name-hash.c

   1 /*
   2  * name-hash.c
   3  *
   4  * Hashing names in the index state
   5  *
   6  * Copyright (C) 2008 Linus Torvalds
   7  */
   8 #define NO_THE_INDEX_COMPATIBILITY_MACROS
   9 #include "cache.h"
  10
  11 /*
  12  * This removes bit 5 if bit 6 is set.
  13  *
  14  * That will make US-ASCII characters hash to their upper-case
  15  * equivalent. We could easily do this one whole word at a time,
  16  * but that's for future worries.
  17  */
  18 static inline unsigned char icase_hash(unsigned char c)
  19 {
  20         return c & ~((c & 0x40) >> 1);
  21 }
  22
  23 static unsigned int hash_name(const char *name, int namelen)
  24 {
  25         unsigned int hash = 0x123;
  26
  27         while (namelen--) {
  28                 unsigned char c = *name++;
  29                 c = icase_hash(c);
  30                 hash = hash*101 + c;
  31         }
  32         return hash;
  33 }
  34
  35 struct dir_entry {
  36         struct dir_entry *next;
  37         struct dir_entry *parent;
  38         struct cache_entry *ce;
  39         int nr;
  40         unsigned int namelen;
  41 };
  42
  43 static struct dir_entry *find_dir_entry(struct index_state *istate,
  44                 const char *name, unsigned int namelen)
  45 {
  46         unsigned int hash = hash_name(name, namelen);
  47         struct dir_entry *dir;
  48
  49         for (dir = lookup_hash(hash, &istate->dir_hash); dir; dir = dir->next)
  50                 if (dir->namelen == namelen &&
  51                     !strncasecmp(dir->ce->name, name, namelen))
  52                         return dir;
  53         return NULL;
  54 }
  55
  56 static struct dir_entry *hash_dir_entry(struct index_state *istate,
  57                 struct cache_entry *ce, int namelen)
  58 {
  59         /*
  60          * Throw each directory component in the hash for quick lookup
  61          * during a git status. Directory components are stored with their
  62          * closing slash.  Despite submodules being a directory, they never
  63          * reach this point, because they are stored without a closing slash
  64          * in index_state.name_hash (as ordinary cache_entries).
  65          *
  66          * Note that the cache_entry stored with the dir_entry merely
  67          * supplies the name of the directory (up to dir_entry.namelen). We
  68          * track the number of 'active' files in a directory in dir_entry.nr,
  69          * so we can tell if the directory is still relevant, e.g. for git
  70          * status. However, if cache_entries are removed, we cannot pinpoint
  71          * an exact cache_entry that's still active. It is very possible that
  72          * multiple dir_entries point to the same cache_entry.
  73          */
  74         struct dir_entry *dir;
  75
  76         /* get length of parent directory */
  77         while (namelen > 0 && !is_dir_sep(ce->name[namelen - 1]))
  78                 namelen--;
  79         if (namelen <= 0)
  80                 return NULL;
  81
  82         /* lookup existing entry for that directory */
  83         dir = find_dir_entry(istate, ce->name, namelen);
  84         if (!dir) {
  85                 /* not found, create it and add to hash table */
  86                 void **pdir;
  87                 unsigned int hash = hash_name(ce->name, namelen);
  88
  89                 dir = xcalloc(1, sizeof(struct dir_entry));
  90                 dir->namelen = namelen;
  91                 dir->ce = ce;
  92
  93                 pdir = insert_hash(hash, dir, &istate->dir_hash);
  94                 if (pdir) {
  95                         dir->next = *pdir;
  96                         *pdir = dir;
  97                 }
  98
  99                 /* recursively add missing parent directories */
 100                 dir->parent = hash_dir_entry(istate, ce, namelen - 1);
 101         }
 102         return dir;
 103 }
 104
 105 static void add_dir_entry(struct index_state *istate, struct cache_entry *ce)
 106 {
 107         /* Add reference to the directory entry (and parents if 0). */
 108         struct dir_entry *dir = hash_dir_entry(istate, ce, ce_namelen(ce));
 109         while (dir && !(dir->nr++))
 110                 dir = dir->parent;
 111 }
 112
 113 static void remove_dir_entry(struct index_state *istate, struct cache_entry *ce)
 114 {
 115         /*
 116          * Release reference to the directory entry (and parents if 0).
 117          *
 118          * Note: we do not remove / free the entry because there's no
 119          * hash.[ch]::remove_hash and dir->next may point to other entries
 120          * that are still valid, so we must not free the memory.
 121          */
 122         struct dir_entry *dir = hash_dir_entry(istate, ce, ce_namelen(ce));
 123         while (dir && dir->nr && !(--dir->nr))
 124                 dir = dir->parent;
 125 }
 126
 127 static void hash_index_entry(struct index_state *istate, struct cache_entry *ce)
 128 {
 129         void **pos;
 130         unsigned int hash;
 131
 132         if (ce->ce_flags & CE_HASHED)
 133                 return;
 134         ce->ce_flags |= CE_HASHED;
 135         ce->next = NULL;
 136         hash = hash_name(ce->name, ce_namelen(ce));
 137         pos = insert_hash(hash, ce, &istate->name_hash);
 138         if (pos) {
 139                 ce->next = *pos;
 140                 *pos = ce;
 141         }
 142
 143         if (ignore_case && !(ce->ce_flags & CE_UNHASHED))
 144                 add_dir_entry(istate, ce);
 145 }
 146
 147 static void lazy_init_name_hash(struct index_state *istate)
 148 {
 149         int nr;
 150
 151         if (istate->name_hash_initialized)
 152                 return;
 153         for (nr = 0; nr < istate->cache_nr; nr++)
 154                 hash_index_entry(istate, istate->cache[nr]);
 155         istate->name_hash_initialized = 1;
 156 }
 157
 158 void add_name_hash(struct index_state *istate, struct cache_entry *ce)
 159 {
 160         /* if already hashed, add reference to directory entries */
 161         if (ignore_case && (ce->ce_flags & CE_STATE_MASK) == CE_STATE_MASK)
 162                 add_dir_entry(istate, ce);
 163
 164         ce->ce_flags &= ~CE_UNHASHED;
 165         if (istate->name_hash_initialized)
 166                 hash_index_entry(istate, ce);
 167 }
 168
 169 /*
 170  * We don't actually *remove* it, we can just mark it invalid so that
 171  * we won't find it in lookups.
 172  *
 173  * Not only would we have to search the lists (simple enough), but
 174  * we'd also have to rehash other hash buckets in case this makes the
 175  * hash bucket empty (common). So it's much better to just mark
 176  * it.
 177  */
 178 void remove_name_hash(struct index_state *istate, struct cache_entry *ce)
 179 {
 180         /* if already hashed, release reference to directory entries */
 181         if (ignore_case && (ce->ce_flags & CE_STATE_MASK) == CE_HASHED)
 182                 remove_dir_entry(istate, ce);
 183
 184         ce->ce_flags |= CE_UNHASHED;
 185 }
 186
 187 static int slow_same_name(const char *name1, int len1, const char *name2, int len2)
 188 {
 189         if (len1 != len2)
 190                 return 0;
 191
 192         while (len1) {
 193                 unsigned char c1 = *name1++;
 194                 unsigned char c2 = *name2++;
 195                 len1--;
 196                 if (c1 != c2) {
 197                         c1 = toupper(c1);
 198                         c2 = toupper(c2);
 199                         if (c1 != c2)
 200                                 return 0;
 201                 }
 202         }
 203         return 1;
 204 }
 205
 206 static int same_name(const struct cache_entry *ce, const char *name, int namelen, int icase)
 207 {
 208         int len = ce_namelen(ce);
 209
 210         /*
 211          * Always do exact compare, even if we want a case-ignoring comparison;
 212          * we do the quick exact one first, because it will be the common case.
 213          */
 214         if (len == namelen && !cache_name_compare(name, namelen, ce->name, len))
 215                 return 1;
 216
 217         if (!icase)
 218                 return 0;
 219
 220         return slow_same_name(name, namelen, ce->name, len);
 221 }
 222
 223 struct cache_entry *index_name_exists(struct index_state *istate, const char *name, int namelen, int icase)
 224 {
 225         unsigned int hash = hash_name(name, namelen);
 226         struct cache_entry *ce;
 227
 228         lazy_init_name_hash(istate);
 229         ce = lookup_hash(hash, &istate->name_hash);
 230
 231         while (ce) {
 232                 if (!(ce->ce_flags & CE_UNHASHED)) {
 233                         if (same_name(ce, name, namelen, icase))
 234                                 return ce;
 235                 }
 236                 ce = ce->next;
 237         }
 238
 239         /*
 240          * When looking for a directory (trailing '/'), it might be a
 241          * submodule or a directory. Despite submodules being directories,
 242          * they are stored in the name hash without a closing slash.
 243          * When ignore_case is 1, directories are stored in a separate hash
 244          * table *with* their closing slash.
 245          *
 246          * The side effect of this storage technique is we have need to
 247          * lookup the directory in a separate hash table, and if not found
 248          * remove the slash from name and perform the lookup again without
 249          * the slash.  If a match is made, S_ISGITLINK(ce->mode) will be
 250          * true.
 251          */
 252         if (icase && name[namelen - 1] == '/') {
 253                 struct dir_entry *dir = find_dir_entry(istate, name, namelen);
 254                 if (dir && dir->nr)
 255                         return dir->ce;
 256
 257                 ce = index_name_exists(istate, name, namelen - 1, icase);
 258                 if (ce && S_ISGITLINK(ce->ce_mode))
 259                         return ce;
 260         }
 261         return NULL;
 262 }
 263
 264 static int free_dir_entry(void *entry, void *unused)
 265 {
 266         struct dir_entry *dir = entry;
 267         while (dir) {
 268                 struct dir_entry *next = dir->next;
 269                 free(dir);
 270                 dir = next;
 271         }
 272         return 0;
 273 }
 274
 275 void free_name_hash(struct index_state *istate)
 276 {
 277         if (!istate->name_hash_initialized)
 278                 return;
 279         istate->name_hash_initialized = 0;
 280         if (ignore_case)
 281                 /* free directory entries */
 282                 for_each_hash(&istate->dir_hash, free_dir_entry, NULL);
 283
 284         free_hash(&istate->name_hash);
 285         free_hash(&istate->dir_hash);
 286 }