xdiff / xutils.con commit Merge branch 'ds/midx-expire-repack' (8ce8a63)
   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, see
  17 *  <http://www.gnu.org/licenses/>.
  18 *
  19 *  Davide Libenzi <davidel@xmailserver.org>
  20 *
  21 */
  22
  23#include "xinclude.h"
  24
  25
  26long xdl_bogosqrt(long n) {
  27        long i;
  28
  29        /*
  30         * Classical integer square root approximation using shifts.
  31         */
  32        for (i = 1; n > 0; n >>= 2)
  33                i <<= 1;
  34
  35        return i;
  36}
  37
  38
  39int xdl_emit_diffrec(char const *rec, long size, char const *pre, long psize,
  40                     xdemitcb_t *ecb) {
  41        int i = 2;
  42        mmbuffer_t mb[3];
  43
  44        mb[0].ptr = (char *) pre;
  45        mb[0].size = psize;
  46        mb[1].ptr = (char *) rec;
  47        mb[1].size = size;
  48        if (size > 0 && rec[size - 1] != '\n') {
  49                mb[2].ptr = (char *) "\n\\ No newline at end of file\n";
  50                mb[2].size = strlen(mb[2].ptr);
  51                i++;
  52        }
  53        if (ecb->out_line(ecb->priv, mb, i) < 0) {
  54
  55                return -1;
  56        }
  57
  58        return 0;
  59}
  60
  61void *xdl_mmfile_first(mmfile_t *mmf, long *size)
  62{
  63        *size = mmf->size;
  64        return mmf->ptr;
  65}
  66
  67
  68long xdl_mmfile_size(mmfile_t *mmf)
  69{
  70        return mmf->size;
  71}
  72
  73
  74int xdl_cha_init(chastore_t *cha, long isize, long icount) {
  75
  76        cha->head = cha->tail = NULL;
  77        cha->isize = isize;
  78        cha->nsize = icount * isize;
  79        cha->ancur = cha->sncur = NULL;
  80        cha->scurr = 0;
  81
  82        return 0;
  83}
  84
  85
  86void xdl_cha_free(chastore_t *cha) {
  87        chanode_t *cur, *tmp;
  88
  89        for (cur = cha->head; (tmp = cur) != NULL;) {
  90                cur = cur->next;
  91                xdl_free(tmp);
  92        }
  93}
  94
  95
  96void *xdl_cha_alloc(chastore_t *cha) {
  97        chanode_t *ancur;
  98        void *data;
  99
 100        if (!(ancur = cha->ancur) || ancur->icurr == cha->nsize) {
 101                if (!(ancur = (chanode_t *) xdl_malloc(sizeof(chanode_t) + cha->nsize))) {
 102
 103                        return NULL;
 104                }
 105                ancur->icurr = 0;
 106                ancur->next = NULL;
 107                if (cha->tail)
 108                        cha->tail->next = ancur;
 109                if (!cha->head)
 110                        cha->head = ancur;
 111                cha->tail = ancur;
 112                cha->ancur = ancur;
 113        }
 114
 115        data = (char *) ancur + sizeof(chanode_t) + ancur->icurr;
 116        ancur->icurr += cha->isize;
 117
 118        return data;
 119}
 120
 121long xdl_guess_lines(mmfile_t *mf, long sample) {
 122        long nl = 0, size, tsize = 0;
 123        char const *data, *cur, *top;
 124
 125        if ((cur = data = xdl_mmfile_first(mf, &size)) != NULL) {
 126                for (top = data + size; nl < sample && cur < top; ) {
 127                        nl++;
 128                        if (!(cur = memchr(cur, '\n', top - cur)))
 129                                cur = top;
 130                        else
 131                                cur++;
 132                }
 133                tsize += (long) (cur - data);
 134        }
 135
 136        if (nl && tsize)
 137                nl = xdl_mmfile_size(mf) / (tsize / nl);
 138
 139        return nl + 1;
 140}
 141
 142int xdl_blankline(const char *line, long size, long flags)
 143{
 144        long i;
 145
 146        if (!(flags & XDF_WHITESPACE_FLAGS))
 147                return (size <= 1);
 148
 149        for (i = 0; i < size && XDL_ISSPACE(line[i]); i++)
 150                ;
 151
 152        return (i == size);
 153}
 154
 155/*
 156 * Have we eaten everything on the line, except for an optional
 157 * CR at the very end?
 158 */
 159static int ends_with_optional_cr(const char *l, long s, long i)
 160{
 161        int complete = s && l[s-1] == '\n';
 162
 163        if (complete)
 164                s--;
 165        if (s == i)
 166                return 1;
 167        /* do not ignore CR at the end of an incomplete line */
 168        if (complete && s == i + 1 && l[i] == '\r')
 169                return 1;
 170        return 0;
 171}
 172
 173int xdl_recmatch(const char *l1, long s1, const char *l2, long s2, long flags)
 174{
 175        int i1, i2;
 176
 177        if (s1 == s2 && !memcmp(l1, l2, s1))
 178                return 1;
 179        if (!(flags & XDF_WHITESPACE_FLAGS))
 180                return 0;
 181
 182        i1 = 0;
 183        i2 = 0;
 184
 185        /*
 186         * -w matches everything that matches with -b, and -b in turn
 187         * matches everything that matches with --ignore-space-at-eol,
 188         * which in turn matches everything that matches with --ignore-cr-at-eol.
 189         *
 190         * Each flavor of ignoring needs different logic to skip whitespaces
 191         * while we have both sides to compare.
 192         */
 193        if (flags & XDF_IGNORE_WHITESPACE) {
 194                goto skip_ws;
 195                while (i1 < s1 && i2 < s2) {
 196                        if (l1[i1++] != l2[i2++])
 197                                return 0;
 198                skip_ws:
 199                        while (i1 < s1 && XDL_ISSPACE(l1[i1]))
 200                                i1++;
 201                        while (i2 < s2 && XDL_ISSPACE(l2[i2]))
 202                                i2++;
 203                }
 204        } else if (flags & XDF_IGNORE_WHITESPACE_CHANGE) {
 205                while (i1 < s1 && i2 < s2) {
 206                        if (XDL_ISSPACE(l1[i1]) && XDL_ISSPACE(l2[i2])) {
 207                                /* Skip matching spaces and try again */
 208                                while (i1 < s1 && XDL_ISSPACE(l1[i1]))
 209                                        i1++;
 210                                while (i2 < s2 && XDL_ISSPACE(l2[i2]))
 211                                        i2++;
 212                                continue;
 213                        }
 214                        if (l1[i1++] != l2[i2++])
 215                                return 0;
 216                }
 217        } else if (flags & XDF_IGNORE_WHITESPACE_AT_EOL) {
 218                while (i1 < s1 && i2 < s2 && l1[i1] == l2[i2]) {
 219                        i1++;
 220                        i2++;
 221                }
 222        } else if (flags & XDF_IGNORE_CR_AT_EOL) {
 223                /* Find the first difference and see how the line ends */
 224                while (i1 < s1 && i2 < s2 && l1[i1] == l2[i2]) {
 225                        i1++;
 226                        i2++;
 227                }
 228                return (ends_with_optional_cr(l1, s1, i1) &&
 229                        ends_with_optional_cr(l2, s2, i2));
 230        }
 231
 232        /*
 233         * After running out of one side, the remaining side must have
 234         * nothing but whitespace for the lines to match.  Note that
 235         * ignore-whitespace-at-eol case may break out of the loop
 236         * while there still are characters remaining on both lines.
 237         */
 238        if (i1 < s1) {
 239                while (i1 < s1 && XDL_ISSPACE(l1[i1]))
 240                        i1++;
 241                if (s1 != i1)
 242                        return 0;
 243        }
 244        if (i2 < s2) {
 245                while (i2 < s2 && XDL_ISSPACE(l2[i2]))
 246                        i2++;
 247                return (s2 == i2);
 248        }
 249        return 1;
 250}
 251
 252static unsigned long xdl_hash_record_with_whitespace(char const **data,
 253                char const *top, long flags) {
 254        unsigned long ha = 5381;
 255        char const *ptr = *data;
 256        int cr_at_eol_only = (flags & XDF_WHITESPACE_FLAGS) == XDF_IGNORE_CR_AT_EOL;
 257
 258        for (; ptr < top && *ptr != '\n'; ptr++) {
 259                if (cr_at_eol_only) {
 260                        /* do not ignore CR at the end of an incomplete line */
 261                        if (*ptr == '\r' &&
 262                            (ptr + 1 < top && ptr[1] == '\n'))
 263                                continue;
 264                }
 265                else if (XDL_ISSPACE(*ptr)) {
 266                        const char *ptr2 = ptr;
 267                        int at_eol;
 268                        while (ptr + 1 < top && XDL_ISSPACE(ptr[1])
 269                                        && ptr[1] != '\n')
 270                                ptr++;
 271                        at_eol = (top <= ptr + 1 || ptr[1] == '\n');
 272                        if (flags & XDF_IGNORE_WHITESPACE)
 273                                ; /* already handled */
 274                        else if (flags & XDF_IGNORE_WHITESPACE_CHANGE
 275                                 && !at_eol) {
 276                                ha += (ha << 5);
 277                                ha ^= (unsigned long) ' ';
 278                        }
 279                        else if (flags & XDF_IGNORE_WHITESPACE_AT_EOL
 280                                 && !at_eol) {
 281                                while (ptr2 != ptr + 1) {
 282                                        ha += (ha << 5);
 283                                        ha ^= (unsigned long) *ptr2;
 284                                        ptr2++;
 285                                }
 286                        }
 287                        continue;
 288                }
 289                ha += (ha << 5);
 290                ha ^= (unsigned long) *ptr;
 291        }
 292        *data = ptr < top ? ptr + 1: ptr;
 293
 294        return ha;
 295}
 296
 297unsigned long xdl_hash_record(char const **data, char const *top, long flags) {
 298        unsigned long ha = 5381;
 299        char const *ptr = *data;
 300
 301        if (flags & XDF_WHITESPACE_FLAGS)
 302                return xdl_hash_record_with_whitespace(data, top, flags);
 303
 304        for (; ptr < top && *ptr != '\n'; ptr++) {
 305                ha += (ha << 5);
 306                ha ^= (unsigned long) *ptr;
 307        }
 308        *data = ptr < top ? ptr + 1: ptr;
 309
 310        return ha;
 311}
 312
 313unsigned int xdl_hashbits(unsigned int size) {
 314        unsigned int val = 1, bits = 0;
 315
 316        for (; val < size && bits < CHAR_BIT * sizeof(unsigned int); val <<= 1, bits++);
 317        return bits ? bits: 1;
 318}
 319
 320
 321int xdl_num_out(char *out, long val) {
 322        char *ptr, *str = out;
 323        char buf[32];
 324
 325        ptr = buf + sizeof(buf) - 1;
 326        *ptr = '\0';
 327        if (val < 0) {
 328                *--ptr = '-';
 329                val = -val;
 330        }
 331        for (; val && ptr > buf; val /= 10)
 332                *--ptr = "0123456789"[val % 10];
 333        if (*ptr)
 334                for (; *ptr; ptr++, str++)
 335                        *str = *ptr;
 336        else
 337                *str++ = '0';
 338        *str = '\0';
 339
 340        return str - out;
 341}
 342
 343static int xdl_format_hunk_hdr(long s1, long c1, long s2, long c2,
 344                               const char *func, long funclen,
 345                               xdemitcb_t *ecb) {
 346        int nb = 0;
 347        mmbuffer_t mb;
 348        char buf[128];
 349
 350        memcpy(buf, "@@ -", 4);
 351        nb += 4;
 352
 353        nb += xdl_num_out(buf + nb, c1 ? s1: s1 - 1);
 354
 355        if (c1 != 1) {
 356                memcpy(buf + nb, ",", 1);
 357                nb += 1;
 358
 359                nb += xdl_num_out(buf + nb, c1);
 360        }
 361
 362        memcpy(buf + nb, " +", 2);
 363        nb += 2;
 364
 365        nb += xdl_num_out(buf + nb, c2 ? s2: s2 - 1);
 366
 367        if (c2 != 1) {
 368                memcpy(buf + nb, ",", 1);
 369                nb += 1;
 370
 371                nb += xdl_num_out(buf + nb, c2);
 372        }
 373
 374        memcpy(buf + nb, " @@", 3);
 375        nb += 3;
 376        if (func && funclen) {
 377                buf[nb++] = ' ';
 378                if (funclen > sizeof(buf) - nb - 1)
 379                        funclen = sizeof(buf) - nb - 1;
 380                memcpy(buf + nb, func, funclen);
 381                nb += funclen;
 382        }
 383        buf[nb++] = '\n';
 384
 385        mb.ptr = buf;
 386        mb.size = nb;
 387        if (ecb->out_line(ecb->priv, &mb, 1) < 0)
 388                return -1;
 389        return 0;
 390}
 391
 392int xdl_emit_hunk_hdr(long s1, long c1, long s2, long c2,
 393                      const char *func, long funclen,
 394                      xdemitcb_t *ecb) {
 395        if (!ecb->out_hunk)
 396                return xdl_format_hunk_hdr(s1, c1, s2, c2, func, funclen, ecb);
 397        if (ecb->out_hunk(ecb->priv,
 398                          c1 ? s1 : s1 - 1, c1,
 399                          c2 ? s2 : s2 - 1, c2,
 400                          func, funclen) < 0)
 401                return -1;
 402        return 0;
 403}
 404
 405int xdl_fall_back_diff(xdfenv_t *diff_env, xpparam_t const *xpp,
 406                int line1, int count1, int line2, int count2)
 407{
 408        /*
 409         * This probably does not work outside Git, since
 410         * we have a very simple mmfile structure.
 411         *
 412         * Note: ideally, we would reuse the prepared environment, but
 413         * the libxdiff interface does not (yet) allow for diffing only
 414         * ranges of lines instead of the whole files.
 415         */
 416        mmfile_t subfile1, subfile2;
 417        xdfenv_t env;
 418
 419        subfile1.ptr = (char *)diff_env->xdf1.recs[line1 - 1]->ptr;
 420        subfile1.size = diff_env->xdf1.recs[line1 + count1 - 2]->ptr +
 421                diff_env->xdf1.recs[line1 + count1 - 2]->size - subfile1.ptr;
 422        subfile2.ptr = (char *)diff_env->xdf2.recs[line2 - 1]->ptr;
 423        subfile2.size = diff_env->xdf2.recs[line2 + count2 - 2]->ptr +
 424                diff_env->xdf2.recs[line2 + count2 - 2]->size - subfile2.ptr;
 425        if (xdl_do_diff(&subfile1, &subfile2, xpp, &env) < 0)
 426                return -1;
 427
 428        memcpy(diff_env->xdf1.rchg + line1 - 1, env.xdf1.rchg, count1);
 429        memcpy(diff_env->xdf2.rchg + line2 - 1, env.xdf2.rchg, count2);
 430
 431        xdl_free_env(&env);
 432
 433        return 0;
 434}