 397         }
 398 }
 399
 400 /*
 401  * level == 0: mark all overlapping changes as conflict
 402  * level == 1: mark overlapping changes as conflict only if not identical
 403  * level == 2: analyze non-identical changes for minimal conflict set
 404  * level == 3: analyze non-identical changes for minimal conflict set, but
 405  *             treat hunks not containing any letter or number as conflicting
 406  *
 407  * returns < 0 on error, == 0 for no conflicts, else number of conflicts
 408  */
 409 static int xdl_do_merge(xdfenv_t *xe1, xdchange_t *xscr1,
 410                 xdfenv_t *xe2, xdchange_t *xscr2,
 411                 xmparam_t const *xmp, mmbuffer_t *result)
 412 {
 413         xdmerge_t *changes, *c;
 414         xpparam_t const *xpp = &xmp->xpp;
 415         const char *const ancestor_name = xmp->ancestor;
 416         const char *const name1 = xmp->file1;
 417         const char *const name2 = xmp->file2;
 418         int i0, i1, i2, chg0, chg1, chg2;
 419         int level = xmp->level;
 420         int style = xmp->style;
 421         int favor = xmp->favor;
 422
 423         if (style == XDL_MERGE_DIFF3) {
 424                 /*
 425                  * "diff3 -m" output does not make sense for anything
 426                  * more aggressive than XDL_MERGE_EAGER.
 427                  */
 428                 if (XDL_MERGE_EAGER < level)
 429                         level = XDL_MERGE_EAGER;
 430         }
 431
 432         c = changes = NULL;
 433
 434         while (xscr1 && xscr2) {
 435                 if (!changes)
 436                         changes = c;
 437                 if (xscr1->i1 + xscr1->chg1 < xscr2->i1) {
 438                         i0 = xscr1->i1;
 439                         i1 = xscr1->i2;
 440                         i2 = xscr2->i2 - xscr2->i1 + xscr1->i1;
 441                         chg0 = xscr1->chg1;
 442                         chg1 = xscr1->chg2;
 443                         chg2 = xscr1->chg1;
 444                         if (xdl_append_merge(&c, 1,
 445                                              i0, chg0, i1, chg1, i2, chg2)) {
 446                                 xdl_cleanup_merge(changes);
 447                                 return -1;
 448                         }
 449                         xscr1 = xscr1->next;
 450                         continue;
 451                 }
 452                 if (xscr2->i1 + xscr2->chg1 < xscr1->i1) {
 453                         i0 = xscr2->i1;
 454                         i1 = xscr1->i2 - xscr1->i1 + xscr2->i1;
 455                         i2 = xscr2->i2;
 456                         chg0 = xscr2->chg1;
 457                         chg1 = xscr2->chg1;
 458                         chg2 = xscr2->chg2;
 459                         if (xdl_append_merge(&c, 2,
 460                                              i0, chg0, i1, chg1, i2, chg2)) {
 461                                 xdl_cleanup_merge(changes);
 462                                 return -1;
 463                         }
 464                         xscr2 = xscr2->next;
 465                         continue;
 466                 }
 467                 if (level == XDL_MERGE_MINIMAL || xscr1->i1 != xscr2->i1 ||
 468                                 xscr1->chg1 != xscr2->chg1 ||
 469                                 xscr1->chg2 != xscr2->chg2 ||
 470                                 xdl_merge_cmp_lines(xe1, xscr1->i2,
 471                                         xe2, xscr2->i2,
 472                                         xscr1->chg2, xpp->flags)) {
 473                         /* conflict */
 474                         int off = xscr1->i1 - xscr2->i1;
 475                         int ffo = off + xscr1->chg1 - xscr2->chg1;
 476
 477                         i0 = xscr1->i1;
 478                         i1 = xscr1->i2;
 479                         i2 = xscr2->i2;
 480                         if (off > 0) {
 481                                 i0 -= off;
 482                                 i1 -= off;
 483                         }
 484                         else
 485                                 i2 += off;
 486                         chg0 = xscr1->i1 + xscr1->chg1 - i0;
 487                         chg1 = xscr1->i2 + xscr1->chg2 - i1;
 488                         chg2 = xscr2->i2 + xscr2->chg2 - i2;
 489                         if (ffo < 0) {
 490                                 chg0 -= ffo;
 491                                 chg1 -= ffo;
 492                         } else
 493                                 chg2 += ffo;
 494                         if (xdl_append_merge(&c, 0,
 495                                              i0, chg0, i1, chg1, i2, chg2)) {
 496                                 xdl_cleanup_merge(changes);
 497                                 return -1;
 498                         }
 499                 }
 500
 501                 i1 = xscr1->i1 + xscr1->chg1;
 502                 i2 = xscr2->i1 + xscr2->chg1;
 503
 504                 if (i1 >= i2)
 505                         xscr2 = xscr2->next;
 506                 if (i2 >= i1)
 507                         xscr1 = xscr1->next;
 508         }
 509         while (xscr1) {
 510                 if (!changes)
 511                         changes = c;
 512                 i0 = xscr1->i1;
 513                 i1 = xscr1->i2;
 514                 i2 = xscr1->i1 + xe2->xdf2.nrec - xe2->xdf1.nrec;
 515                 chg0 = xscr1->chg1;
 516                 chg1 = xscr1->chg2;
 517                 chg2 = xscr1->chg1;
 518                 if (xdl_append_merge(&c, 1,
 519                                      i0, chg0, i1, chg1, i2, chg2)) {
 520                         xdl_cleanup_merge(changes);
 521                         return -1;
 522                 }
 523                 xscr1 = xscr1->next;
 524         }
 525         while (xscr2) {
 526                 if (!changes)
 527                         changes = c;
 528                 i0 = xscr2->i1;
 529                 i1 = xscr2->i1 + xe1->xdf2.nrec - xe1->xdf1.nrec;
 530                 i2 = xscr2->i2;
 531                 chg0 = xscr2->chg1;
 532                 chg1 = xscr2->chg1;
 533                 chg2 = xscr2->chg2;
 534                 if (xdl_append_merge(&c, 2,
 535                                      i0, chg0, i1, chg1, i2, chg2)) {
 536                         xdl_cleanup_merge(changes);
 537                         return -1;
 538                 }
 539                 xscr2 = xscr2->next;
 540         }
 541         if (!changes)
 542                 changes = c;
 543         /* refine conflicts */
 544         if (XDL_MERGE_ZEALOUS <= level &&
 545             (xdl_refine_conflicts(xe1, xe2, changes, xpp) < 0 ||
 546              xdl_simplify_non_conflicts(xe1, changes,
 547                                         XDL_MERGE_ZEALOUS < level) < 0)) {
 548                 xdl_cleanup_merge(changes);
 549                 return -1;
 550         }
 551         /* output */
 552         if (result) {
 553                 int marker_size = xmp->marker_size;
 554                 int size = xdl_fill_merge_buffer(xe1, name1, xe2, name2,
 555                                                  ancestor_name,
 556                                                  favor, changes, NULL, style,
 557                                                  marker_size);
 558                 result->ptr = xdl_malloc(size);
 559                 if (!result->ptr) {
 560                         xdl_cleanup_merge(changes);
 561                         return -1;
 562                 }
 563                 result->size = size;
 564                 xdl_fill_merge_buffer(xe1, name1, xe2, name2,
 565                                       ancestor_name, favor, changes,
 566                                       result->ptr, style, marker_size);
 567         }
 568         return xdl_cleanup_merge(changes);
 569 }
 570
 571 int xdl_merge(mmfile_t *orig, mmfile_t *mf1, mmfile_t *mf2,
 572                 xmparam_t const *xmp, mmbuffer_t *result)
 573 {
 574         xdchange_t *xscr1, *xscr2;
 575         xdfenv_t xe1, xe2;
 576         int status;
 577         xpparam_t const *xpp = &xmp->xpp;
 578
 579         result->ptr = NULL;
 580         result->size = 0;
 581
 582         if (xdl_do_diff(orig, mf1, xpp, &xe1) < 0 ||
 583                         xdl_do_diff(orig, mf2, xpp, &xe2) < 0) {
 584                 return -1;
 585         }
 586         if (xdl_change_compact(&xe1.xdf1, &xe1.xdf2, xpp->flags) < 0 ||
 587             xdl_change_compact(&xe1.xdf2, &xe1.xdf1, xpp->flags) < 0 ||
 588             xdl_build_script(&xe1, &xscr1) < 0) {
 589                 xdl_free_env(&xe1);
 590                 return -1;
 591         }
 592         if (xdl_change_compact(&xe2.xdf1, &xe2.xdf2, xpp->flags) < 0 ||
 593             xdl_change_compact(&xe2.xdf2, &xe2.xdf1, xpp->flags) < 0 ||
 594             xdl_build_script(&xe2, &xscr2) < 0) {
 595                 xdl_free_env(&xe2);
 596                 return -1;
 597         }
 598         status = 0;
 599         if (!xscr1) {
 600                 result->ptr = xdl_malloc(mf2->size);
 601                 memcpy(result->ptr, mf2->ptr, mf2->size);
 602                 result->size = mf2->size;
 603         } else if (!xscr2) {
 604                 result->ptr = xdl_malloc(mf1->size);
 605                 memcpy(result->ptr, mf1->ptr, mf1->size);
 606                 result->size = mf1->size;
 607         } else {
 608                 status = xdl_do_merge(&xe1, xscr1,
 609                                       &xe2, xscr2,
 610                                       xmp, result);
 611         }
 612         xdl_free_script(xscr1);
 613         xdl_free_script(xscr2);
 614
 615         xdl_free_env(&xe1);
 616         xdl_free_env(&xe2);
 617
 618         return status;
 619 }