xdiff / xutils.con commit diff-delta.c: Rationalize culling of hash buckets (02e665c)
   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_GUESS_NLINES 256
  28
  29
  30
  31
  32long xdl_bogosqrt(long n) {
  33        long i;
  34
  35        /*
  36         * Classical integer square root approximation using shifts.
  37         */
  38        for (i = 1; n > 0; n >>= 2)
  39                i <<= 1;
  40
  41        return i;
  42}
  43
  44
  45int xdl_emit_diffrec(char const *rec, long size, char const *pre, long psize,
  46                     xdemitcb_t *ecb) {
  47        int i = 2;
  48        mmbuffer_t mb[3];
  49
  50        mb[0].ptr = (char *) pre;
  51        mb[0].size = psize;
  52        mb[1].ptr = (char *) rec;
  53        mb[1].size = size;
  54        if (size > 0 && rec[size - 1] != '\n') {
  55                mb[2].ptr = (char *) "\n\\ No newline at end of file\n";
  56                mb[2].size = strlen(mb[2].ptr);
  57                i++;
  58        }
  59        if (ecb->outf(ecb->priv, mb, i) < 0) {
  60
  61                return -1;
  62        }
  63
  64        return 0;
  65}
  66
  67void *xdl_mmfile_first(mmfile_t *mmf, long *size)
  68{
  69        *size = mmf->size;
  70        return mmf->ptr;
  71}
  72
  73
  74void *xdl_mmfile_next(mmfile_t *mmf, long *size)
  75{
  76        return NULL;
  77}
  78
  79
  80long xdl_mmfile_size(mmfile_t *mmf)
  81{
  82        return mmf->size;
  83}
  84
  85
  86int xdl_cha_init(chastore_t *cha, long isize, long icount) {
  87
  88        cha->head = cha->tail = NULL;
  89        cha->isize = isize;
  90        cha->nsize = icount * isize;
  91        cha->ancur = cha->sncur = NULL;
  92        cha->scurr = 0;
  93
  94        return 0;
  95}
  96
  97
  98void xdl_cha_free(chastore_t *cha) {
  99        chanode_t *cur, *tmp;
 100
 101        for (cur = cha->head; (tmp = cur) != NULL;) {
 102                cur = cur->next;
 103                xdl_free(tmp);
 104        }
 105}
 106
 107
 108void *xdl_cha_alloc(chastore_t *cha) {
 109        chanode_t *ancur;
 110        void *data;
 111
 112        if (!(ancur = cha->ancur) || ancur->icurr == cha->nsize) {
 113                if (!(ancur = (chanode_t *) xdl_malloc(sizeof(chanode_t) + cha->nsize))) {
 114
 115                        return NULL;
 116                }
 117                ancur->icurr = 0;
 118                ancur->next = NULL;
 119                if (cha->tail)
 120                        cha->tail->next = ancur;
 121                if (!cha->head)
 122                        cha->head = ancur;
 123                cha->tail = ancur;
 124                cha->ancur = ancur;
 125        }
 126
 127        data = (char *) ancur + sizeof(chanode_t) + ancur->icurr;
 128        ancur->icurr += cha->isize;
 129
 130        return data;
 131}
 132
 133
 134void *xdl_cha_first(chastore_t *cha) {
 135        chanode_t *sncur;
 136
 137        if (!(cha->sncur = sncur = cha->head))
 138                return NULL;
 139
 140        cha->scurr = 0;
 141
 142        return (char *) sncur + sizeof(chanode_t) + cha->scurr;
 143}
 144
 145
 146void *xdl_cha_next(chastore_t *cha) {
 147        chanode_t *sncur;
 148
 149        if (!(sncur = cha->sncur))
 150                return NULL;
 151        cha->scurr += cha->isize;
 152        if (cha->scurr == sncur->icurr) {
 153                if (!(sncur = cha->sncur = sncur->next))
 154                        return NULL;
 155                cha->scurr = 0;
 156        }
 157
 158        return (char *) sncur + sizeof(chanode_t) + cha->scurr;
 159}
 160
 161
 162long xdl_guess_lines(mmfile_t *mf) {
 163        long nl = 0, size, tsize = 0;
 164        char const *data, *cur, *top;
 165
 166        if ((cur = data = xdl_mmfile_first(mf, &size)) != NULL) {
 167                for (top = data + size; nl < XDL_GUESS_NLINES;) {
 168                        if (cur >= top) {
 169                                tsize += (long) (cur - data);
 170                                if (!(cur = data = xdl_mmfile_next(mf, &size)))
 171                                        break;
 172                                top = data + size;
 173                        }
 174                        nl++;
 175                        if (!(cur = memchr(cur, '\n', top - cur)))
 176                                cur = top;
 177                        else
 178                                cur++;
 179                }
 180                tsize += (long) (cur - data);
 181        }
 182
 183        if (nl && tsize)
 184                nl = xdl_mmfile_size(mf) / (tsize / nl);
 185
 186        return nl + 1;
 187}
 188
 189int xdl_recmatch(const char *l1, long s1, const char *l2, long s2, long flags)
 190{
 191        int i1, i2;
 192
 193        if (flags & XDF_IGNORE_WHITESPACE) {
 194                for (i1 = i2 = 0; i1 < s1 && i2 < s2; ) {
 195                        if (isspace(l1[i1]))
 196                                while (isspace(l1[i1]) && i1 < s1)
 197                                        i1++;
 198                        if (isspace(l2[i2]))
 199                                while (isspace(l2[i2]) && i2 < s2)
 200                                        i2++;
 201                        if (i1 < s1 && i2 < s2 && l1[i1++] != l2[i2++])
 202                                return 0;
 203                }
 204                return (i1 >= s1 && i2 >= s2);
 205        } else if (flags & XDF_IGNORE_WHITESPACE_CHANGE) {
 206                for (i1 = i2 = 0; i1 < s1 && i2 < s2; ) {
 207                        if (isspace(l1[i1])) {
 208                                if (!isspace(l2[i2]))
 209                                        return 0;
 210                                while (isspace(l1[i1]) && i1 < s1)
 211                                        i1++;
 212                                while (isspace(l2[i2]) && i2 < s2)
 213                                        i2++;
 214                        } else if (l1[i1++] != l2[i2++])
 215                                return 0;
 216                }
 217                return (i1 >= s1 && i2 >= s2);
 218        } else if (flags & XDF_IGNORE_WHITESPACE_AT_EOL) {
 219                for (i1 = i2 = 0; i1 < s1 && i2 < s2; ) {
 220                        if (l1[i1] != l2[i2]) {
 221                                while (i1 < s1 && isspace(l1[i1]))
 222                                        i1++;
 223                                while (i2 < s2 && isspace(l2[i2]))
 224                                        i2++;
 225                                if (i1 < s1 || i2 < s2)
 226                                        return 0;
 227                                return 1;
 228                        }
 229                        i1++;
 230                        i2++;
 231                }
 232                return i1 >= s1 && i2 >= s2;
 233        } else
 234                return s1 == s2 && !memcmp(l1, l2, s1);
 235
 236        return 0;
 237}
 238
 239static unsigned long xdl_hash_record_with_whitespace(char const **data,
 240                char const *top, long flags) {
 241        unsigned long ha = 5381;
 242        char const *ptr = *data;
 243
 244        for (; ptr < top && *ptr != '\n'; ptr++) {
 245                if (isspace(*ptr)) {
 246                        const char *ptr2 = ptr;
 247                        while (ptr + 1 < top && isspace(ptr[1])
 248                                        && ptr[1] != '\n')
 249                                ptr++;
 250                        if (flags & XDF_IGNORE_WHITESPACE_CHANGE
 251                                        && ptr[1] != '\n') {
 252                                ha += (ha << 5);
 253                                ha ^= (unsigned long) ' ';
 254                        }
 255                        if (flags & XDF_IGNORE_WHITESPACE_AT_EOL
 256                                        && ptr[1] != '\n') {
 257                                while (ptr2 != ptr + 1) {
 258                                        ha += (ha << 5);
 259                                        ha ^= (unsigned long) *ptr2;
 260                                        ptr2++;
 261                                }
 262                        }
 263                        continue;
 264                }
 265                ha += (ha << 5);
 266                ha ^= (unsigned long) *ptr;
 267        }
 268        *data = ptr < top ? ptr + 1: ptr;
 269
 270        return ha;
 271}
 272
 273
 274unsigned long xdl_hash_record(char const **data, char const *top, long flags) {
 275        unsigned long ha = 5381;
 276        char const *ptr = *data;
 277
 278        if (flags & XDF_WHITESPACE_FLAGS)
 279                return xdl_hash_record_with_whitespace(data, top, flags);
 280
 281        for (; ptr < top && *ptr != '\n'; ptr++) {
 282                ha += (ha << 5);
 283                ha ^= (unsigned long) *ptr;
 284        }
 285        *data = ptr < top ? ptr + 1: ptr;
 286
 287        return ha;
 288}
 289
 290
 291unsigned int xdl_hashbits(unsigned int size) {
 292        unsigned int val = 1, bits = 0;
 293
 294        for (; val < size && bits < CHAR_BIT * sizeof(unsigned int); val <<= 1, bits++);
 295        return bits ? bits: 1;
 296}
 297
 298
 299int xdl_num_out(char *out, long val) {
 300        char *ptr, *str = out;
 301        char buf[32];
 302
 303        ptr = buf + sizeof(buf) - 1;
 304        *ptr = '\0';
 305        if (val < 0) {
 306                *--ptr = '-';
 307                val = -val;
 308        }
 309        for (; val && ptr > buf; val /= 10)
 310                *--ptr = "0123456789"[val % 10];
 311        if (*ptr)
 312                for (; *ptr; ptr++, str++)
 313                        *str = *ptr;
 314        else
 315                *str++ = '0';
 316        *str = '\0';
 317
 318        return str - out;
 319}
 320
 321
 322long xdl_atol(char const *str, char const **next) {
 323        long val, base;
 324        char const *top;
 325
 326        for (top = str; XDL_ISDIGIT(*top); top++);
 327        if (next)
 328                *next = top;
 329        for (val = 0, base = 1, top--; top >= str; top--, base *= 10)
 330                val += base * (long)(*top - '0');
 331        return val;
 332}
 333
 334
 335int xdl_emit_hunk_hdr(long s1, long c1, long s2, long c2,
 336                      const char *func, long funclen, xdemitcb_t *ecb) {
 337        int nb = 0;
 338        mmbuffer_t mb;
 339        char buf[128];
 340
 341        memcpy(buf, "@@ -", 4);
 342        nb += 4;
 343
 344        nb += xdl_num_out(buf + nb, c1 ? s1: s1 - 1);
 345
 346        if (c1 != 1) {
 347                memcpy(buf + nb, ",", 1);
 348                nb += 1;
 349
 350                nb += xdl_num_out(buf + nb, c1);
 351        }
 352
 353        memcpy(buf + nb, " +", 2);
 354        nb += 2;
 355
 356        nb += xdl_num_out(buf + nb, c2 ? s2: s2 - 1);
 357
 358        if (c2 != 1) {
 359                memcpy(buf + nb, ",", 1);
 360                nb += 1;
 361
 362                nb += xdl_num_out(buf + nb, c2);
 363        }
 364
 365        memcpy(buf + nb, " @@", 3);
 366        nb += 3;
 367        if (func && funclen) {
 368                buf[nb++] = ' ';
 369                if (funclen > sizeof(buf) - nb - 1)
 370                        funclen = sizeof(buf) - nb - 1;
 371                memcpy(buf + nb, func, funclen);
 372                nb += funclen;
 373        }
 374        buf[nb++] = '\n';
 375
 376        mb.ptr = buf;
 377        mb.size = nb;
 378        if (ecb->outf(ecb->priv, &mb, 1) < 0)
 379                return -1;
 380
 381        return 0;
 382}