locale/weight.h

   1 /* Copyright (C) 1996, 1997 Free Software Foundation, Inc.
   2    This file is part of the GNU C Library.
   3    Written by Ulrich Drepper, <drepper@gnu.ai.mit.edu>.
   4
   5    The GNU C Library is free software; you can redistribute it and/or
   6    modify it under the terms of the GNU Library General Public License as
   7    published by the Free Software Foundation; either version 2 of the
   8    License, or (at your option) any later version.
   9
  10    The GNU C Library is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13    Library General Public License for more details.
  14
  15    You should have received a copy of the GNU Library General Public
  16    License along with the GNU C Library; see the file COPYING.LIB.  If not,
  17    write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  18    Boston, MA 02111-1307, USA.  */
  19
  20 #include <alloca.h>
  21 #include <langinfo.h>
  22 #include "localeinfo.h"
  23
  24 #ifndef STRING_TYPE
  25 # error STRING_TYPE not defined
  26 #endif
  27
  28 #ifndef USTRING_TYPE
  29 # error USTRING_TYPE not defined
  30 #endif
  31
  32 typedef struct weight_t
  33 {
  34   struct weight_t *prev;
  35   struct weight_t *next;
  36   struct data_pair
  37     {
  38       int number;
  39       const u_int32_t *value;
  40     } data[0];
  41 } weight_t;
  42
  43
  44 /* The following five macros grant access to the values in the
  45    collate locale file that do not depend on byte order.  */
  46 #ifndef USE_IN_EXTENDED_LOCALE_MODEL
  47 # define collate_nrules \
  48   (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES))
  49 # define collate_hash_size \
  50   (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_HASH_SIZE))
  51 # define collate_hash_layers \
  52   (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_HASH_LAYERS))
  53 # define collate_undefined \
  54   (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_UNDEFINED))
  55 # define collate_rules \
  56   ((u_int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_RULES))
  57
  58 static __inline void get_weight (const STRING_TYPE **str, weight_t *result);
  59 static __inline void
  60 get_weight (const STRING_TYPE **str, weight_t *result)
  61 #else
  62 # define collate_nrules \
  63   current->values[_NL_ITEM_INDEX (_NL_COLLATE_NRULES)].word
  64 # define collate_hash_size \
  65   current->values[_NL_ITEM_INDEX (_NL_COLLATE_HASH_SIZE)].word
  66 # define collate_hash_layers \
  67   current->values[_NL_ITEM_INDEX (_NL_COLLATE_HASH_LAYERS)].word
  68 # define collate_undefined \
  69   current->values[_NL_ITEM_INDEX (_NL_COLLATE_UNDEFINED)].word
  70 # define collate_rules \
  71   ((u_int32_t *) current->values[_NL_ITEM_INDEX (_NL_COLLATE_RULES)].string)
  72
  73 static __inline void get_weight (const STRING_TYPE **str, weight_t *result,
  74                                  struct locale_data *current,
  75                                  const u_int32_t *__collate_table,
  76                                  const u_int32_t *__collate_extra);
  77 static __inline void
  78 get_weight (const STRING_TYPE **str, weight_t *result,
  79             struct locale_data *current, const u_int32_t *__collate_table,
  80             const u_int32_t *__collate_extra)
  81 #endif
  82 {
  83   unsigned int ch = *((USTRING_TYPE *) (*str))++;
  84   size_t slot;
  85
  86   if (sizeof (STRING_TYPE) == 1)
  87     slot = ch * (collate_nrules + 1);
  88   else
  89     {
  90       const size_t level_size = collate_hash_size * (collate_nrules + 1);
  91       size_t level;
  92
  93       slot = (ch % collate_hash_size) * (collate_nrules + 1);
  94
  95       level = 0;
  96       while (__collate_table[slot] != (u_int32_t) ch)
  97         {
  98           if (__collate_table[slot + 1] == 0
  99               || ++level >= collate_hash_layers)
 100             {
 101               size_t idx = collate_undefined;
 102               size_t cnt;
 103
 104               for (cnt = 0; cnt < collate_nrules; ++cnt)
 105                 {
 106                   result->data[cnt].number = __collate_extra[idx++];
 107                   result->data[cnt].value = &__collate_extra[idx];
 108                   idx += result->data[cnt].number;
 109                 }
 110               return;
 111             }
 112           slot += level_size;
 113         }
 114     }
 115
 116   if (__collate_table[slot + 1] != (u_int32_t) FORWARD_CHAR)
 117     {
 118       /* We have a simple form.  One one value for each weight.  */
 119       size_t cnt;
 120
 121       for (cnt = 0; cnt < collate_nrules; ++cnt)
 122         {
 123           result->data[cnt].number = 1;
 124           result->data[cnt].value = &__collate_table[slot + 1 + cnt];
 125         }
 126       return;
 127     }
 128
 129   /* We now look for any collation element which starts with CH.
 130      There might none, but the last list member is a catch-all case
 131      because it is simple the character CH.  The value of this entry
 132      might be the same as UNDEFINED.  */
 133   slot = __collate_table[slot + 2];
 134
 135   while (1)
 136     {
 137       size_t idx;
 138
 139       /* This is a comparison between a u_int32_t array (aka wchar_t) and
 140          an 8-bit string.  */
 141       for (idx = 0; __collate_extra[slot + 2 + idx] != 0; ++idx)
 142         if (__collate_extra[slot + 2 + idx] != (u_int32_t) (*str)[idx])
 143           break;
 144
 145       /* When the loop finished with all character of the collation
 146          element used, we found the longest prefix.  */
 147       if (__collate_extra[slot + 2 + idx] == 0)
 148         {
 149           size_t cnt;
 150
 151           *str += idx;
 152           idx += slot + 3;
 153           for (cnt = 0; cnt < collate_nrules; ++cnt)
 154             {
 155               result->data[cnt].number = __collate_extra[idx++];
 156               result->data[cnt].value = &__collate_extra[idx];
 157               idx += result->data[cnt].number;
 158             }
 159           return;
 160         }
 161
 162       /* To next entry in list.  */
 163       slot += __collate_extra[slot];
 164     }
 165 }
 166
 167
 168 /* To process a string efficiently we retrieve all information about
 169    the string at once.  The following macro constructs a double linked
 170    list of this information.  It is a macro because we use `alloca'
 171    and we use a double linked list because of the backward collation
 172    order.
 173
 174    We have this strange extra macro since the functions which use the
 175    given locale (not the global one) cannot use the global tables.  */
 176 #ifndef USE_IN_EXTENDED_LOCALE_MODEL
 177 # define call_get_weight(strp, newp) get_weight ((strp), (newp))
 178 #else
 179 # define call_get_weight(strp, newp) \
 180   get_weight ((strp), (newp), current, collate_table, collate_extra)
 181 #endif
 182
 183 #define get_string(str, forw, backw) \
 184   do                                                                          \
 185     {                                                                         \
 186       weight_t *newp;                                                         \
 187       while (*str != '\0')                                                    \
 188         {                                                                     \
 189           newp = (weight_t *) alloca (sizeof (weight_t)                       \
 190                                       + (collate_nrules                       \
 191                                          * sizeof (struct data_pair)));       \
 192                                                                               \
 193           newp->prev = backw;                                                 \
 194           if (backw == NULL)                                                  \
 195             forw = newp;                                                      \
 196           else                                                                \
 197             backw->next = newp;                                               \
 198           newp->next = NULL;                                                  \
 199           backw = newp;                                                       \
 200           call_get_weight (&str, newp);                                       \
 201         }                                                                     \
 202     }                                                                         \
 203   while (0)