exclude.c

   1 /* -*- c-file-style: "linux" -*-
   2  *
   3  * Copyright (C) 1996-2001 by Andrew Tridgell <tridge@samba.org>
   4  * Copyright (C) 1996 by Paul Mackerras
   5  * Copyright (C) 2002 by Martin Pool
   6  *
   7  * This program is free software; you can redistribute it and/or modify
   8  * it under the terms of the GNU General Public License as published by
   9  * the Free Software Foundation; either version 2 of the License, or
  10  * (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  * GNU General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU General Public License
  18  * along with this program; if not, write to the Free Software
  19  * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  20  */
  21
  22 /* a lot of this stuff was originally derived from GNU tar, although
  23    it has now changed so much that it is hard to tell :) */
  24
  25 /* include/exclude cluestick added by Martin Pool <mbp@samba.org> */
  26
  27 #include "rsync.h"
  28
  29 extern int verbose;
  30 extern int am_server;
  31 extern int am_sender;
  32 extern int eol_nulls;
  33 extern int list_only;
  34 extern int recurse;
  35 extern int io_error;
  36 extern int local_server;
  37 extern int delete_mode;
  38 extern int delete_excluded;
  39 extern int cvs_exclude;
  40 extern int sanitize_paths;
  41 extern int protocol_version;
  42 extern int module_id;
  43
  44 extern char curr_dir[];
  45 extern unsigned int curr_dir_len;
  46 extern unsigned int module_dirlen;
  47
  48 struct filter_list_struct filter_list = { 0, 0, "" };
  49 struct filter_list_struct cvs_filter_list = { 0, 0, " [cvsignore]" };
  50 struct filter_list_struct server_filter_list = { 0, 0, " [server]" };
  51
  52 /* Need room enough for ":MODS " prefix plus some room to grow. */
  53 #define MAX_RULE_PREFIX (16)
  54
  55 #define MODIFIERS_MERGE_FILE "-+Cenw"
  56 #define MODIFIERS_INCL_EXCL "/!Crs"
  57 #define MODIFIERS_HIDE_PROTECT "/!"
  58
  59 /* The dirbuf is set by push_local_filters() to the current subdirectory
  60  * relative to curr_dir that is being processed.  The path always has a
  61  * trailing slash appended, and the variable dirbuf_len contains the length
  62  * of this path prefix.  The path is always absolute. */
  63 static char dirbuf[MAXPATHLEN+1];
  64 static unsigned int dirbuf_len = 0;
  65 static int dirbuf_depth;
  66
  67 /* This is True when we're scanning parent dirs for per-dir merge-files. */
  68 static BOOL parent_dirscan = False;
  69
  70 /* This array contains a list of all the currently active per-dir merge
  71  * files.  This makes it easier to save the appropriate values when we
  72  * "push" down into each subdirectory. */
  73 static struct filter_struct **mergelist_parents;
  74 static int mergelist_cnt = 0;
  75 static int mergelist_size = 0;
  76
  77 /* Each filter_list_struct describes a singly-linked list by keeping track
  78  * of both the head and tail pointers.  The list is slightly unusual in that
  79  * a parent-dir's content can be appended to the end of the local list in a
  80  * special way:  the last item in the local list has its "next" pointer set
  81  * to point to the inherited list, but the local list's tail pointer points
  82  * at the end of the local list.  Thus, if the local list is empty, the head
  83  * will be pointing at the inherited content but the tail will be NULL.  To
  84  * help you visualize this, here are the possible list arrangements:
  85  *
  86  * Completely Empty                     Local Content Only
  87  * ==================================   ====================================
  88  * head -> NULL                         head -> Local1 -> Local2 -> NULL
  89  * tail -> NULL                         tail -------------^
  90  *
  91  * Inherited Content Only               Both Local and Inherited Content
  92  * ==================================   ====================================
  93  * head -> Parent1 -> Parent2 -> NULL   head -> L1 -> L2 -> P1 -> P2 -> NULL
  94  * tail -> NULL                         tail ---------^
  95  *
  96  * This means that anyone wanting to traverse the whole list to use it just
  97  * needs to start at the head and use the "next" pointers until it goes
  98  * NULL.  To add new local content, we insert the item after the tail item
  99  * and update the tail (obviously, if "tail" was NULL, we insert it at the
 100  * head).  To clear the local list, WE MUST NOT FREE THE INHERITED CONTENT
 101  * because it is shared between the current list and our parent list(s).
 102  * The easiest way to handle this is to simply truncate the list after the
 103  * tail item and then free the local list from the head.  When inheriting
 104  * the list for a new local dir, we just save off the filter_list_struct
 105  * values (so we can pop back to them later) and set the tail to NULL.
 106  */
 107
 108 static void free_filter(struct filter_struct *ex)
 109 {
 110         if (ex->match_flags & MATCHFLG_PERDIR_MERGE) {
 111                 free(ex->u.mergelist->debug_type);
 112                 free(ex->u.mergelist);
 113                 mergelist_cnt--;
 114         }
 115         free(ex->pattern);
 116         free(ex);
 117 }
 118
 119 /* Build a filter structure given a filter pattern.  The value in "pat"
 120  * is not null-terminated. */
 121 static void add_rule(struct filter_list_struct *listp, const char *pat,
 122                      unsigned int pat_len, uint32 mflags, int xflags)
 123 {
 124         struct filter_struct *ret;
 125         const char *cp;
 126         unsigned int ex_len;
 127
 128         if (verbose > 2) {
 129                 rprintf(FINFO, "[%s] add_rule(%s%.*s%s)%s\n",
 130                         who_am_i(), get_rule_prefix(mflags, pat, 0, NULL),
 131                         (int)pat_len, pat,
 132                         (mflags & MATCHFLG_DIRECTORY) ? "/" : "",
 133                         listp->debug_type);
 134         }
 135
 136         /* This flag also indicates that we're reading a list that
 137          * needs to be filtered now, not post-filtered later. */
 138         if (xflags & XFLG_ANCHORED2ABS) {
 139                 uint32 mf = mflags & (MATCHFLG_RECEIVER_SIDE|MATCHFLG_SENDER_SIDE);
 140                 if (am_sender) {
 141                         if (mf == MATCHFLG_RECEIVER_SIDE)
 142                                 return;
 143                 } else {
 144                         if (mf == MATCHFLG_SENDER_SIDE)
 145                                 return;
 146                 }
 147         }
 148
 149         if (!(ret = new(struct filter_struct)))
 150                 out_of_memory("add_rule");
 151         memset(ret, 0, sizeof ret[0]);
 152
 153         if (xflags & XFLG_ANCHORED2ABS && *pat == '/'
 154             && !(mflags & (MATCHFLG_ABS_PATH | MATCHFLG_MERGE_FILE))) {
 155                 mflags |= MATCHFLG_ABS_PATH;
 156                 ex_len = dirbuf_len - module_dirlen - 1;
 157         } else
 158                 ex_len = 0;
 159         if (!(ret->pattern = new_array(char, ex_len + pat_len + 1)))
 160                 out_of_memory("add_rule");
 161         if (ex_len)
 162                 memcpy(ret->pattern, dirbuf + module_dirlen, ex_len);
 163         strlcpy(ret->pattern + ex_len, pat, pat_len + 1);
 164         pat_len += ex_len;
 165
 166         if (strpbrk(ret->pattern, "*[?")) {
 167                 mflags |= MATCHFLG_WILD;
 168                 if ((cp = strstr(ret->pattern, "**")) != NULL) {
 169                         mflags |= MATCHFLG_WILD2;
 170                         /* If the pattern starts with **, note that. */
 171                         if (cp == ret->pattern)
 172                                 mflags |= MATCHFLG_WILD2_PREFIX;
 173                 }
 174         }
 175
 176         if (pat_len > 1 && ret->pattern[pat_len-1] == '/') {
 177                 ret->pattern[pat_len-1] = 0;
 178                 mflags |= MATCHFLG_DIRECTORY;
 179         }
 180
 181         if (mflags & MATCHFLG_PERDIR_MERGE) {
 182                 struct filter_list_struct *lp;
 183                 unsigned int len;
 184                 int i;
 185
 186                 if ((cp = strrchr(ret->pattern, '/')) != NULL)
 187                         cp++;
 188                 else
 189                         cp = ret->pattern;
 190
 191                 /* If the local merge file was already mentioned, don't
 192                  * add it again. */
 193                 for (i = 0; i < mergelist_cnt; i++) {
 194                         struct filter_struct *ex = mergelist_parents[i];
 195                         const char *s = strrchr(ex->pattern, '/');
 196                         if (s)
 197                                 s++;
 198                         else
 199                                 s = ex->pattern;
 200                         len = strlen(s);
 201                         if (len == pat_len - (cp - ret->pattern)
 202                             && memcmp(s, cp, len) == 0) {
 203                                 free_filter(ret);
 204                                 return;
 205                         }
 206                 }
 207
 208                 if (!(lp = new_array(struct filter_list_struct, 1)))
 209                         out_of_memory("add_rule");
 210                 lp->head = lp->tail = NULL;
 211                 if (asprintf(&lp->debug_type, " [per-dir %s]", cp) < 0)
 212                         out_of_memory("add_rule");
 213                 ret->u.mergelist = lp;
 214
 215                 if (mergelist_cnt == mergelist_size) {
 216                         mergelist_size += 5;
 217                         mergelist_parents = realloc_array(mergelist_parents,
 218                                                 struct filter_struct *,
 219                                                 mergelist_size);
 220                         if (!mergelist_parents)
 221                                 out_of_memory("add_rule");
 222                 }
 223                 mergelist_parents[mergelist_cnt++] = ret;
 224         } else {
 225                 for (cp = ret->pattern; (cp = strchr(cp, '/')) != NULL; cp++)
 226                         ret->u.slash_cnt++;
 227         }
 228
 229         ret->match_flags = mflags;
 230
 231         if (!listp->tail) {
 232                 ret->next = listp->head;
 233                 listp->head = listp->tail = ret;
 234         } else {
 235                 ret->next = listp->tail->next;
 236                 listp->tail->next = ret;
 237                 listp->tail = ret;
 238         }
 239 }
 240
 241 static void clear_filter_list(struct filter_list_struct *listp)
 242 {
 243         if (listp->tail) {
 244                 struct filter_struct *ent, *next;
 245                 /* Truncate any inherited items from the local list. */
 246                 listp->tail->next = NULL;
 247                 /* Now free everything that is left. */
 248                 for (ent = listp->head; ent; ent = next) {
 249                         next = ent->next;
 250                         free_filter(ent);
 251                 }
 252         }
 253
 254         listp->head = listp->tail = NULL;
 255 }
 256
 257 /* This returns an expanded (absolute) filename for the merge-file name if
 258  * the name has any slashes in it OR if the parent_dirscan var is True;
 259  * otherwise it returns the original merge_file name.  If the len_ptr value
 260  * is non-NULL the merge_file name is limited by the referenced length
 261  * value and will be updated with the length of the resulting name.  We
 262  * always return a name that is null terminated, even if the merge_file
 263  * name was not. */
 264 static char *parse_merge_name(const char *merge_file, unsigned int *len_ptr,
 265                               unsigned int prefix_skip)
 266 {
 267         static char buf[MAXPATHLEN];
 268         char *fn, tmpbuf[MAXPATHLEN];
 269         unsigned int fn_len;
 270
 271         if (!parent_dirscan && *merge_file != '/') {
 272                 /* Return the name unchanged it doesn't have any slashes. */
 273                 if (len_ptr) {
 274                         const char *p = merge_file + *len_ptr;
 275                         while (--p > merge_file && *p != '/') {}
 276                         if (p == merge_file) {
 277                                 strlcpy(buf, merge_file, *len_ptr + 1);
 278                                 return buf;
 279                         }
 280                 } else if (strchr(merge_file, '/') == NULL)
 281                         return (char *)merge_file;
 282         }
 283
 284         fn = *merge_file == '/' ? buf : tmpbuf;
 285         if (sanitize_paths) {
 286                 const char *r = prefix_skip ? "/" : NULL;
 287                 /* null-terminate the name if it isn't already */
 288                 if (len_ptr && merge_file[*len_ptr]) {
 289                         char *to = fn == buf ? tmpbuf : buf;
 290                         strlcpy(to, merge_file, *len_ptr + 1);
 291                         merge_file = to;
 292                 }
 293                 if (!sanitize_path(fn, merge_file, r, dirbuf_depth)) {
 294                         rprintf(FERROR, "merge-file name overflows: %s\n",
 295                                 safe_fname(merge_file));
 296                         return NULL;
 297                 }
 298         } else {
 299                 strlcpy(fn, merge_file, len_ptr ? *len_ptr + 1 : MAXPATHLEN);
 300                 clean_fname(fn, 1);
 301         }
 302
 303         fn_len = strlen(fn);
 304         if (fn == buf)
 305                 goto done;
 306
 307         if (dirbuf_len + fn_len >= MAXPATHLEN) {
 308                 rprintf(FERROR, "merge-file name overflows: %s\n",
 309                         safe_fname(fn));
 310                 return NULL;
 311         }
 312         memcpy(buf, dirbuf + prefix_skip, dirbuf_len - prefix_skip);
 313         memcpy(buf + dirbuf_len - prefix_skip, fn, fn_len + 1);
 314         fn_len = clean_fname(buf, 1);
 315
 316     done:
 317         if (len_ptr)
 318                 *len_ptr = fn_len;
 319         return buf;
 320 }
 321
 322 /* Sets the dirbuf and dirbuf_len values. */
 323 void set_filter_dir(const char *dir, unsigned int dirlen)
 324 {
 325         unsigned int len;
 326         if (*dir != '/') {
 327                 memcpy(dirbuf, curr_dir, curr_dir_len);
 328                 dirbuf[curr_dir_len] = '/';
 329                 len = curr_dir_len + 1;
 330                 if (len + dirlen >= MAXPATHLEN)
 331                         dirlen = 0;
 332         } else
 333                 len = 0;
 334         memcpy(dirbuf + len, dir, dirlen);
 335         dirbuf[dirlen + len] = '\0';
 336         dirbuf_len = clean_fname(dirbuf, 1);
 337         if (dirbuf_len > 1 && dirbuf[dirbuf_len-1] == '.'
 338             && dirbuf[dirbuf_len-2] == '/')
 339                 dirbuf_len -= 2;
 340         if (dirbuf_len != 1)
 341                 dirbuf[dirbuf_len++] = '/';
 342         dirbuf[dirbuf_len] = '\0';
 343         if (sanitize_paths)
 344                 dirbuf_depth = count_dir_elements(dirbuf + module_dirlen);
 345 }
 346
 347 /* This routine takes a per-dir merge-file entry and finishes its setup.
 348  * If the name has a path portion then we check to see if it refers to a
 349  * parent directory of the first transfer dir.  If it does, we scan all the
 350  * dirs from that point through the parent dir of the transfer dir looking
 351  * for the per-dir merge-file in each one. */
 352 static BOOL setup_merge_file(struct filter_struct *ex,
 353                              struct filter_list_struct *lp)
 354 {
 355         char buf[MAXPATHLEN];
 356         char *x, *y, *pat = ex->pattern;
 357         unsigned int len;
 358
 359         if (!(x = parse_merge_name(pat, NULL, 0)) || *x != '/')
 360                 return 0;
 361
 362         y = strrchr(x, '/');
 363         *y = '\0';
 364         ex->pattern = strdup(y+1);
 365         if (!*x)
 366                 x = "/";
 367         if (*x == '/')
 368                 strlcpy(buf, x, MAXPATHLEN);
 369         else
 370                 pathjoin(buf, MAXPATHLEN, dirbuf, x);
 371
 372         len = clean_fname(buf, 1);
 373         if (len != 1 && len < MAXPATHLEN-1) {
 374                 buf[len++] = '/';
 375                 buf[len] = '\0';
 376         }
 377         /* This ensures that the specified dir is a parent of the transfer. */
 378         for (x = buf, y = dirbuf; *x && *x == *y; x++, y++) {}
 379         if (*x)
 380                 y += strlen(y); /* nope -- skip the scan */
 381
 382         parent_dirscan = True;
 383         while (*y) {
 384                 char save[MAXPATHLEN];
 385                 strlcpy(save, y, MAXPATHLEN);
 386                 *y = '\0';
 387                 dirbuf_len = y - dirbuf;
 388                 strlcpy(x, ex->pattern, MAXPATHLEN - (x - buf));
 389                 parse_filter_file(lp, buf, ex->match_flags, XFLG_ANCHORED2ABS);
 390                 if (ex->match_flags & MATCHFLG_NO_INHERIT)
 391                         lp->head = NULL;
 392                 lp->tail = NULL;
 393                 strlcpy(y, save, MAXPATHLEN);
 394                 while ((*x++ = *y++) != '/') {}
 395         }
 396         parent_dirscan = False;
 397         free(pat);
 398         return 1;
 399 }
 400
 401 /* Each time rsync changes to a new directory it call this function to
 402  * handle all the per-dir merge-files.  The "dir" value is the current path
 403  * relative to curr_dir (which might not be null-terminated).  We copy it
 404  * into dirbuf so that we can easily append a file name on the end. */
 405 void *push_local_filters(const char *dir, unsigned int dirlen)
 406 {
 407         struct filter_list_struct *ap, *push;
 408         int i;
 409
 410         set_filter_dir(dir, dirlen);
 411
 412         if (!mergelist_cnt)
 413                 return NULL;
 414
 415         push = new_array(struct filter_list_struct, mergelist_cnt);
 416         if (!push)
 417                 out_of_memory("push_local_filters");
 418
 419         for (i = 0, ap = push; i < mergelist_cnt; i++) {
 420                 memcpy(ap++, mergelist_parents[i]->u.mergelist,
 421                        sizeof (struct filter_list_struct));
 422         }
 423
 424         /* Note: parse_filter_file() might increase mergelist_cnt, so keep
 425          * this loop separate from the above loop. */
 426         for (i = 0; i < mergelist_cnt; i++) {
 427                 struct filter_struct *ex = mergelist_parents[i];
 428                 struct filter_list_struct *lp = ex->u.mergelist;
 429
 430                 if (verbose > 2) {
 431                         rprintf(FINFO, "[%s] pushing filter list%s\n",
 432                                 who_am_i(), lp->debug_type);
 433                 }
 434
 435                 lp->tail = NULL; /* Switch any local rules to inherited. */
 436                 if (ex->match_flags & MATCHFLG_NO_INHERIT)
 437                         lp->head = NULL;
 438
 439                 if (ex->match_flags & MATCHFLG_FINISH_SETUP) {
 440                         ex->match_flags &= ~MATCHFLG_FINISH_SETUP;
 441                         if (setup_merge_file(ex, lp))
 442                                 set_filter_dir(dir, dirlen);
 443                 }
 444
 445                 if (strlcpy(dirbuf + dirbuf_len, ex->pattern,
 446                     MAXPATHLEN - dirbuf_len) < MAXPATHLEN - dirbuf_len) {
 447                         parse_filter_file(lp, dirbuf, ex->match_flags,
 448                                           XFLG_ANCHORED2ABS);
 449                 } else {
 450                         io_error |= IOERR_GENERAL;
 451                         rprintf(FINFO,
 452                             "cannot add local filter rules in long-named directory: %s\n",
 453                             full_fname(dirbuf));
 454                 }
 455                 dirbuf[dirbuf_len] = '\0';
 456         }
 457
 458         return (void*)push;
 459 }
 460
 461 void pop_local_filters(void *mem)
 462 {
 463         struct filter_list_struct *ap, *pop = (struct filter_list_struct*)mem;
 464         int i;
 465
 466         for (i = mergelist_cnt; i-- > 0; ) {
 467                 struct filter_struct *ex = mergelist_parents[i];
 468                 struct filter_list_struct *lp = ex->u.mergelist;
 469
 470                 if (verbose > 2) {
 471                         rprintf(FINFO, "[%s] popping filter list%s\n",
 472                                 who_am_i(), lp->debug_type);
 473                 }
 474
 475                 clear_filter_list(lp);
 476         }
 477
 478         if (!pop)
 479                 return;
 480
 481         for (i = 0, ap = pop; i < mergelist_cnt; i++) {
 482                 memcpy(mergelist_parents[i]->u.mergelist, ap++,
 483                        sizeof (struct filter_list_struct));
 484         }
 485
 486         free(pop);
 487 }
 488
 489 static int rule_matches(char *name, struct filter_struct *ex, int name_is_dir)
 490 {
 491         char *p, full_name[MAXPATHLEN];
 492         int match_start = 0;
 493         int ret_match = ex->match_flags & MATCHFLG_NEGATE ? 0 : 1;
 494         char *pattern = ex->pattern;
 495
 496         if (!*name)
 497                 return 0;
 498
 499         /* If the pattern does not have any slashes AND it does not have
 500          * a "**" (which could match a slash), then we just match the
 501          * name portion of the path. */
 502         if (!ex->u.slash_cnt && !(ex->match_flags & MATCHFLG_WILD2)) {
 503                 if ((p = strrchr(name,'/')) != NULL)
 504                         name = p+1;
 505         }
 506         else if (ex->match_flags & MATCHFLG_ABS_PATH && *name != '/'
 507             && curr_dir_len > module_dirlen + 1) {
 508                 pathjoin(full_name, sizeof full_name,
 509                          curr_dir + module_dirlen + 1, name);
 510                 name = full_name;
 511         }
 512
 513         if (ex->match_flags & MATCHFLG_DIRECTORY && !name_is_dir)
 514                 return !ret_match;
 515
 516         if (*pattern == '/') {
 517                 match_start = 1;
 518                 pattern++;
 519                 if (*name == '/')
 520                         name++;
 521         }
 522
 523         if (ex->match_flags & MATCHFLG_WILD) {
 524                 /* A non-anchored match with an infix slash and no "**"
 525                  * needs to match the last slash_cnt+1 name elements. */
 526                 if (!match_start && ex->u.slash_cnt
 527                     && !(ex->match_flags & MATCHFLG_WILD2)) {
 528                         int cnt = ex->u.slash_cnt + 1;
 529                         for (p = name + strlen(name) - 1; p >= name; p--) {
 530                                 if (*p == '/' && !--cnt)
 531                                         break;
 532                         }
 533                         name = p+1;
 534                 }
 535                 if (wildmatch(pattern, name))
 536                         return ret_match;
 537                 if (ex->match_flags & MATCHFLG_WILD2_PREFIX) {
 538                         /* If the **-prefixed pattern has a '/' as the next
 539                          * character, then try to match the rest of the
 540                          * pattern at the root. */
 541                         if (pattern[2] == '/' && wildmatch(pattern+3, name))
 542                                 return ret_match;
 543                 }
 544                 else if (!match_start && ex->match_flags & MATCHFLG_WILD2) {
 545                         /* A non-anchored match with an infix or trailing "**"
 546                          * (but not a prefixed "**") needs to try matching
 547                          * after every slash. */
 548                         while ((name = strchr(name, '/')) != NULL) {
 549                                 name++;
 550                                 if (wildmatch(pattern, name))
 551                                         return ret_match;
 552                         }
 553                 }
 554         } else if (match_start) {
 555                 if (strcmp(name,pattern) == 0)
 556                         return ret_match;
 557         } else {
 558                 int l1 = strlen(name);
 559                 int l2 = strlen(pattern);
 560                 if (l2 <= l1 &&
 561                     strcmp(name+(l1-l2),pattern) == 0 &&
 562                     (l1==l2 || name[l1-(l2+1)] == '/')) {
 563                         return ret_match;
 564                 }
 565         }
 566
 567         return !ret_match;
 568 }
 569
 570
 571 static void report_filter_result(char const *name,
 572                                  struct filter_struct const *ent,
 573                                  int name_is_dir, const char *type)
 574 {
 575         /* If a trailing slash is present to match only directories,
 576          * then it is stripped out by add_rule().  So as a special
 577          * case we add it back in here. */
 578
 579         if (verbose >= 2) {
 580                 rprintf(FINFO, "[%s] %scluding %s %s because of pattern %s%s%s\n",
 581                         who_am_i(),
 582                         ent->match_flags & MATCHFLG_INCLUDE ? "in" : "ex",
 583                         name_is_dir ? "directory" : "file", name, ent->pattern,
 584                         ent->match_flags & MATCHFLG_DIRECTORY ? "/" : "", type);
 585         }
 586 }
 587
 588
 589 /*
 590  * Return -1 if file "name" is defined to be excluded by the specified
 591  * exclude list, 1 if it is included, and 0 if it was not matched.
 592  */
 593 int check_filter(struct filter_list_struct *listp, char *name, int name_is_dir)
 594 {
 595         struct filter_struct *ent;
 596
 597         for (ent = listp->head; ent; ent = ent->next) {
 598                 if (ent->match_flags & MATCHFLG_PERDIR_MERGE) {
 599                         int rc = check_filter(ent->u.mergelist, name,
 600                                               name_is_dir);
 601                         if (rc)
 602                                 return rc;
 603                         continue;
 604                 }
 605                 if (ent->match_flags & MATCHFLG_CVS_IGNORE) {
 606                         int rc = check_filter(&cvs_filter_list, name,
 607                                               name_is_dir);
 608                         if (rc)
 609                                 return rc;
 610                         continue;
 611                 }
 612                 if (rule_matches(name, ent, name_is_dir)) {
 613                         report_filter_result(name, ent, name_is_dir,
 614                                               listp->debug_type);
 615                         return ent->match_flags & MATCHFLG_INCLUDE ? 1 : -1;
 616                 }
 617         }
 618
 619         return 0;
 620 }
 621
 622 #define RULE_STRCMP(s,r) rule_strcmp((s), (r), sizeof (r) - 1)
 623
 624 static const uchar *rule_strcmp(const uchar *str, const char *rule, int rule_len)
 625 {
 626         if (strncmp((char*)str, rule, rule_len) != 0)
 627                 return NULL;
 628         if (isspace(str[rule_len]) || str[rule_len] == '_' || !str[rule_len])
 629                 return str + rule_len - 1;
 630         if (str[rule_len] == ',')
 631                 return str + rule_len;
 632         return NULL;
 633 }
 634
 635 /* Get the next include/exclude arg from the string.  The token will not
 636  * be '\0' terminated, so use the returned length to limit the string.
 637  * Also, be sure to add this length to the returned pointer before passing
 638  * it back to ask for the next token.  This routine parses the "!" (list-
 639  * clearing) token and (depending on the mflags) the various prefixes.
 640  * The *mflags_ptr value will be set on exit to the new MATCHFLG_* bits
 641  * for the current token. */
 642 static const char *parse_rule_tok(const char *p, uint32 mflags, int xflags,
 643                                   unsigned int *len_ptr, uint32 *mflags_ptr)
 644 {
 645         const uchar *s = (const uchar *)p;
 646         uint32 new_mflags;
 647         unsigned int len;
 648
 649         if (mflags & MATCHFLG_WORD_SPLIT) {
 650                 /* Skip over any initial whitespace. */
 651                 while (isspace(*s))
 652                         s++;
 653                 /* Update to point to real start of rule. */
 654                 p = (const char *)s;
 655         }
 656         if (!*s)
 657                 return NULL;
 658
 659         new_mflags = mflags & MATCHFLGS_FROM_CONTAINER;
 660
 661         /* Figure out what kind of a filter rule "s" is pointing at.  Note
 662          * that if MATCHFLG_NO_PREFIXES is set, the rule is either an include
 663          * or an exclude based on the inheritance of the MATCHFLG_INCLUDE
 664          * flag (above).  XFLG_OLD_PREFIXES indicates a compatibility mode
 665          * for old include/exclude patterns where just "+ " and "- " are
 666          * allowed as optional prefixes.  */
 667         if (mflags & MATCHFLG_NO_PREFIXES) {
 668                 if (*s == '!' && mflags & MATCHFLG_CVS_IGNORE)
 669                         new_mflags |= MATCHFLG_CLEAR_LIST; /* Tentative! */
 670         } else if (xflags & XFLG_OLD_PREFIXES) {
 671                 if (*s == '-' && s[1] == ' ') {
 672                         new_mflags &= ~MATCHFLG_INCLUDE;
 673                         s += 2;
 674                 } else if (*s == '+' && s[1] == ' ') {
 675                         new_mflags |= MATCHFLG_INCLUDE;
 676                         s += 2;
 677                 }
 678                 if (*s == '!')
 679                         new_mflags |= MATCHFLG_CLEAR_LIST; /* Tentative! */
 680         } else {
 681                 char ch = 0, *mods = "";
 682                 switch (*s) {
 683                 case 'c':
 684                         if ((s = RULE_STRCMP(s, "clear")) != NULL)
 685                                 ch = '!';
 686                         break;
 687                 case 'd':
 688                         if ((s = RULE_STRCMP(s, "dir-merge")) != NULL)
 689                                 ch = ':';
 690                         break;
 691                 case 'e':
 692                         if ((s = RULE_STRCMP(s, "exclude")) != NULL)
 693                                 ch = '-';
 694                         break;
 695                 case 'h':
 696                         if ((s = RULE_STRCMP(s, "hide")) != NULL)
 697                                 ch = 'H';
 698                         break;
 699                 case 'i':
 700                         if ((s = RULE_STRCMP(s, "include")) != NULL)
 701                                 ch = '+';
 702                         break;
 703                 case 'm':
 704                         if ((s = RULE_STRCMP(s, "merge")) != NULL)
 705                                 ch = '.';
 706                         break;
 707                 case 'p':
 708                         if ((s = RULE_STRCMP(s, "protect")) != NULL)
 709                                 ch = 'P';
 710                         break;
 711                 case 'r':
 712                         if ((s = RULE_STRCMP(s, "risk")) != NULL)
 713                                 ch = 'R';
 714                         break;
 715                 case 's':
 716                         if ((s = RULE_STRCMP(s, "show")) != NULL)
 717                                 ch = 'S';
 718                         break;
 719
 720                 default:
 721                         ch = *s;
 722                         if (s[1] == ',')
 723                                 s++;
 724                         break;
 725                 }
 726                 switch (ch) {
 727                 case ':':
 728                         new_mflags |= MATCHFLG_PERDIR_MERGE
 729                                     | MATCHFLG_FINISH_SETUP;
 730                         /* FALL THROUGH */
 731                 case '.':
 732                         new_mflags |= MATCHFLG_MERGE_FILE;
 733                         mods = MODIFIERS_INCL_EXCL MODIFIERS_MERGE_FILE;
 734                         break;
 735                 case '+':
 736                         new_mflags |= MATCHFLG_INCLUDE;
 737                         /* FALL THROUGH */
 738                 case '-':
 739                         mods = MODIFIERS_INCL_EXCL;
 740                         break;
 741                 case 'S':
 742                         new_mflags |= MATCHFLG_INCLUDE;
 743                         /* FALL THROUGH */
 744                 case 'H':
 745                         new_mflags |= MATCHFLG_SENDER_SIDE;
 746                         mods = MODIFIERS_HIDE_PROTECT;
 747                         break;
 748                 case 'R':
 749                         new_mflags |= MATCHFLG_INCLUDE;
 750                         /* FALL THROUGH */
 751                 case 'P':
 752                         new_mflags |= MATCHFLG_RECEIVER_SIDE;
 753                         mods = MODIFIERS_HIDE_PROTECT;
 754                         break;
 755                 case '!':
 756                         new_mflags |= MATCHFLG_CLEAR_LIST;
 757                         mods = NULL;
 758                         break;
 759                 default:
 760                         rprintf(FERROR, "Unknown filter rule: `%s'\n", p);
 761                         exit_cleanup(RERR_SYNTAX);
 762                 }
 763                 while (mods && *++s && *s != ' ' && *s != '_') {
 764                         if (strchr(mods, *s) == NULL) {
 765                                 if (mflags & MATCHFLG_WORD_SPLIT && isspace(*s)) {
 766                                         s--;
 767                                         break;
 768                                 }
 769                             invalid:
 770                                 rprintf(FERROR,
 771                                         "invalid modifier sequence at '%c' in filter rule: %s\n",
 772                                         *s, p);
 773                                 exit_cleanup(RERR_SYNTAX);
 774                         }
 775                         switch (*s) {
 776                         case '-':
 777                                 if (new_mflags & MATCHFLG_NO_PREFIXES)
 778                                     goto invalid;
 779                                 new_mflags |= MATCHFLG_NO_PREFIXES;
 780                                 break;
 781                         case '+':
 782                                 if (new_mflags & MATCHFLG_NO_PREFIXES)
 783                                     goto invalid;
 784                                 new_mflags |= MATCHFLG_NO_PREFIXES
 785                                             | MATCHFLG_INCLUDE;
 786                                 break;
 787                         case '/':
 788                                 new_mflags |= MATCHFLG_ABS_PATH;
 789                                 break;
 790                         case '!':
 791                                 new_mflags |= MATCHFLG_NEGATE;
 792                                 break;
 793                         case 'C':
 794                                 if (new_mflags & MATCHFLG_NO_PREFIXES)
 795                                     goto invalid;
 796                                 new_mflags |= MATCHFLG_NO_PREFIXES
 797                                             | MATCHFLG_WORD_SPLIT
 798                                             | MATCHFLG_NO_INHERIT
 799                                             | MATCHFLG_CVS_IGNORE;
 800                                 break;
 801                         case 'e':
 802                                 new_mflags |= MATCHFLG_EXCLUDE_SELF;
 803                                 break;
 804                         case 'n':
 805                                 new_mflags |= MATCHFLG_NO_INHERIT;
 806                                 break;
 807                         case 'r':
 808                                 new_mflags |= MATCHFLG_RECEIVER_SIDE;
 809                                 break;
 810                         case 's':
 811                                 new_mflags |= MATCHFLG_SENDER_SIDE;
 812                                 break;
 813                         case 'w':
 814                                 new_mflags |= MATCHFLG_WORD_SPLIT;
 815                                 break;
 816                         }
 817                 }
 818                 if (*s)
 819                         s++;
 820         }
 821
 822         if (mflags & MATCHFLG_WORD_SPLIT) {
 823                 const uchar *cp = s;
 824                 /* Token ends at whitespace or the end of the string. */
 825                 while (!isspace(*cp) && *cp != '\0')
 826                         cp++;
 827                 len = cp - s;
 828         } else
 829                 len = strlen((char*)s);
 830
 831         if (new_mflags & MATCHFLG_CLEAR_LIST) {
 832                 if (!(xflags & XFLG_OLD_PREFIXES) && len) {
 833                         rprintf(FERROR,
 834                                 "'!' rule has trailing characters: %s\n", p);
 835                         exit_cleanup(RERR_SYNTAX);
 836                 }
 837                 if (len > 1)
 838                         new_mflags &= ~MATCHFLG_CLEAR_LIST;
 839         } else if (!len && !(new_mflags & MATCHFLG_CVS_IGNORE)) {
 840                 rprintf(FERROR, "unexpected end of filter rule: %s\n", p);
 841                 exit_cleanup(RERR_SYNTAX);
 842         }
 843
 844         *len_ptr = len;
 845         *mflags_ptr = new_mflags;
 846         return (const char *)s;
 847 }
 848
 849
 850 static char default_cvsignore[] =
 851         /* These default ignored items come from the CVS manual. */
 852         "RCS SCCS CVS CVS.adm RCSLOG cvslog.* tags TAGS"
 853         " .make.state .nse_depinfo *~ #* .#* ,* _$* *$"
 854         " *.old *.bak *.BAK *.orig *.rej .del-*"
 855         " *.a *.olb *.o *.obj *.so *.exe"
 856         " *.Z *.elc *.ln core"
 857         /* The rest we added to suit ourself. */
 858         " .svn/";
 859
 860 static void get_cvs_excludes(uint32 mflags)
 861 {
 862         char *p, fname[MAXPATHLEN];
 863         static int initialized = 0;
 864
 865         if (initialized)
 866                 return;
 867         initialized = 1;
 868
 869         parse_rule(&cvs_filter_list, default_cvsignore, mflags, 0);
 870
 871         p = module_id >= 0 && lp_use_chroot(module_id) ? "/" : getenv("HOME");
 872         if (p && pathjoin(fname, MAXPATHLEN, p, ".cvsignore") < MAXPATHLEN)
 873                 parse_filter_file(&cvs_filter_list, fname, mflags, 0);
 874
 875         parse_rule(&cvs_filter_list, getenv("CVSIGNORE"), mflags, 0);
 876 }
 877
 878
 879 void parse_rule(struct filter_list_struct *listp, const char *pattern,
 880                 uint32 mflags, int xflags)
 881 {
 882         unsigned int pat_len;
 883         uint32 new_mflags;
 884         const char *cp, *p;
 885
 886         if (!pattern)
 887                 return;
 888
 889         while (1) {
 890                 /* Remember that the returned string is NOT '\0' terminated! */
 891                 cp = parse_rule_tok(pattern, mflags, xflags,
 892                                     &pat_len, &new_mflags);
 893                 if (!cp)
 894                         break;
 895                 if (pat_len >= MAXPATHLEN) {
 896                         rprintf(FERROR, "discarding over-long filter: %s\n",
 897                                 cp);
 898                         continue;
 899                 }
 900                 pattern = cp + pat_len;
 901
 902                 if (new_mflags & MATCHFLG_CLEAR_LIST) {
 903                         if (verbose > 2) {
 904                                 rprintf(FINFO,
 905                                         "[%s] clearing filter list%s\n",
 906                                         who_am_i(), listp->debug_type);
 907                         }
 908                         clear_filter_list(listp);
 909                         continue;
 910                 }
 911
 912                 if (new_mflags & MATCHFLG_MERGE_FILE) {
 913                         unsigned int len;
 914                         if (!pat_len) {
 915                                 cp = ".cvsignore";
 916                                 pat_len = 10;
 917                         }
 918                         len = pat_len;
 919                         if (new_mflags & MATCHFLG_EXCLUDE_SELF) {
 920                                 const char *name = strrchr(cp, '/');
 921                                 if (name)
 922                                         len -= ++name - cp;
 923                                 else
 924                                         name = cp;
 925                                 add_rule(listp, name, len, 0, 0);
 926                                 new_mflags &= ~MATCHFLG_EXCLUDE_SELF;
 927                                 len = pat_len;
 928                         }
 929                         if (new_mflags & MATCHFLG_PERDIR_MERGE) {
 930                                 if (parent_dirscan) {
 931                                         if (!(p = parse_merge_name(cp, &len,
 932                                                                 module_dirlen)))
 933                                                 continue;
 934                                         add_rule(listp, p, len, new_mflags, 0);
 935                                         continue;
 936                                 }
 937                         } else {
 938                                 if (!(p = parse_merge_name(cp, &len, 0)))
 939                                         continue;
 940                                 parse_filter_file(listp, p, new_mflags,
 941                                                   XFLG_FATAL_ERRORS);
 942                                 continue;
 943                         }
 944                 }
 945
 946                 add_rule(listp, cp, pat_len, new_mflags, xflags);
 947
 948                 if (new_mflags & MATCHFLG_CVS_IGNORE
 949                     && !(new_mflags & MATCHFLG_MERGE_FILE))
 950                         get_cvs_excludes(new_mflags);
 951         }
 952 }
 953
 954
 955 void parse_filter_file(struct filter_list_struct *listp, const char *fname,
 956                        uint32 mflags, int xflags)
 957 {
 958         FILE *fp;
 959         char line[MAXPATHLEN+MAX_RULE_PREFIX+1]; /* +1 for trailing slash. */
 960         char *eob = line + sizeof line - 1;
 961         int word_split = mflags & MATCHFLG_WORD_SPLIT;
 962
 963         if (!fname || !*fname)
 964                 return;
 965
 966         if (*fname != '-' || fname[1] || am_server) {
 967                 if (server_filter_list.head) {
 968                         strlcpy(line, fname, sizeof line);
 969                         clean_fname(line, 1);
 970                         if (check_filter(&server_filter_list, line, 0) < 0)
 971                                 fp = NULL;
 972                         else
 973                                 fp = fopen(line, "rb");
 974                 } else
 975                         fp = fopen(fname, "rb");
 976         } else
 977                 fp = stdin;
 978
 979         if (verbose > 2) {
 980                 rprintf(FINFO, "[%s] parse_filter_file(%s,%x,%x)%s\n",
 981                         who_am_i(), safe_fname(fname), mflags, xflags,
 982                         fp ? "" : " [not found]");
 983         }
 984
 985         if (!fp) {
 986                 if (xflags & XFLG_FATAL_ERRORS) {
 987                         rsyserr(FERROR, errno,
 988                                 "failed to open %sclude file %s",
 989                                 mflags & MATCHFLG_INCLUDE ? "in" : "ex",
 990                                 safe_fname(fname));
 991                         exit_cleanup(RERR_FILEIO);
 992                 }
 993                 return;
 994         }
 995         dirbuf[dirbuf_len] = '\0';
 996
 997         while (1) {
 998                 char *s = line;
 999                 int ch, overflow = 0;
1000                 while (1) {
1001                         if ((ch = getc(fp)) == EOF) {
1002                                 if (ferror(fp) && errno == EINTR)
1003                                         continue;
1004                                 break;
1005                         }
1006                         if (word_split && isspace(ch))
1007                                 break;
1008                         if (eol_nulls? !ch : (ch == '\n' || ch == '\r'))
1009                                 break;
1010                         if (s < eob)
1011                                 *s++ = ch;
1012                         else
1013                                 overflow = 1;
1014                 }
1015                 if (overflow) {
1016                         rprintf(FERROR, "discarding over-long filter: %s...\n", line);
1017                         s = line;
1018                 }
1019                 *s = '\0';
1020                 /* Skip an empty token and (when line parsing) comments. */
1021                 if (*line && (word_split || (*line != ';' && *line != '#')))
1022                         parse_rule(listp, line, mflags, xflags);
1023                 if (ch == EOF)
1024                         break;
1025         }
1026         fclose(fp);
1027 }
1028
1029 /* If the "for_xfer" flag is set, the prefix is made compatible with the
1030  * current protocol_version (if possible) or a NULL is returned (if not
1031  * possible). */
1032 char *get_rule_prefix(int match_flags, const char *pat, int for_xfer,
1033                       unsigned int *plen_ptr)
1034 {
1035         static char buf[MAX_RULE_PREFIX+1];
1036         char *op = buf;
1037         int legal_len = for_xfer && protocol_version < 29 ? 1 : MAX_RULE_PREFIX-1;
1038
1039         if (match_flags & MATCHFLG_PERDIR_MERGE) {
1040                 if (legal_len == 1)
1041                         return NULL;
1042                 *op++ = ':';
1043         } else if (match_flags & MATCHFLG_INCLUDE)
1044                 *op++ = '+';
1045         else if (legal_len != 1
1046             || ((*pat == '-' || *pat == '+') && pat[1] == ' '))
1047                 *op++ = '-';
1048         else
1049                 legal_len = 0;
1050
1051         if (match_flags & MATCHFLG_CVS_IGNORE)
1052                 *op++ = 'C';
1053         else {
1054                 if (match_flags & MATCHFLG_NO_INHERIT)
1055                         *op++ = 'n';
1056                 if (match_flags & MATCHFLG_WORD_SPLIT)
1057                         *op++ = 'w';
1058                 if (match_flags & MATCHFLG_NO_PREFIXES) {
1059                         if (match_flags & MATCHFLG_INCLUDE)
1060                                 *op++ = '+';
1061                         else
1062                                 *op++ = '-';
1063                 }
1064         }
1065         if (match_flags & MATCHFLG_EXCLUDE_SELF)
1066                 *op++ = 'e';
1067         if (match_flags & MATCHFLG_SENDER_SIDE
1068             && (!for_xfer || protocol_version >= 29))
1069                 *op++ = 's';
1070         if (match_flags & MATCHFLG_RECEIVER_SIDE
1071             && (!for_xfer || protocol_version >= 29
1072              || (delete_excluded && am_sender)))
1073                 *op++ = 'r';
1074         if (op - buf > legal_len)
1075                 return NULL;
1076         if (legal_len)
1077                 *op++ = ' ';
1078         *op = '\0';
1079         if (plen_ptr)
1080                 *plen_ptr = op - buf;
1081         return buf;
1082 }
1083
1084 static void send_rules(int f_out, struct filter_list_struct *flp)
1085 {
1086         struct filter_struct *ent, *prev = NULL;
1087
1088         for (ent = flp->head; ent; ent = ent->next) {
1089                 unsigned int len, plen, dlen;
1090                 int elide = 0;
1091                 char *p;
1092
1093                 if (ent->match_flags & MATCHFLG_SENDER_SIDE)
1094                         elide = am_sender ? 1 : -1;
1095                 if (ent->match_flags & MATCHFLG_RECEIVER_SIDE)
1096                         elide = elide ? 0 : am_sender ? -1 : 1;
1097                 else if (delete_excluded && !elide)
1098                         elide = am_sender ? 1 : -1;
1099                 if (elide < 0) {
1100                         if (prev)
1101                                 prev->next = ent->next;
1102                         else
1103                                 flp->head = ent->next;
1104                 } else
1105                         prev = ent;
1106                 if (elide > 0)
1107                         continue;
1108                 if (ent->match_flags & MATCHFLG_CVS_IGNORE
1109                     && !(ent->match_flags & MATCHFLG_MERGE_FILE)) {
1110                         int f = am_sender || protocol_version < 29 ? f_out : -1;
1111                         send_rules(f, &cvs_filter_list);
1112                         if (f >= 0)
1113                                 continue;
1114                 }
1115                 p = get_rule_prefix(ent->match_flags, ent->pattern, 1, &plen);
1116                 if (!p) {
1117                         rprintf(FERROR,
1118                                 "filter rules are too modern for remote rsync.\n");
1119                         exit_cleanup(RERR_SYNTAX);
1120                 }
1121                 if (f_out < 0)
1122                         continue;
1123                 len = strlen(ent->pattern);
1124                 dlen = ent->match_flags & MATCHFLG_DIRECTORY ? 1 : 0;
1125                 if (!(plen + len + dlen))
1126                         continue;
1127                 write_int(f_out, plen + len + dlen);
1128                 if (plen)
1129                         write_buf(f_out, p, plen);
1130                 write_buf(f_out, ent->pattern, len);
1131                 if (dlen)
1132                         write_byte(f_out, '/');
1133         }
1134         flp->tail = prev;
1135 }
1136
1137 /* This is only called by the client. */
1138 void send_filter_list(int f_out)
1139 {
1140         int receiver_wants_list = delete_mode
1141                 && (!delete_excluded || protocol_version >= 29);
1142
1143         if (local_server || (am_sender && !receiver_wants_list))
1144                 f_out = -1;
1145         if (cvs_exclude && am_sender) {
1146                 if (protocol_version >= 29)
1147                         parse_rule(&filter_list, ":C", 0, 0);
1148                 parse_rule(&filter_list, "-C", 0, 0);
1149         }
1150
1151         /* This is a complete hack - blame Rusty.  FIXME!
1152          * Remove this hack when older rsyncs (below 2.6.4) are gone. */
1153         if (list_only == 1 && !recurse)
1154                 parse_rule(&filter_list, "/*/*", MATCHFLG_NO_PREFIXES, 0);
1155
1156         send_rules(f_out, &filter_list);
1157
1158         if (f_out >= 0)
1159                 write_int(f_out, 0);
1160
1161         if (cvs_exclude) {
1162                 if (!am_sender || protocol_version < 29)
1163                         parse_rule(&filter_list, ":C", 0, 0);
1164                 if (!am_sender)
1165                         parse_rule(&filter_list, "-C", 0, 0);
1166         }
1167 }
1168
1169 /* This is only called by the server. */
1170 void recv_filter_list(int f_in)
1171 {
1172         char line[MAXPATHLEN+MAX_RULE_PREFIX+1]; /* +1 for trailing slash. */
1173         int xflags = protocol_version >= 29 ? 0 : XFLG_OLD_PREFIXES;
1174         int receiver_wants_list = delete_mode
1175                 && (!delete_excluded || protocol_version >= 29);
1176         unsigned int len;
1177
1178         if (!local_server && (am_sender || receiver_wants_list)) {
1179                 while ((len = read_int(f_in)) != 0) {
1180                         if (len >= sizeof line)
1181                                 overflow("recv_rules");
1182                         read_sbuf(f_in, line, len);
1183                         parse_rule(&filter_list, line, 0, xflags);
1184                 }
1185         }
1186
1187         if (cvs_exclude) {
1188                 if (local_server || am_sender || protocol_version < 29)
1189                         parse_rule(&filter_list, ":C", 0, 0);
1190                 if (local_server || am_sender)
1191                         parse_rule(&filter_list, "-C", 0, 0);
1192         }
1193
1194         if (local_server) /* filter out any rules that aren't for us. */
1195                 send_rules(-1, &filter_list);
1196 }