iconv/strtab.c

   1 /* C string table handling.
   2    Copyright (C) 2000, 2001 Free Software Foundation, Inc.
   3    Written by Ulrich Drepper <drepper@redhat.com>, 2000.
   4
   5    This program is free software; you can redistribute it and/or modify
   6    it under the terms of the GNU General Public License as published by
   7    the Free Software Foundation; either version 2, or (at your option)
   8    any later version.
   9
  10    This program 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
  13    GNU General Public License for more details.
  14
  15    You should have received a copy of the GNU General Public License
  16    along with this program; if not, write to the Free Software Foundation,
  17    Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
  18
  19 #ifdef HAVE_CONFIG_H
  20 # include <config.h>
  21 #endif
  22
  23 #include <assert.h>
  24 #include <inttypes.h>
  25 #include <stddef.h>
  26 #include <stdlib.h>
  27 #include <string.h>
  28 #include <unistd.h>
  29 #include <sys/cdefs.h>
  30 #include <sys/param.h>
  31
  32
  33 struct Strent
  34 {
  35   const char *string;
  36   size_t len;
  37   struct Strent *next;
  38   struct Strent *left;
  39   struct Strent *right;
  40   size_t offset;
  41   char reverse[0];
  42 };
  43
  44
  45 struct memoryblock
  46 {
  47   struct memoryblock *next;
  48   char memory[0];
  49 };
  50
  51
  52 struct Strtab
  53 {
  54   struct Strent *root;
  55   struct memoryblock *memory;
  56   char *backp;
  57   size_t left;
  58   size_t total;
  59
  60   struct Strent null;
  61 };
  62
  63
  64 /* Cache for the pagesize.  We correct this value a bit so that `malloc'
  65    is not allocating more than a page.  */
  66 static size_t ps;
  67
  68
  69 extern void *xmalloc (size_t n) __attribute_malloc__;
  70
  71 /* Prototypes for our functions that are used from iconvconfig.c.  If
  72    you change these, change also iconvconfig.c.  */
  73 /* Create new C string table object in memory.  */
  74 extern struct Strtab *strtabinit (void);
  75
  76 /* Free resources allocated for C string table ST.  */
  77 extern void strtabfree (struct Strtab *st);
  78
  79 /* Add string STR (length LEN is != 0) to C string table ST.  */
  80 extern struct Strent *strtabadd (struct Strtab *st, const char *str,
  81                                  size_t len);
  82
  83 /* Finalize string table ST and store size in *SIZE and return a pointer.  */
  84 extern void *strtabfinalize (struct Strtab *st, size_t *size);
  85
  86 /* Get offset in string table for string associated with SE.  */
  87 extern size_t strtaboffset (struct Strent *se);
  88
  89
  90 struct Strtab *
  91 strtabinit (void)
  92 {
  93   if (ps == 0)
  94     {
  95       ps = sysconf (_SC_PAGESIZE) - 2 * sizeof (void *);
  96       assert (sizeof (struct memoryblock) < ps);
  97     }
  98
  99   return (struct Strtab *) calloc (1, sizeof (struct Strtab));
 100 }
 101
 102
 103 static void
 104 morememory (struct Strtab *st, size_t len)
 105 {
 106   struct memoryblock *newmem;
 107
 108   if (len < ps)
 109     len = ps;
 110   newmem = (struct memoryblock *) malloc (len);
 111   if (newmem == NULL)
 112     abort ();
 113
 114   newmem->next = st->memory;
 115   st->memory = newmem;
 116   st->backp = newmem->memory;
 117   st->left = len - offsetof (struct memoryblock, memory);
 118 }
 119
 120
 121 void
 122 strtabfree (struct Strtab *st)
 123 {
 124   struct memoryblock *mb = st->memory;
 125
 126   while (mb != NULL)
 127     {
 128       void *old = mb;
 129       mb = mb->next;
 130       free (old);
 131     }
 132
 133   free (st);
 134 }
 135
 136
 137 static struct Strent *
 138 newstring (struct Strtab *st, const char *str, size_t len)
 139 {
 140   struct Strent *newstr;
 141   size_t align;
 142   int i;
 143
 144   /* Compute the string length if the caller doesn't know it.  */
 145   if (len == 0)
 146     len = strlen (str) + 1;
 147
 148   /* Compute the amount of padding needed to make the structure aligned.  */
 149   align = ((__alignof__ (struct Strent)
 150             - (((uintptr_t) st->backp)
 151                & (__alignof__ (struct Strent) - 1)))
 152            & (__alignof__ (struct Strent) - 1));
 153
 154   /* Make sure there is enough room in the memory block.  */
 155   if (st->left < align + sizeof (struct Strent) + len)
 156     {
 157       morememory (st, sizeof (struct Strent) + len);
 158       align = 0;
 159     }
 160
 161   /* Create the reserved string.  */
 162   newstr = (struct Strent *) (st->backp + align);
 163   newstr->string = str;
 164   newstr->len = len;
 165   newstr->next = NULL;
 166   newstr->left = NULL;
 167   newstr->right = NULL;
 168   newstr->offset = 0;
 169   for (i = len - 2; i >= 0; --i)
 170     newstr->reverse[i] = str[len - 2 - i];
 171   newstr->reverse[len - 1] = '\0';
 172   st->backp += align + sizeof (struct Strent) + len;
 173   st->left -= align + sizeof (struct Strent) + len;
 174
 175   return newstr;
 176 }
 177
 178
 179 /* XXX This function should definitely be rewritten to use a balancing
 180    tree algorith (AVL, red-black trees).  For now a simple, correct
 181    implementation is enough.  */
 182 static struct Strent **
 183 searchstring (struct Strent **sep, struct Strent *newstr)
 184 {
 185   int cmpres;
 186
 187   /* More strings?  */
 188   if (*sep == NULL)
 189     {
 190       *sep = newstr;
 191       return sep;
 192     }
 193
 194   /* Compare the strings.  */
 195   cmpres = memcmp ((*sep)->reverse, newstr->reverse,
 196                    MIN ((*sep)->len, newstr->len));
 197   if (cmpres == 0)
 198     /* We found a matching string.  */
 199     return sep;
 200   else if (cmpres > 0)
 201     return searchstring (&(*sep)->left, newstr);
 202   else
 203     return searchstring (&(*sep)->right, newstr);
 204 }
 205
 206
 207 /* Add new string.  The actual string is assumed to be permanent.  */
 208 struct Strent *
 209 strtabadd (struct Strtab *st, const char *str, size_t len)
 210 {
 211   struct Strent *newstr;
 212   struct Strent **sep;
 213
 214   /* Allocate memory for the new string and its associated information.  */
 215   newstr = newstring (st, str, len);
 216
 217   /* Search in the array for the place to insert the string.  If there
 218      is no string with matching prefix and no string with matching
 219      leading substring, create a new entry.  */
 220   sep = searchstring (&st->root, newstr);
 221   if (*sep != newstr)
 222     {
 223       /* This is not the same entry.  This means we have a prefix match.  */
 224       if ((*sep)->len > newstr->len)
 225         {
 226           /* We have a new substring.  This means we don't need the reverse
 227              string of this entry anymore.  */
 228           st->backp -= newstr->len;
 229           st->left += newstr->len;
 230
 231           newstr->next = (*sep)->next;
 232           (*sep)->next = newstr;
 233         }
 234       else if ((*sep)->len != newstr->len)
 235         {
 236           /* When we get here it means that the string we are about to
 237              add has a common prefix with a string we already have but
 238              it is longer.  In this case we have to put it first.  */
 239           newstr->next = *sep;
 240           *sep = newstr;
 241
 242           st->total += newstr->len - (*sep)->len;
 243         }
 244       else
 245         {
 246           /* We have an exact match.  Free the memory we allocated.  */
 247           st->left += st->backp - (char *) newstr;
 248           st->backp = (char *) newstr;
 249
 250           newstr = *sep;
 251         }
 252     }
 253   else
 254     st->total += newstr->len;
 255
 256   return newstr;
 257 }
 258
 259
 260 static void
 261 copystrings (struct Strent *nodep, char **freep, size_t *offsetp)
 262 {
 263   struct Strent *subs;
 264
 265   if (nodep->left != NULL)
 266     copystrings (nodep->left, freep, offsetp);
 267
 268   /* Process the current node.  */
 269   nodep->offset = *offsetp;
 270   *freep = (char *) mempcpy (*freep, nodep->string, nodep->len);
 271   *offsetp += nodep->len;
 272
 273   for (subs = nodep->next; subs != NULL; subs = subs->next)
 274     {
 275       assert (subs->len < nodep->len);
 276       subs->offset = nodep->offset + nodep->len - subs->len;
 277     }
 278
 279   if (nodep->right != NULL)
 280     copystrings (nodep->right, freep, offsetp);
 281 }
 282
 283
 284 void *
 285 strtabfinalize (struct Strtab *st, size_t *size)
 286 {
 287   size_t copylen;
 288   char *endp;
 289   char *retval;
 290
 291   /* Fill in the information.  */
 292   endp = retval = (char *) xmalloc (st->total + 1);
 293
 294   /* Always put an empty string at the beginning so that a zero offset
 295      can mean error.  */
 296   *endp++ = '\0';
 297
 298   /* Now run through the tree and add all the string while also updating
 299      the offset members of the elfstrent records.  */
 300   copylen = 1;
 301   copystrings (st->root, &endp, &copylen);
 302   assert (copylen == st->total + 1);
 303   assert (endp = retval + st->total + 1);
 304   *size = copylen;
 305
 306   return retval;
 307 }
 308
 309
 310 size_t
 311 strtaboffset (struct Strent *se)
 312 {
 313   return se->offset;
 314 }