Andrew's git - gitweb.git/blob

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