symlinks.c

   1 #include "cache.h"
   2
   3 static int threaded_check_leading_path(struct cache_def *cache, const char *name, int len);
   4 static int threaded_has_dirs_only_path(struct cache_def *cache, const char *name, int len, int prefix_len);
   5
   6 /*
   7  * Returns the length (on a path component basis) of the longest
   8  * common prefix match of 'name_a' and 'name_b'.
   9  */
  10 static int longest_path_match(const char *name_a, int len_a,
  11                               const char *name_b, int len_b,
  12                               int *previous_slash)
  13 {
  14         int max_len, match_len = 0, match_len_prev = 0, i = 0;
  15
  16         max_len = len_a < len_b ? len_a : len_b;
  17         while (i < max_len && name_a[i] == name_b[i]) {
  18                 if (name_a[i] == '/') {
  19                         match_len_prev = match_len;
  20                         match_len = i;
  21                 }
  22                 i++;
  23         }
  24         /*
  25          * Is 'name_b' a substring of 'name_a', the other way around,
  26          * or is 'name_a' and 'name_b' the exact same string?
  27          */
  28         if (i >= max_len && ((len_a > len_b && name_a[len_b] == '/') ||
  29                              (len_a < len_b && name_b[len_a] == '/') ||
  30                              (len_a == len_b))) {
  31                 match_len_prev = match_len;
  32                 match_len = i;
  33         }
  34         *previous_slash = match_len_prev;
  35         return match_len;
  36 }
  37
  38 static struct cache_def default_cache = CACHE_DEF_INIT;
  39
  40 static inline void reset_lstat_cache(struct cache_def *cache)
  41 {
  42         strbuf_reset(&cache->path);
  43         cache->flags = 0;
  44         /*
  45          * The track_flags and prefix_len_stat_func members is only
  46          * set by the safeguard rule inside lstat_cache()
  47          */
  48 }
  49
  50 #define FL_DIR      (1 << 0)
  51 #define FL_NOENT    (1 << 1)
  52 #define FL_SYMLINK  (1 << 2)
  53 #define FL_LSTATERR (1 << 3)
  54 #define FL_ERR      (1 << 4)
  55 #define FL_FULLPATH (1 << 5)
  56
  57 /*
  58  * Check if name 'name' of length 'len' has a symlink leading
  59  * component, or if the directory exists and is real, or not.
  60  *
  61  * To speed up the check, some information is allowed to be cached.
  62  * This can be indicated by the 'track_flags' argument, which also can
  63  * be used to indicate that we should check the full path.
  64  *
  65  * The 'prefix_len_stat_func' parameter can be used to set the length
  66  * of the prefix, where the cache should use the stat() function
  67  * instead of the lstat() function to test each path component.
  68  */
  69 static int lstat_cache_matchlen(struct cache_def *cache,
  70                                 const char *name, int len,
  71                                 int *ret_flags, int track_flags,
  72                                 int prefix_len_stat_func)
  73 {
  74         int match_len, last_slash, last_slash_dir, previous_slash;
  75         int save_flags, ret;
  76         struct stat st;
  77
  78         if (cache->track_flags != track_flags ||
  79             cache->prefix_len_stat_func != prefix_len_stat_func) {
  80                 /*
  81                  * As a safeguard rule we clear the cache if the
  82                  * values of track_flags and/or prefix_len_stat_func
  83                  * does not match with the last supplied values.
  84                  */
  85                 reset_lstat_cache(cache);
  86                 cache->track_flags = track_flags;
  87                 cache->prefix_len_stat_func = prefix_len_stat_func;
  88                 match_len = last_slash = 0;
  89         } else {
  90                 /*
  91                  * Check to see if we have a match from the cache for
  92                  * the 2 "excluding" path types.
  93                  */
  94                 match_len = last_slash =
  95                         longest_path_match(name, len, cache->path.buf,
  96                                            cache->path.len, &previous_slash);
  97                 *ret_flags = cache->flags & track_flags & (FL_NOENT|FL_SYMLINK);
  98
  99                 if (!(track_flags & FL_FULLPATH) && match_len == len)
 100                         match_len = last_slash = previous_slash;
 101
 102                 if (*ret_flags && match_len == cache->path.len)
 103                         return match_len;
 104                 /*
 105                  * If we now have match_len > 0, we would know that
 106                  * the matched part will always be a directory.
 107                  *
 108                  * Also, if we are tracking directories and 'name' is
 109                  * a substring of the cache on a path component basis,
 110                  * we can return immediately.
 111                  */
 112                 *ret_flags = track_flags & FL_DIR;
 113                 if (*ret_flags && len == match_len)
 114                         return match_len;
 115         }
 116
 117         /*
 118          * Okay, no match from the cache so far, so now we have to
 119          * check the rest of the path components.
 120          */
 121         *ret_flags = FL_DIR;
 122         last_slash_dir = last_slash;
 123         if (len > cache->path.len)
 124                 strbuf_grow(&cache->path, len - cache->path.len);
 125         while (match_len < len) {
 126                 do {
 127                         cache->path.buf[match_len] = name[match_len];
 128                         match_len++;
 129                 } while (match_len < len && name[match_len] != '/');
 130                 if (match_len >= len && !(track_flags & FL_FULLPATH))
 131                         break;
 132                 last_slash = match_len;
 133                 cache->path.buf[last_slash] = '\0';
 134
 135                 if (last_slash <= prefix_len_stat_func)
 136                         ret = stat(cache->path.buf, &st);
 137                 else
 138                         ret = lstat(cache->path.buf, &st);
 139
 140                 if (ret) {
 141                         *ret_flags = FL_LSTATERR;
 142                         if (errno == ENOENT)
 143                                 *ret_flags |= FL_NOENT;
 144                 } else if (S_ISDIR(st.st_mode)) {
 145                         last_slash_dir = last_slash;
 146                         continue;
 147                 } else if (S_ISLNK(st.st_mode)) {
 148                         *ret_flags = FL_SYMLINK;
 149                 } else {
 150                         *ret_flags = FL_ERR;
 151                 }
 152                 break;
 153         }
 154
 155         /*
 156          * At the end update the cache.  Note that max 3 different
 157          * path types, FL_NOENT, FL_SYMLINK and FL_DIR, can be cached
 158          * for the moment!
 159          */
 160         save_flags = *ret_flags & track_flags & (FL_NOENT|FL_SYMLINK);
 161         if (save_flags && last_slash > 0) {
 162                 cache->path.buf[last_slash] = '\0';
 163                 cache->path.len = last_slash;
 164                 cache->flags = save_flags;
 165         } else if ((track_flags & FL_DIR) && last_slash_dir > 0) {
 166                 /*
 167                  * We have a separate test for the directory case,
 168                  * since it could be that we have found a symlink or a
 169                  * non-existing directory and the track_flags says
 170                  * that we cannot cache this fact, so the cache would
 171                  * then have been left empty in this case.
 172                  *
 173                  * But if we are allowed to track real directories, we
 174                  * can still cache the path components before the last
 175                  * one (the found symlink or non-existing component).
 176                  */
 177                 cache->path.buf[last_slash_dir] = '\0';
 178                 cache->path.len = last_slash_dir;
 179                 cache->flags = FL_DIR;
 180         } else {
 181                 reset_lstat_cache(cache);
 182         }
 183         return match_len;
 184 }
 185
 186 static int lstat_cache(struct cache_def *cache, const char *name, int len,
 187                        int track_flags, int prefix_len_stat_func)
 188 {
 189         int flags;
 190         (void)lstat_cache_matchlen(cache, name, len, &flags, track_flags,
 191                         prefix_len_stat_func);
 192         return flags;
 193 }
 194
 195 #define USE_ONLY_LSTAT  0
 196
 197 /*
 198  * Return non-zero if path 'name' has a leading symlink component
 199  */
 200 int threaded_has_symlink_leading_path(struct cache_def *cache, const char *name, int len)
 201 {
 202         return lstat_cache(cache, name, len, FL_SYMLINK|FL_DIR, USE_ONLY_LSTAT) & FL_SYMLINK;
 203 }
 204
 205 /*
 206  * Return non-zero if path 'name' has a leading symlink component
 207  */
 208 int has_symlink_leading_path(const char *name, int len)
 209 {
 210         return threaded_has_symlink_leading_path(&default_cache, name, len);
 211 }
 212
 213 /*
 214  * Return zero if path 'name' has a leading symlink component or
 215  * if some leading path component does not exists.
 216  *
 217  * Return -1 if leading path exists and is a directory.
 218  *
 219  * Return path length if leading path exists and is neither a
 220  * directory nor a symlink.
 221  */
 222 int check_leading_path(const char *name, int len)
 223 {
 224         return threaded_check_leading_path(&default_cache, name, len);
 225 }
 226
 227 /*
 228  * Return zero if path 'name' has a leading symlink component or
 229  * if some leading path component does not exists.
 230  *
 231  * Return -1 if leading path exists and is a directory.
 232  *
 233  * Return path length if leading path exists and is neither a
 234  * directory nor a symlink.
 235  */
 236 static int threaded_check_leading_path(struct cache_def *cache, const char *name, int len)
 237 {
 238         int flags;
 239         int match_len = lstat_cache_matchlen(cache, name, len, &flags,
 240                            FL_SYMLINK|FL_NOENT|FL_DIR, USE_ONLY_LSTAT);
 241         if (flags & FL_NOENT)
 242                 return 0;
 243         else if (flags & FL_DIR)
 244                 return -1;
 245         else
 246                 return match_len;
 247 }
 248
 249 /*
 250  * Return non-zero if all path components of 'name' exists as a
 251  * directory.  If prefix_len > 0, we will test with the stat()
 252  * function instead of the lstat() function for a prefix length of
 253  * 'prefix_len', thus we then allow for symlinks in the prefix part as
 254  * long as those points to real existing directories.
 255  */
 256 int has_dirs_only_path(const char *name, int len, int prefix_len)
 257 {
 258         return threaded_has_dirs_only_path(&default_cache, name, len, prefix_len);
 259 }
 260
 261 /*
 262  * Return non-zero if all path components of 'name' exists as a
 263  * directory.  If prefix_len > 0, we will test with the stat()
 264  * function instead of the lstat() function for a prefix length of
 265  * 'prefix_len', thus we then allow for symlinks in the prefix part as
 266  * long as those points to real existing directories.
 267  */
 268 static int threaded_has_dirs_only_path(struct cache_def *cache, const char *name, int len, int prefix_len)
 269 {
 270         return lstat_cache(cache, name, len,
 271                            FL_DIR|FL_FULLPATH, prefix_len) &
 272                 FL_DIR;
 273 }
 274
 275 static struct strbuf removal = STRBUF_INIT;
 276
 277 static void do_remove_scheduled_dirs(int new_len)
 278 {
 279         while (removal.len > new_len) {
 280                 removal.buf[removal.len] = '\0';
 281                 if (rmdir(removal.buf))
 282                         break;
 283                 do {
 284                         removal.len--;
 285                 } while (removal.len > new_len &&
 286                          removal.buf[removal.len] != '/');
 287         }
 288         removal.len = new_len;
 289 }
 290
 291 void schedule_dir_for_removal(const char *name, int len)
 292 {
 293         int match_len, last_slash, i, previous_slash;
 294
 295         match_len = last_slash = i =
 296                 longest_path_match(name, len, removal.buf, removal.len,
 297                                    &previous_slash);
 298         /* Find last slash inside 'name' */
 299         while (i < len) {
 300                 if (name[i] == '/')
 301                         last_slash = i;
 302                 i++;
 303         }
 304
 305         /*
 306          * If we are about to go down the directory tree, we check if
 307          * we must first go upwards the tree, such that we then can
 308          * remove possible empty directories as we go upwards.
 309          */
 310         if (match_len < last_slash && match_len < removal.len)
 311                 do_remove_scheduled_dirs(match_len);
 312         /*
 313          * If we go deeper down the directory tree, we only need to
 314          * save the new path components as we go down.
 315          */
 316         if (match_len < last_slash)
 317                 strbuf_add(&removal, &name[match_len], last_slash - match_len);
 318 }
 319
 320 void remove_scheduled_dirs(void)
 321 {
 322         do_remove_scheduled_dirs(0);
 323 }