xdiff/xprepare.c

   1 /*
   2  *  LibXDiff by Davide Libenzi ( File Differential Library )
   3  *  Copyright (C) 2003  Davide Libenzi
   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
  26
  27 #define XDL_KPDIS_RUN 4
  28 #define XDL_MAX_EQLIMIT 1024
  29
  30
  31
  32 typedef struct s_xdlclass {
  33         struct s_xdlclass *next;
  34         unsigned long ha;
  35         char const *line;
  36         long size;
  37         long idx;
  38 } xdlclass_t;
  39
  40 typedef struct s_xdlclassifier {
  41         unsigned int hbits;
  42         long hsize;
  43         xdlclass_t **rchash;
  44         chastore_t ncha;
  45         long count;
  46 } xdlclassifier_t;
  47
  48
  49
  50
  51 static int xdl_init_classifier(xdlclassifier_t *cf, long size);
  52 static void xdl_free_classifier(xdlclassifier_t *cf);
  53 static int xdl_classify_record(xdlclassifier_t *cf, xrecord_t **rhash, unsigned int hbits,
  54                                xrecord_t *rec);
  55 static int xdl_prepare_ctx(mmfile_t *mf, long narec, xpparam_t const *xpp,
  56                            xdlclassifier_t *cf, xdfile_t *xdf);
  57 static void xdl_free_ctx(xdfile_t *xdf);
  58 static int xdl_clean_mmatch(char const *dis, long i, long s, long e);
  59 static int xdl_cleanup_records(xdfile_t *xdf1, xdfile_t *xdf2);
  60 static int xdl_trim_ends(xdfile_t *xdf1, xdfile_t *xdf2);
  61 static int xdl_optimize_ctxs(xdfile_t *xdf1, xdfile_t *xdf2);
  62
  63
  64
  65
  66 static int xdl_init_classifier(xdlclassifier_t *cf, long size) {
  67         long i;
  68
  69         cf->hbits = xdl_hashbits((unsigned int) size);
  70         cf->hsize = 1 << cf->hbits;
  71
  72         if (xdl_cha_init(&cf->ncha, sizeof(xdlclass_t), size / 4 + 1) < 0) {
  73
  74                 return -1;
  75         }
  76         if (!(cf->rchash = (xdlclass_t **) xdl_malloc(cf->hsize * sizeof(xdlclass_t *)))) {
  77
  78                 xdl_cha_free(&cf->ncha);
  79                 return -1;
  80         }
  81         for (i = 0; i < cf->hsize; i++)
  82                 cf->rchash[i] = NULL;
  83
  84         cf->count = 0;
  85
  86         return 0;
  87 }
  88
  89
  90 static void xdl_free_classifier(xdlclassifier_t *cf) {
  91
  92         xdl_free(cf->rchash);
  93         xdl_cha_free(&cf->ncha);
  94 }
  95
  96
  97 static int xdl_classify_record(xdlclassifier_t *cf, xrecord_t **rhash, unsigned int hbits,
  98                                xrecord_t *rec) {
  99         long hi;
 100         char const *line;
 101         xdlclass_t *rcrec;
 102
 103         line = rec->ptr;
 104         hi = (long) XDL_HASHLONG(rec->ha, cf->hbits);
 105         for (rcrec = cf->rchash[hi]; rcrec; rcrec = rcrec->next)
 106                 if (rcrec->ha == rec->ha && rcrec->size == rec->size &&
 107                     !memcmp(line, rcrec->line, rec->size))
 108                         break;
 109
 110         if (!rcrec) {
 111                 if (!(rcrec = xdl_cha_alloc(&cf->ncha))) {
 112
 113                         return -1;
 114                 }
 115                 rcrec->idx = cf->count++;
 116                 rcrec->line = line;
 117                 rcrec->size = rec->size;
 118                 rcrec->ha = rec->ha;
 119                 rcrec->next = cf->rchash[hi];
 120                 cf->rchash[hi] = rcrec;
 121         }
 122
 123         rec->ha = (unsigned long) rcrec->idx;
 124
 125         hi = (long) XDL_HASHLONG(rec->ha, hbits);
 126         rec->next = rhash[hi];
 127         rhash[hi] = rec;
 128
 129         return 0;
 130 }
 131
 132
 133 static int xdl_prepare_ctx(mmfile_t *mf, long narec, xpparam_t const *xpp,
 134                            xdlclassifier_t *cf, xdfile_t *xdf) {
 135         unsigned int hbits;
 136         long i, nrec, hsize, bsize;
 137         unsigned long hav;
 138         char const *blk, *cur, *top, *prev;
 139         xrecord_t *crec;
 140         xrecord_t **recs, **rrecs;
 141         xrecord_t **rhash;
 142         unsigned long *ha;
 143         char *rchg;
 144         long *rindex;
 145
 146         if (xdl_cha_init(&xdf->rcha, sizeof(xrecord_t), narec / 4 + 1) < 0) {
 147
 148                 return -1;
 149         }
 150         if (!(recs = (xrecord_t **) xdl_malloc(narec * sizeof(xrecord_t *)))) {
 151
 152                 xdl_cha_free(&xdf->rcha);
 153                 return -1;
 154         }
 155
 156         hbits = xdl_hashbits((unsigned int) narec);
 157         hsize = 1 << hbits;
 158         if (!(rhash = (xrecord_t **) xdl_malloc(hsize * sizeof(xrecord_t *)))) {
 159
 160                 xdl_free(recs);
 161                 xdl_cha_free(&xdf->rcha);
 162                 return -1;
 163         }
 164         for (i = 0; i < hsize; i++)
 165                 rhash[i] = NULL;
 166
 167         nrec = 0;
 168         if ((cur = blk = xdl_mmfile_first(mf, &bsize)) != NULL) {
 169                 for (top = blk + bsize;;) {
 170                         if (cur >= top) {
 171                                 if (!(cur = blk = xdl_mmfile_next(mf, &bsize)))
 172                                         break;
 173                                 top = blk + bsize;
 174                         }
 175                         prev = cur;
 176                         hav = xdl_hash_record(&cur, top);
 177                         if (nrec >= narec) {
 178                                 narec *= 2;
 179                                 if (!(rrecs = (xrecord_t **) xdl_realloc(recs, narec * sizeof(xrecord_t *)))) {
 180
 181                                         xdl_free(rhash);
 182                                         xdl_free(recs);
 183                                         xdl_cha_free(&xdf->rcha);
 184                                         return -1;
 185                                 }
 186                                 recs = rrecs;
 187                         }
 188                         if (!(crec = xdl_cha_alloc(&xdf->rcha))) {
 189
 190                                 xdl_free(rhash);
 191                                 xdl_free(recs);
 192                                 xdl_cha_free(&xdf->rcha);
 193                                 return -1;
 194                         }
 195                         crec->ptr = prev;
 196                         crec->size = (long) (cur - prev);
 197                         crec->ha = hav;
 198                         recs[nrec++] = crec;
 199
 200                         if (xdl_classify_record(cf, rhash, hbits, crec) < 0) {
 201
 202                                 xdl_free(rhash);
 203                                 xdl_free(recs);
 204                                 xdl_cha_free(&xdf->rcha);
 205                                 return -1;
 206                         }
 207                 }
 208         }
 209
 210         if (!(rchg = (char *) xdl_malloc((nrec + 2) * sizeof(char)))) {
 211
 212                 xdl_free(rhash);
 213                 xdl_free(recs);
 214                 xdl_cha_free(&xdf->rcha);
 215                 return -1;
 216         }
 217         memset(rchg, 0, (nrec + 2) * sizeof(char));
 218
 219         if (!(rindex = (long *) xdl_malloc((nrec + 1) * sizeof(long)))) {
 220
 221                 xdl_free(rchg);
 222                 xdl_free(rhash);
 223                 xdl_free(recs);
 224                 xdl_cha_free(&xdf->rcha);
 225                 return -1;
 226         }
 227         if (!(ha = (unsigned long *) xdl_malloc((nrec + 1) * sizeof(unsigned long)))) {
 228
 229                 xdl_free(rindex);
 230                 xdl_free(rchg);
 231                 xdl_free(rhash);
 232                 xdl_free(recs);
 233                 xdl_cha_free(&xdf->rcha);
 234                 return -1;
 235         }
 236
 237         xdf->nrec = nrec;
 238         xdf->recs = recs;
 239         xdf->hbits = hbits;
 240         xdf->rhash = rhash;
 241         xdf->rchg = rchg + 1;
 242         xdf->rindex = rindex;
 243         xdf->nreff = 0;
 244         xdf->ha = ha;
 245         xdf->dstart = 0;
 246         xdf->dend = nrec - 1;
 247
 248         return 0;
 249 }
 250
 251
 252 static void xdl_free_ctx(xdfile_t *xdf) {
 253
 254         xdl_free(xdf->rhash);
 255         xdl_free(xdf->rindex);
 256         xdl_free(xdf->rchg - 1);
 257         xdl_free(xdf->ha);
 258         xdl_free(xdf->recs);
 259         xdl_cha_free(&xdf->rcha);
 260 }
 261
 262
 263 int xdl_prepare_env(mmfile_t *mf1, mmfile_t *mf2, xpparam_t const *xpp,
 264                     xdfenv_t *xe) {
 265         long enl1, enl2;
 266         xdlclassifier_t cf;
 267
 268         enl1 = xdl_guess_lines(mf1) + 1;
 269         enl2 = xdl_guess_lines(mf2) + 1;
 270
 271         if (xdl_init_classifier(&cf, enl1 + enl2 + 1) < 0) {
 272
 273                 return -1;
 274         }
 275
 276         if (xdl_prepare_ctx(mf1, enl1, xpp, &cf, &xe->xdf1) < 0) {
 277
 278                 xdl_free_classifier(&cf);
 279                 return -1;
 280         }
 281         if (xdl_prepare_ctx(mf2, enl2, xpp, &cf, &xe->xdf2) < 0) {
 282
 283                 xdl_free_ctx(&xe->xdf1);
 284                 xdl_free_classifier(&cf);
 285                 return -1;
 286         }
 287
 288         xdl_free_classifier(&cf);
 289
 290         if (xdl_optimize_ctxs(&xe->xdf1, &xe->xdf2) < 0) {
 291
 292                 xdl_free_ctx(&xe->xdf2);
 293                 xdl_free_ctx(&xe->xdf1);
 294                 return -1;
 295         }
 296
 297         return 0;
 298 }
 299
 300
 301 void xdl_free_env(xdfenv_t *xe) {
 302
 303         xdl_free_ctx(&xe->xdf2);
 304         xdl_free_ctx(&xe->xdf1);
 305 }
 306
 307
 308 static int xdl_clean_mmatch(char const *dis, long i, long s, long e) {
 309         long r, rdis0, rpdis0, rdis1, rpdis1;
 310
 311         /*
 312          * Scans the lines before 'i' to find a run of lines that either
 313          * have no match (dis[j] == 0) or have multiple matches (dis[j] > 1).
 314          * Note that we always call this function with dis[i] > 1, so the
 315          * current line (i) is already a multimatch line.
 316          */
 317         for (r = 1, rdis0 = 0, rpdis0 = 1; (i - r) >= s; r++) {
 318                 if (!dis[i - r])
 319                         rdis0++;
 320                 else if (dis[i - r] == 2)
 321                         rpdis0++;
 322                 else
 323                         break;
 324         }
 325         /*
 326          * If the run before the line 'i' found only multimatch lines, we
 327          * return 0 and hence we don't make the current line (i) discarded.
 328          * We want to discard multimatch lines only when they appear in the
 329          * middle of runs with nomatch lines (dis[j] == 0).
 330          */
 331         if (rdis0 == 0)
 332                 return 0;
 333         for (r = 1, rdis1 = 0, rpdis1 = 1; (i + r) <= e; r++) {
 334                 if (!dis[i + r])
 335                         rdis1++;
 336                 else if (dis[i + r] == 2)
 337                         rpdis1++;
 338                 else
 339                         break;
 340         }
 341         /*
 342          * If the run after the line 'i' found only multimatch lines, we
 343          * return 0 and hence we don't make the current line (i) discarded.
 344          */
 345         if (rdis1 == 0)
 346                 return 0;
 347         rdis1 += rdis0;
 348         rpdis1 += rpdis0;
 349
 350         return rpdis1 * XDL_KPDIS_RUN < (rpdis1 + rdis1);
 351 }
 352
 353
 354 /*
 355  * Try to reduce the problem complexity, discard records that have no
 356  * matches on the other file. Also, lines that have multiple matches
 357  * might be potentially discarded if they happear in a run of discardable.
 358  */
 359 static int xdl_cleanup_records(xdfile_t *xdf1, xdfile_t *xdf2) {
 360         long i, nm, rhi, nreff, mlim;
 361         unsigned long hav;
 362         xrecord_t **recs;
 363         xrecord_t *rec;
 364         char *dis, *dis1, *dis2;
 365
 366         if (!(dis = (char *) xdl_malloc(xdf1->nrec + xdf2->nrec + 2))) {
 367
 368                 return -1;
 369         }
 370         memset(dis, 0, xdf1->nrec + xdf2->nrec + 2);
 371         dis1 = dis;
 372         dis2 = dis1 + xdf1->nrec + 1;
 373
 374         if ((mlim = xdl_bogosqrt(xdf1->nrec)) > XDL_MAX_EQLIMIT)
 375                 mlim = XDL_MAX_EQLIMIT;
 376         for (i = xdf1->dstart, recs = &xdf1->recs[xdf1->dstart]; i <= xdf1->dend; i++, recs++) {
 377                 hav = (*recs)->ha;
 378                 rhi = (long) XDL_HASHLONG(hav, xdf2->hbits);
 379                 for (nm = 0, rec = xdf2->rhash[rhi]; rec; rec = rec->next)
 380                         if (rec->ha == hav && ++nm == mlim)
 381                                 break;
 382                 dis1[i] = (nm == 0) ? 0: (nm >= mlim) ? 2: 1;
 383         }
 384
 385         if ((mlim = xdl_bogosqrt(xdf2->nrec)) > XDL_MAX_EQLIMIT)
 386                 mlim = XDL_MAX_EQLIMIT;
 387         for (i = xdf2->dstart, recs = &xdf2->recs[xdf2->dstart]; i <= xdf2->dend; i++, recs++) {
 388                 hav = (*recs)->ha;
 389                 rhi = (long) XDL_HASHLONG(hav, xdf1->hbits);
 390                 for (nm = 0, rec = xdf1->rhash[rhi]; rec; rec = rec->next)
 391                         if (rec->ha == hav && ++nm == mlim)
 392                                 break;
 393                 dis2[i] = (nm == 0) ? 0: (nm >= mlim) ? 2: 1;
 394         }
 395
 396         for (nreff = 0, i = xdf1->dstart, recs = &xdf1->recs[xdf1->dstart];
 397              i <= xdf1->dend; i++, recs++) {
 398                 if (dis1[i] == 1 ||
 399                     (dis1[i] == 2 && !xdl_clean_mmatch(dis1, i, xdf1->dstart, xdf1->dend))) {
 400                         xdf1->rindex[nreff] = i;
 401                         xdf1->ha[nreff] = (*recs)->ha;
 402                         nreff++;
 403                 } else
 404                         xdf1->rchg[i] = 1;
 405         }
 406         xdf1->nreff = nreff;
 407
 408         for (nreff = 0, i = xdf2->dstart, recs = &xdf2->recs[xdf2->dstart];
 409              i <= xdf2->dend; i++, recs++) {
 410                 if (dis2[i] == 1 ||
 411                     (dis2[i] == 2 && !xdl_clean_mmatch(dis2, i, xdf2->dstart, xdf2->dend))) {
 412                         xdf2->rindex[nreff] = i;
 413                         xdf2->ha[nreff] = (*recs)->ha;
 414                         nreff++;
 415                 } else
 416                         xdf2->rchg[i] = 1;
 417         }
 418         xdf2->nreff = nreff;
 419
 420         xdl_free(dis);
 421
 422         return 0;
 423 }
 424
 425
 426 /*
 427  * Early trim initial and terminal matching records.
 428  */
 429 static int xdl_trim_ends(xdfile_t *xdf1, xdfile_t *xdf2) {
 430         long i, lim;
 431         xrecord_t **recs1, **recs2;
 432
 433         recs1 = xdf1->recs;
 434         recs2 = xdf2->recs;
 435         for (i = 0, lim = XDL_MIN(xdf1->nrec, xdf2->nrec); i < lim;
 436              i++, recs1++, recs2++)
 437                 if ((*recs1)->ha != (*recs2)->ha)
 438                         break;
 439
 440         xdf1->dstart = xdf2->dstart = i;
 441
 442         recs1 = xdf1->recs + xdf1->nrec - 1;
 443         recs2 = xdf2->recs + xdf2->nrec - 1;
 444         for (lim -= i, i = 0; i < lim; i++, recs1--, recs2--)
 445                 if ((*recs1)->ha != (*recs2)->ha)
 446                         break;
 447
 448         xdf1->dend = xdf1->nrec - i - 1;
 449         xdf2->dend = xdf2->nrec - i - 1;
 450
 451         return 0;
 452 }
 453
 454
 455 static int xdl_optimize_ctxs(xdfile_t *xdf1, xdfile_t *xdf2) {
 456
 457         if (xdl_trim_ends(xdf1, xdf2) < 0 ||
 458             xdl_cleanup_records(xdf1, xdf2) < 0) {
 459
 460                 return -1;
 461         }
 462
 463         return 0;
 464 }
 465