uniqstr.c

   1 /* uniqstr.c -- implementation of unique strings used for shortage
   2    memory footprint by lot of strings.
   3
   4    Copyright (C) 2022 Sergey Sushilin <sergeysushilin@protonmail.com>
   5
   6    This file is part of Build.
   7
   8    Build is free software: you can redistribute it and/or
   9    modify it under the terms of either the GNU General Public License
  10    as published by the Free Software Foundation;
  11    either version 2 of the License, or version 3 of the License,
  12    or both in parallel, as here.
  13
  14    Build is distributed in the hope that it will be useful,
  15    but WITHOUT ANY WARRANTY; without even the implied warranty of
  16    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17    General Public License for more details.
  18
  19    You should have received copies of the GNU General Public License
  20    version 2 and 3 along with this program.
  21    If not, see http://www.gnu.org/licenses/.  */
  22
  23 #include <errno.h>
  24 #include <stddef.h>
  25 #include <stdint.h>
  26 #include <stdlib.h>
  27 #include <string.h>
  28
  29 #include "diagnostic.h"
  30 #include "hash.h"
  31 #include "system.h"
  32 #include "uniqstr.h"
  33 #include "thread.h"
  34
  35 /* Use global variable to simplify API.  */
  36 static Hash uniqstrs_table;
  37 static struct mutex uniqstrs_table_mutex;
  38
  39 uniqstr
  40 uniqstr_new_from_buffer (char const *str, size_t size)
  41 {
  42   uniqstr res;
  43   char *r;
  44   uint32_t size32;
  45
  46   if (UNLIKELY (size > UINT32_MAX))
  47     {
  48       warning (_("string is too long"));
  49       abort ();
  50     }
  51
  52   size32 = (uint32_t) size;
  53   mutex_acquire (&uniqstrs_table_mutex);
  54   res = HashFind (&uniqstrs_table, str, size32);
  55   mutex_release (&uniqstrs_table_mutex);
  56
  57   if (res == NULL)
  58     {
  59       /* First insertion in the hash.  */
  60       r = malloc (sizeof (uint32_t) + size32 + 1);
  61
  62       if (UNLIKELY (r == NULL))
  63         goto lfail;
  64
  65       *(uint32_t *) r = size32;
  66       r += sizeof (uint32_t);
  67       memcpy (r, str, size32);
  68       r[size32] = '\0';
  69       mutex_acquire (&uniqstrs_table_mutex);
  70
  71       if (UNLIKELY (HashInsert (&uniqstrs_table, r, size32, r) == NULL))
  72         goto lfail;
  73
  74       mutex_release (&uniqstrs_table_mutex);
  75       res = r;
  76     }
  77
  78   return res;
  79
  80 lfail:
  81   warning (_("memory exhausted"));
  82   abort ();
  83 }
  84
  85 uniqstr
  86 uniqstr_new (const char *str)
  87 {
  88   return uniqstr_new_from_buffer (str, strlen (str));
  89 }
  90
  91 uint32_t
  92 uniqstr_len (uniqstr str)
  93 {
  94   return *(const uint32_t *) (str - sizeof (uint32_t));
  95 }
  96
  97 bool
  98 uniqstr_cmp (uniqstr str1, uniqstr str2)
  99 {
 100   return str1 == str2;
 101 }
 102
 103 void *
 104 uniqstr_chr (uniqstr str, char c)
 105 {
 106   return memchr (str, c, uniqstr_len (str));
 107 }
 108
 109 void
 110 uniqstrs_table_create (void)
 111 {
 112   if (UNLIKELY (!mutex_create (&uniqstrs_table_mutex)))
 113     abort ();
 114
 115   /* Can not provide uniqstr_hash() since HashFind can be called
 116      with non-uniqstr pointer.  */
 117   HashInit (&uniqstrs_table);
 118 }
 119
 120 static void
 121 uniqstr_destroy (void *p)
 122 {
 123   free ((char *) p - sizeof (uint32_t));
 124 }
 125
 126 void
 127 uniqstrs_table_destroy (void)
 128 {
 129   HashClearWithDestructor (&uniqstrs_table, uniqstr_destroy);
 130   mutex_destroy (&uniqstrs_table_mutex);
 131 }