lib/hash.h

   1 /* hash - hashing table processing.
   2    Copyright (C) 1998-1999, 2001, 2003, 2009-2024 Free Software Foundation,
   3    Inc.
   4    Written by Jim Meyering <meyering@ascend.com>, 1998.
   5
   6    This file is free software: you can redistribute it and/or modify
   7    it under the terms of the GNU Lesser General Public License as
   8    published by the Free Software Foundation; either version 2.1 of the
   9    License, or (at your option) any later version.
  10
  11    This file is distributed in the hope that it will be useful,
  12    but WITHOUT ANY WARRANTY; without even the implied warranty of
  13    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14    GNU Lesser General Public License for more details.
  15
  16    You should have received a copy of the GNU Lesser General Public License
  17    along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
  18
  19 /* A generic hash table package.  */
  20
  21 /* Make sure USE_OBSTACK is defined to 1 if you want the allocator to use
  22    obstacks instead of malloc, and recompile 'hash.c' with same setting.  */
  23
  24 #ifndef HASH_H_
  25 # define HASH_H_
  26
  27 /* This file uses _GL_ATTRIBUTE_DEALLOC, _GL_ATTRIBUTE_DEPRECATED,
  28    _GL_ATTRIBUTE_MALLOC, _GL_ATTRIBUTE_NODISCARD, _GL_ATTRIBUTE_PURE,
  29    _GL_ATTRIBUTE_RETURNS_NONNULL.  */
  30 #if !_GL_CONFIG_H_INCLUDED
  31  #error "Please include config.h first."
  32 #endif
  33
  34 # include <stdio.h>
  35
  36 # ifdef __cplusplus
  37 extern "C" {
  38 # endif
  39
  40 struct hash_tuning
  41   {
  42     /* This structure is mainly used for 'hash_initialize', see the block
  43        documentation of 'hash_reset_tuning' for more complete comments.  */
  44
  45     float shrink_threshold;     /* ratio of used buckets to trigger a shrink */
  46     float shrink_factor;        /* ratio of new smaller size to original size */
  47     float growth_threshold;     /* ratio of used buckets to trigger a growth */
  48     float growth_factor;        /* ratio of new bigger size to original size */
  49     bool is_n_buckets;          /* if CANDIDATE really means table size */
  50   };
  51
  52 typedef struct hash_tuning Hash_tuning;
  53
  54 struct hash_table;
  55
  56 typedef struct hash_table Hash_table;
  57
  58 /*
  59  * Information and lookup.
  60  */
  61
  62 /* The following few functions provide information about the overall hash
  63    table organization: the number of entries, number of buckets and maximum
  64    length of buckets.  */
  65
  66 /* Return the number of buckets in the hash table.  The table size, the total
  67    number of buckets (used plus unused), or the maximum number of slots, are
  68    the same quantity.  */
  69 extern size_t hash_get_n_buckets (const Hash_table *table)
  70        _GL_ATTRIBUTE_PURE;
  71
  72 /* Return the number of slots in use (non-empty buckets).  */
  73 extern size_t hash_get_n_buckets_used (const Hash_table *table)
  74        _GL_ATTRIBUTE_PURE;
  75
  76 /* Return the number of active entries.  */
  77 extern size_t hash_get_n_entries (const Hash_table *table)
  78        _GL_ATTRIBUTE_PURE;
  79
  80 /* Return the length of the longest chain (bucket).  */
  81 extern size_t hash_get_max_bucket_length (const Hash_table *table)
  82        _GL_ATTRIBUTE_PURE;
  83
  84 /* Do a mild validation of a hash table, by traversing it and checking two
  85    statistics.  */
  86 extern bool hash_table_ok (const Hash_table *table)
  87        _GL_ATTRIBUTE_PURE;
  88
  89 extern void hash_print_statistics (const Hash_table *table, FILE *stream);
  90
  91 /* If ENTRY matches an entry already in the hash table, return the
  92    entry from the table.  Otherwise, return NULL.  */
  93 extern void *hash_lookup (const Hash_table *table, const void *entry);
  94
  95 /*
  96  * Walking.
  97  */
  98
  99 /* The functions in this page traverse the hash table and process the
 100    contained entries.  For the traversal to work properly, the hash table
 101    should not be resized nor modified while any particular entry is being
 102    processed.  In particular, entries should not be added, and an entry
 103    may be removed only if there is no shrink threshold and the entry being
 104    removed has already been passed to hash_get_next.  */
 105
 106 /* Return the first data in the table, or NULL if the table is empty.  */
 107 extern void *hash_get_first (const Hash_table *table)
 108        _GL_ATTRIBUTE_PURE;
 109
 110 /* Return the user data for the entry following ENTRY, where ENTRY has been
 111    returned by a previous call to either 'hash_get_first' or 'hash_get_next'.
 112    Return NULL if there are no more entries.  */
 113 extern void *hash_get_next (const Hash_table *table, const void *entry);
 114
 115 /* Fill BUFFER with pointers to active user entries in the hash table, then
 116    return the number of pointers copied.  Do not copy more than BUFFER_SIZE
 117    pointers.  */
 118 extern size_t hash_get_entries (const Hash_table *table, void **buffer,
 119                                 size_t buffer_size);
 120
 121 typedef bool (*Hash_processor) (void *entry, void *processor_data);
 122
 123 /* Call a PROCESSOR function for each entry of a hash table, and return the
 124    number of entries for which the processor function returned success.  A
 125    pointer to some PROCESSOR_DATA which will be made available to each call to
 126    the processor function.  The PROCESSOR accepts two arguments: the first is
 127    the user entry being walked into, the second is the value of PROCESSOR_DATA
 128    as received.  The walking continue for as long as the PROCESSOR function
 129    returns nonzero.  When it returns zero, the walking is interrupted.  */
 130 extern size_t hash_do_for_each (const Hash_table *table,
 131                                 Hash_processor processor, void *processor_data);
 132
 133 /*
 134  * Allocation and clean-up.
 135  */
 136
 137 /* Return a hash index for a NUL-terminated STRING between 0 and N_BUCKETS-1.
 138    This is a convenience routine for constructing other hashing functions.  */
 139 extern size_t hash_string (const char *string, size_t n_buckets)
 140        _GL_ATTRIBUTE_PURE;
 141
 142 extern void hash_reset_tuning (Hash_tuning *tuning);
 143
 144 typedef size_t (*Hash_hasher) (const void *entry, size_t table_size);
 145 typedef bool (*Hash_comparator) (const void *entry1, const void *entry2);
 146 typedef void (*Hash_data_freer) (void *entry);
 147
 148 /* Reclaim all storage associated with a hash table.  If a data_freer
 149    function has been supplied by the user when the hash table was created,
 150    this function applies it to the data of each entry before freeing that
 151    entry.  This function preserves errno, like 'free'.  */
 152 extern void hash_free (Hash_table *table);
 153
 154 /* Allocate and return a new hash table, or NULL upon failure.  The initial
 155    number of buckets is automatically selected so as to _guarantee_ that you
 156    may insert at least CANDIDATE different user entries before any growth of
 157    the hash table size occurs.  So, if have a reasonably tight a-priori upper
 158    bound on the number of entries you intend to insert in the hash table, you
 159    may save some table memory and insertion time, by specifying it here.  If
 160    the IS_N_BUCKETS field of the TUNING structure is true, the CANDIDATE
 161    argument has its meaning changed to the wanted number of buckets.
 162
 163    TUNING points to a structure of user-supplied values, in case some fine
 164    tuning is wanted over the default behavior of the hasher.  If TUNING is
 165    NULL, the default tuning parameters are used instead.  If TUNING is
 166    provided but the values requested are out of bounds or might cause
 167    rounding errors, return NULL.
 168
 169    The user-supplied HASHER function, when not NULL, accepts two
 170    arguments ENTRY and TABLE_SIZE.  It computes, by hashing ENTRY contents, a
 171    slot number for that entry which should be in the range 0..TABLE_SIZE-1.
 172    This slot number is then returned.
 173
 174    The user-supplied COMPARATOR function, when not NULL, accepts two
 175    arguments pointing to user data, it then returns true for a pair of entries
 176    that compare equal, or false otherwise.  This function is internally called
 177    on entries which are already known to hash to the same bucket index,
 178    but which are distinct pointers.
 179
 180    The user-supplied DATA_FREER function, when not NULL, may be later called
 181    with the user data as an argument, just before the entry containing the
 182    data gets freed.  This happens from within 'hash_free' or 'hash_clear'.
 183    You should specify this function only if you want these functions to free
 184    all of your 'data' data.  This is typically the case when your data is
 185    simply an auxiliary struct that you have malloc'd to aggregate several
 186    values.
 187
 188    Set errno on failure; otherwise errno is unspecified.  */
 189 _GL_ATTRIBUTE_NODISCARD
 190 extern Hash_table *hash_initialize (size_t candidate,
 191                                     const Hash_tuning *tuning,
 192                                     Hash_hasher hasher,
 193                                     Hash_comparator comparator,
 194                                     Hash_data_freer data_freer)
 195   _GL_ATTRIBUTE_MALLOC _GL_ATTRIBUTE_DEALLOC (hash_free, 1);
 196
 197 /* Like hash_initialize, but invokes xalloc_die instead of returning NULL.  */
 198 /* This function is defined by module 'xhash'.  */
 199 _GL_ATTRIBUTE_NODISCARD
 200 extern Hash_table *hash_xinitialize (size_t candidate,
 201                                      const Hash_tuning *tuning,
 202                                      Hash_hasher hasher,
 203                                      Hash_comparator comparator,
 204                                      Hash_data_freer data_freer)
 205   _GL_ATTRIBUTE_MALLOC _GL_ATTRIBUTE_DEALLOC (hash_free, 1)
 206   _GL_ATTRIBUTE_RETURNS_NONNULL;
 207
 208 /* Make all buckets empty, placing any chained entries on the free list.
 209    Apply the user-specified function data_freer (if any) to the data of any
 210    affected entries.  */
 211 extern void hash_clear (Hash_table *table);
 212
 213 /*
 214  * Insertion and deletion.
 215  */
 216
 217 /* For an already existing hash table, change the number of buckets through
 218    specifying CANDIDATE.  The contents of the hash table are preserved.  The
 219    new number of buckets is automatically selected so as to _guarantee_ that
 220    the table may receive at least CANDIDATE different user entries, including
 221    those already in the table, before any other growth of the hash table size
 222    occurs.  If TUNING->IS_N_BUCKETS is true, then CANDIDATE specifies the
 223    exact number of buckets desired.  Return true iff the rehash succeeded,
 224    false (setting errno) otherwise.  */
 225 _GL_ATTRIBUTE_NODISCARD
 226 extern bool hash_rehash (Hash_table *table, size_t candidate);
 227
 228 /* If ENTRY matches an entry already in the hash table, return the pointer
 229    to the entry from the table.  Otherwise, insert ENTRY and return ENTRY.
 230    Return NULL (setting errno) if the storage required for insertion
 231    cannot be allocated.  This implementation does not support
 232    duplicate entries or insertion of NULL.  */
 233 _GL_ATTRIBUTE_NODISCARD
 234 extern void *hash_insert (Hash_table *table, const void *entry);
 235
 236 /* Same as hash_insert, but invokes xalloc_die instead of returning NULL.  */
 237 /* This function is defined by module 'xhash'.  */
 238 extern void *hash_xinsert (Hash_table *table, const void *entry);
 239
 240 /* Insert ENTRY into hash TABLE if there is not already a matching entry.
 241
 242    Return -1 (setting errno) upon memory allocation failure.
 243    Return 1 if insertion succeeded.
 244    Return 0 if there is already a matching entry in the table,
 245    and in that case, if MATCHED_ENT is non-NULL, set *MATCHED_ENT
 246    to that entry.
 247
 248    This interface is easier to use than hash_insert when you must
 249    distinguish between the latter two cases.  More importantly,
 250    hash_insert is unusable for some types of ENTRY values.  When using
 251    hash_insert, the only way to distinguish those cases is to compare
 252    the return value and ENTRY.  That works only when you can have two
 253    different ENTRY values that point to data that compares "equal".  Thus,
 254    when the ENTRY value is a simple scalar, you must use
 255    hash_insert_if_absent.  ENTRY must not be NULL.  */
 256 extern int hash_insert_if_absent (Hash_table *table, const void *entry,
 257                                   const void **matched_ent);
 258
 259 /* If ENTRY is already in the table, remove it and return the just-deleted
 260    data (the user may want to deallocate its storage).  If ENTRY is not in the
 261    table, don't modify the table and return NULL.  */
 262 extern void *hash_remove (Hash_table *table, const void *entry);
 263
 264 /* Same as hash_remove.  This interface is deprecated.
 265    FIXME: Remove in 2022.  */
 266 _GL_ATTRIBUTE_DEPRECATED
 267 extern void *hash_delete (Hash_table *table, const void *entry);
 268
 269 # ifdef __cplusplus
 270 }
 271 # endif
 272
 273 #endif