ltable.c

   1 /*
   2 ** $Id: ltable.c,v 2.32.1.2 2007/12/28 15:32:23 roberto Exp $
   3 ** Lua tables (hash)
   4 ** See Copyright Notice in lua.h
   5 */
   6
   7
   8 /*
   9 ** Implementation of tables (aka arrays, objects, or hash tables).
  10 ** Tables keep its elements in two parts: an array part and a hash part.
  11 ** Non-negative integer keys are all candidates to be kept in the array
  12 ** part. The actual size of the array is the largest `n' such that at
  13 ** least half the slots between 0 and n are in use.
  14 ** Hash uses a mix of chained scatter table with Brent's variation.
  15 ** A main invariant of these tables is that, if an element is not
  16 ** in its main position (i.e. the `original' position that its hash gives
  17 ** to it), then the colliding element is in its own main position.
  18 ** Hence even when the load factor reaches 100%, performance remains good.
  19 */
  20
  21 #if 0
  22 #include <math.h>
  23 #include <string.h>
  24 #endif
  25
  26 #define ltable_c
  27 #define LUA_CORE
  28
  29 #include "lua.h"
  30
  31 #include "ldebug.h"
  32 #include "ldo.h"
  33 #include "lgc.h"
  34 #include "lmem.h"
  35 #include "lobject.h"
  36 #include "lstate.h"
  37 #include "ltable.h"
  38
  39
  40 /*
  41 ** max size of array part is 2^MAXBITS
  42 */
  43 #if LUAI_BITSINT > 26
  44 #define MAXBITS         26
  45 #else
  46 #define MAXBITS         (LUAI_BITSINT-2)
  47 #endif
  48
  49 #define MAXASIZE        (1 << MAXBITS)
  50
  51
  52 #define hashpow2(t,n)      (gnode(t, lmod((n), sizenode(t))))
  53
  54 #define hashstr(t,str)  hashpow2(t, (str)->tsv.hash)
  55 #define hashboolean(t,p)        hashpow2(t, p)
  56
  57
  58 /*
  59 ** for some types, it is better to avoid modulus by power of 2, as
  60 ** they tend to have many 2 factors.
  61 */
  62 #define hashmod(t,n)    (gnode(t, ((n) % ((sizenode(t)-1)|1))))
  63
  64
  65 #define hashpointer(t,p)        hashmod(t, IntPoint(p))
  66
  67
  68 /*
  69 ** number of ints inside a lua_Number
  70 */
  71 #define numints         cast_int(sizeof(lua_Number)/sizeof(int))
  72
  73
  74
  75 #define dummynode               (&dummynode_)
  76
  77 static const Node dummynode_ = {
  78   {{NULL}, LUA_TNIL},  /* value */
  79   {{{NULL}, LUA_TNIL, NULL}}  /* key */
  80 };
  81
  82
  83 /*
  84 ** hash for lua_Numbers
  85 */
  86 static Node *hashnum (const Table *t, lua_Number n) {
  87   unsigned int a[numints];
  88   int i;
  89   if (luai_numeq(n, 0))  /* avoid problems with -0 */
  90     return gnode(t, 0);
  91   memcpy(a, &n, sizeof(a));
  92   for (i = 1; i < numints; i++) a[0] += a[i];
  93   return hashmod(t, a[0]);
  94 }
  95
  96
  97
  98 /*
  99 ** returns the `main' position of an element in a table (that is, the index
 100 ** of its hash value)
 101 */
 102 static Node *mainposition (const Table *t, const TValue *key) {
 103   switch (ttype(key)) {
 104     case LUA_TNUMBER:
 105       return hashnum(t, nvalue(key));
 106     case LUA_TSTRING:
 107       return hashstr(t, rawtsvalue(key));
 108     case LUA_TBOOLEAN:
 109       return hashboolean(t, bvalue(key));
 110     case LUA_TLIGHTUSERDATA:
 111       return hashpointer(t, pvalue(key));
 112     default:
 113       return hashpointer(t, gcvalue(key));
 114   }
 115 }
 116
 117
 118 /*
 119 ** returns the index for `key' if `key' is an appropriate key to live in
 120 ** the array part of the table, -1 otherwise.
 121 */
 122 static int arrayindex (const TValue *key) {
 123   if (ttisnumber(key)) {
 124     lua_Number n = nvalue(key);
 125     int k;
 126     lua_number2int(k, n);
 127     if (luai_numeq(cast_num(k), n))
 128       return k;
 129   }
 130   return -1;  /* `key' did not match some condition */
 131 }
 132
 133
 134 /*
 135 ** returns the index of a `key' for table traversals. First goes all
 136 ** elements in the array part, then elements in the hash part. The
 137 ** beginning of a traversal is signalled by -1.
 138 */
 139 static int findindex (lua_State *L, Table *t, StkId key) {
 140   int i;
 141   if (ttisnil(key)) return -1;  /* first iteration */
 142   i = arrayindex(key);
 143   if (0 < i && i <= t->sizearray)  /* is `key' inside array part? */
 144     return i-1;  /* yes; that's the index (corrected to C) */
 145   else {
 146     Node *n = mainposition(t, key);
 147     do {  /* check whether `key' is somewhere in the chain */
 148       /* key may be dead already, but it is ok to use it in `next' */
 149       if (luaO_rawequalObj(key2tval(n), key) ||
 150             (ttype(gkey(n)) == LUA_TDEADKEY && iscollectable(key) &&
 151              gcvalue(gkey(n)) == gcvalue(key))) {
 152         i = cast_int(n - gnode(t, 0));  /* key index in hash table */
 153         /* hash elements are numbered after array ones */
 154         return i + t->sizearray;
 155       }
 156       else n = gnext(n);
 157     } while (n);
 158     luaG_runerror(L, "invalid key to " LUA_QL("next"));  /* key not found */
 159     return 0;  /* to avoid warnings */
 160   }
 161 }
 162
 163
 164 int luaH_next (lua_State *L, Table *t, StkId key) {
 165   int i = findindex(L, t, key);  /* find original element */
 166   for (i++; i < t->sizearray; i++) {  /* try first array part */
 167     if (!ttisnil(&t->array[i])) {  /* a non-nil value? */
 168       setnvalue(key, cast_num(i+1));
 169       setobj2s(L, key+1, &t->array[i]);
 170       return 1;
 171     }
 172   }
 173   for (i -= t->sizearray; i < sizenode(t); i++) {  /* then hash part */
 174     if (!ttisnil(gval(gnode(t, i)))) {  /* a non-nil value? */
 175       setobj2s(L, key, key2tval(gnode(t, i)));
 176       setobj2s(L, key+1, gval(gnode(t, i)));
 177       return 1;
 178     }
 179   }
 180   return 0;  /* no more elements */
 181 }
 182
 183
 184 /*
 185 ** {=============================================================
 186 ** Rehash
 187 ** ==============================================================
 188 */
 189
 190
 191 static int computesizes (int nums[], int *narray) {
 192   int i;
 193   int twotoi;  /* 2^i */
 194   int a = 0;  /* number of elements smaller than 2^i */
 195   int na = 0;  /* number of elements to go to array part */
 196   int n = 0;  /* optimal size for array part */
 197   for (i = 0, twotoi = 1; twotoi/2 < *narray; i++, twotoi *= 2) {
 198     if (nums[i] > 0) {
 199       a += nums[i];
 200       if (a > twotoi/2) {  /* more than half elements present? */
 201         n = twotoi;  /* optimal size (till now) */
 202         na = a;  /* all elements smaller than n will go to array part */
 203       }
 204     }
 205     if (a == *narray) break;  /* all elements already counted */
 206   }
 207   *narray = n;
 208   lua_assert(*narray/2 <= na && na <= *narray);
 209   return na;
 210 }
 211
 212
 213 static int countint (const TValue *key, int *nums) {
 214   int k = arrayindex(key);
 215   if (0 < k && k <= MAXASIZE) {  /* is `key' an appropriate array index? */
 216     nums[ceillog2(k)]++;  /* count as such */
 217     return 1;
 218   }
 219   else
 220     return 0;
 221 }
 222
 223
 224 static int numusearray (const Table *t, int *nums) {
 225   int lg;
 226   int ttlg;  /* 2^lg */
 227   int ause = 0;  /* summation of `nums' */
 228   int i = 1;  /* count to traverse all array keys */
 229   for (lg=0, ttlg=1; lg<=MAXBITS; lg++, ttlg*=2) {  /* for each slice */
 230     int lc = 0;  /* counter */
 231     int lim = ttlg;
 232     if (lim > t->sizearray) {
 233       lim = t->sizearray;  /* adjust upper limit */
 234       if (i > lim)
 235         break;  /* no more elements to count */
 236     }
 237     /* count elements in range (2^(lg-1), 2^lg] */
 238     for (; i <= lim; i++) {
 239       if (!ttisnil(&t->array[i-1]))
 240         lc++;
 241     }
 242     nums[lg] += lc;
 243     ause += lc;
 244   }
 245   return ause;
 246 }
 247
 248
 249 static int numusehash (const Table *t, int *nums, int *pnasize) {
 250   int totaluse = 0;  /* total number of elements */
 251   int ause = 0;  /* summation of `nums' */
 252   int i = sizenode(t);
 253   while (i--) {
 254     Node *n = &t->node[i];
 255     if (!ttisnil(gval(n))) {
 256       ause += countint(key2tval(n), nums);
 257       totaluse++;
 258     }
 259   }
 260   *pnasize += ause;
 261   return totaluse;
 262 }
 263
 264
 265 static void setarrayvector (lua_State *L, Table *t, int size) {
 266   int i;
 267   luaM_reallocvector(L, t->array, t->sizearray, size, TValue);
 268   for (i=t->sizearray; i<size; i++)
 269      setnilvalue(&t->array[i]);
 270   t->sizearray = size;
 271 }
 272
 273
 274 static void setnodevector (lua_State *L, Table *t, int size) {
 275   int lsize;
 276   if (size == 0) {  /* no elements to hash part? */
 277     t->node = cast(Node *, dummynode);  /* use common `dummynode' */
 278     lsize = 0;
 279   }
 280   else {
 281     int i;
 282     lsize = ceillog2(size);
 283     if (lsize > MAXBITS)
 284       luaG_runerror(L, "table overflow");
 285     size = twoto(lsize);
 286     t->node = luaM_newvector(L, size, Node);
 287     for (i=0; i<size; i++) {
 288       Node *n = gnode(t, i);
 289       gnext(n) = NULL;
 290       setnilvalue(gkey(n));
 291       setnilvalue(gval(n));
 292     }
 293   }
 294   t->lsizenode = cast_byte(lsize);
 295   t->lastfree = gnode(t, size);  /* all positions are free */
 296 }
 297
 298
 299 static void resize (lua_State *L, Table *t, int nasize, int nhsize) {
 300   int i;
 301   int oldasize = t->sizearray;
 302   int oldhsize = t->lsizenode;
 303   Node *nold = t->node;  /* save old hash ... */
 304   if (nasize > oldasize)  /* array part must grow? */
 305     setarrayvector(L, t, nasize);
 306   /* create new hash part with appropriate size */
 307   setnodevector(L, t, nhsize);
 308   if (nasize < oldasize) {  /* array part must shrink? */
 309     t->sizearray = nasize;
 310     /* re-insert elements from vanishing slice */
 311     for (i=nasize; i<oldasize; i++) {
 312       if (!ttisnil(&t->array[i]))
 313         setobjt2t(L, luaH_setnum(L, t, i+1), &t->array[i]);
 314     }
 315     /* shrink array */
 316     luaM_reallocvector(L, t->array, oldasize, nasize, TValue);
 317   }
 318   /* re-insert elements from hash part */
 319   for (i = twoto(oldhsize) - 1; i >= 0; i--) {
 320     Node *old = nold+i;
 321     if (!ttisnil(gval(old)))
 322       setobjt2t(L, luaH_set(L, t, key2tval(old)), gval(old));
 323   }
 324   if (nold != dummynode)
 325     luaM_freearray(L, nold, twoto(oldhsize), Node);  /* free old array */
 326 }
 327
 328
 329 void luaH_resizearray (lua_State *L, Table *t, int nasize) {
 330   int nsize = (t->node == dummynode) ? 0 : sizenode(t);
 331   resize(L, t, nasize, nsize);
 332 }
 333
 334
 335 static void rehash (lua_State *L, Table *t, const TValue *ek) {
 336   int nasize, na;
 337   int nums[MAXBITS+1];  /* nums[i] = number of keys between 2^(i-1) and 2^i */
 338   int i;
 339   int totaluse;
 340   for (i=0; i<=MAXBITS; i++) nums[i] = 0;  /* reset counts */
 341   nasize = numusearray(t, nums);  /* count keys in array part */
 342   totaluse = nasize;  /* all those keys are integer keys */
 343   totaluse += numusehash(t, nums, &nasize);  /* count keys in hash part */
 344   /* count extra key */
 345   nasize += countint(ek, nums);
 346   totaluse++;
 347   /* compute new size for array part */
 348   na = computesizes(nums, &nasize);
 349   /* resize the table to new computed sizes */
 350   resize(L, t, nasize, totaluse - na);
 351 }
 352
 353
 354
 355 /*
 356 ** }=============================================================
 357 */
 358
 359
 360 Table *luaH_new (lua_State *L, int narray, int nhash) {
 361   Table *t = luaM_new(L, Table);
 362   luaC_link(L, obj2gco(t), LUA_TTABLE);
 363   t->metatable = NULL;
 364   t->flags = cast_byte(~0);
 365   /* temporary values (kept only if some malloc fails) */
 366   t->array = NULL;
 367   t->sizearray = 0;
 368   t->lsizenode = 0;
 369   t->node = cast(Node *, dummynode);
 370   setarrayvector(L, t, narray);
 371   setnodevector(L, t, nhash);
 372   return t;
 373 }
 374
 375
 376 void luaH_free (lua_State *L, Table *t) {
 377   if (t->node != dummynode)
 378     luaM_freearray(L, t->node, sizenode(t), Node);
 379   luaM_freearray(L, t->array, t->sizearray, TValue);
 380   luaM_free(L, t);
 381 }
 382
 383
 384 static Node *getfreepos (Table *t) {
 385   while (t->lastfree-- > t->node) {
 386     if (ttisnil(gkey(t->lastfree)))
 387       return t->lastfree;
 388   }
 389   return NULL;  /* could not find a free place */
 390 }
 391
 392
 393
 394 /*
 395 ** inserts a new key into a hash table; first, check whether key's main
 396 ** position is free. If not, check whether colliding node is in its main
 397 ** position or not: if it is not, move colliding node to an empty place and
 398 ** put new key in its main position; otherwise (colliding node is in its main
 399 ** position), new key goes to an empty position.
 400 */
 401 static TValue *newkey (lua_State *L, Table *t, const TValue *key) {
 402   Node *mp = mainposition(t, key);
 403   if (!ttisnil(gval(mp)) || mp == dummynode) {
 404     Node *othern;
 405     Node *n = getfreepos(t);  /* get a free place */
 406     if (n == NULL) {  /* cannot find a free place? */
 407       rehash(L, t, key);  /* grow table */
 408       return luaH_set(L, t, key);  /* re-insert key into grown table */
 409     }
 410     lua_assert(n != dummynode);
 411     othern = mainposition(t, key2tval(mp));
 412     if (othern != mp) {  /* is colliding node out of its main position? */
 413       /* yes; move colliding node into free position */
 414       while (gnext(othern) != mp) othern = gnext(othern);  /* find previous */
 415       gnext(othern) = n;  /* redo the chain with `n' in place of `mp' */
 416 #if 0
 417       *n = *mp;  /* copy colliding node into free pos. (mp->next also goes) */
 418 #else
 419       memcpy (n, mp, sizeof (*n));
 420 #endif
 421       gnext(mp) = NULL;  /* now `mp' is free */
 422       setnilvalue(gval(mp));
 423     }
 424     else {  /* colliding node is in its own main position */
 425       /* new node will go into free position */
 426       gnext(n) = gnext(mp);  /* chain new position */
 427       gnext(mp) = n;
 428       mp = n;
 429     }
 430   }
 431   gkey(mp)->value = key->value; gkey(mp)->tt = key->tt;
 432   luaC_barriert(L, t, key);
 433   lua_assert(ttisnil(gval(mp)));
 434   return gval(mp);
 435 }
 436
 437
 438 /*
 439 ** search function for integers
 440 */
 441 const TValue *luaH_getnum (Table *t, int key) {
 442   /* (1 <= key && key <= t->sizearray) */
 443   if (cast(unsigned int, key-1) < cast(unsigned int, t->sizearray))
 444     return &t->array[key-1];
 445   else {
 446     lua_Number nk = cast_num(key);
 447     Node *n = hashnum(t, nk);
 448     do {  /* check whether `key' is somewhere in the chain */
 449       if (ttisnumber(gkey(n)) && luai_numeq(nvalue(gkey(n)), nk))
 450         return gval(n);  /* that's it */
 451       else n = gnext(n);
 452     } while (n);
 453     return luaO_nilobject;
 454   }
 455 }
 456
 457
 458 /*
 459 ** search function for strings
 460 */
 461 const TValue *luaH_getstr (Table *t, TString *key) {
 462   Node *n = hashstr(t, key);
 463   do {  /* check whether `key' is somewhere in the chain */
 464     if (ttisstring(gkey(n)) && rawtsvalue(gkey(n)) == key)
 465       return gval(n);  /* that's it */
 466     else n = gnext(n);
 467   } while (n);
 468   return luaO_nilobject;
 469 }
 470
 471
 472 /*
 473 ** main search function
 474 */
 475 const TValue *luaH_get (Table *t, const TValue *key) {
 476   switch (ttype(key)) {
 477     case LUA_TNIL: return luaO_nilobject;
 478     case LUA_TSTRING: return luaH_getstr(t, rawtsvalue(key));
 479     case LUA_TNUMBER: {
 480       int k;
 481       lua_Number n = nvalue(key);
 482       lua_number2int(k, n);
 483       if (luai_numeq(cast_num(k), nvalue(key))) /* index is int? */
 484         return luaH_getnum(t, k);  /* use specialized version */
 485       /* else go through */
 486     }
 487     default: {
 488       Node *n = mainposition(t, key);
 489       do {  /* check whether `key' is somewhere in the chain */
 490         if (luaO_rawequalObj(key2tval(n), key))
 491           return gval(n);  /* that's it */
 492         else n = gnext(n);
 493       } while (n);
 494       return luaO_nilobject;
 495     }
 496   }
 497 }
 498
 499
 500 TValue *luaH_set (lua_State *L, Table *t, const TValue *key) {
 501   const TValue *p = luaH_get(t, key);
 502   t->flags = 0;
 503   if (p != luaO_nilobject)
 504     return cast(TValue *, p);
 505   else {
 506     if (ttisnil(key)) luaG_runerror(L, "table index is nil");
 507     else if (ttisnumber(key) && luai_numisnan(nvalue(key)))
 508       luaG_runerror(L, "table index is NaN");
 509     return newkey(L, t, key);
 510   }
 511 }
 512
 513
 514 TValue *luaH_setnum (lua_State *L, Table *t, int key) {
 515   const TValue *p = luaH_getnum(t, key);
 516   if (p != luaO_nilobject)
 517     return cast(TValue *, p);
 518   else {
 519     TValue k;
 520     setnvalue(&k, cast_num(key));
 521     return newkey(L, t, &k);
 522   }
 523 }
 524
 525
 526 TValue *luaH_setstr (lua_State *L, Table *t, TString *key) {
 527   const TValue *p = luaH_getstr(t, key);
 528   if (p != luaO_nilobject)
 529     return cast(TValue *, p);
 530   else {
 531     TValue k;
 532     setsvalue(L, &k, key);
 533     return newkey(L, t, &k);
 534   }
 535 }
 536
 537
 538 static int unbound_search (Table *t, unsigned int j) {
 539   unsigned int i = j;  /* i is zero or a present index */
 540   j++;
 541   /* find `i' and `j' such that i is present and j is not */
 542   while (!ttisnil(luaH_getnum(t, j))) {
 543     i = j;
 544     j *= 2;
 545     if (j > cast(unsigned int, MAX_INT)) {  /* overflow? */
 546       /* table was built with bad purposes: resort to linear search */
 547       i = 1;
 548       while (!ttisnil(luaH_getnum(t, i))) i++;
 549       return i - 1;
 550     }
 551   }
 552   /* now do a binary search between them */
 553   while (j - i > 1) {
 554     unsigned int m = (i+j)/2;
 555     if (ttisnil(luaH_getnum(t, m))) j = m;
 556     else i = m;
 557   }
 558   return i;
 559 }
 560
 561
 562 /*
 563 ** Try to find a boundary in table `t'. A `boundary' is an integer index
 564 ** such that t[i] is non-nil and t[i+1] is nil (and 0 if t[1] is nil).
 565 */
 566 int luaH_getn (Table *t) {
 567   unsigned int j = t->sizearray;
 568   if (j > 0 && ttisnil(&t->array[j - 1])) {
 569     /* there is a boundary in the array part: (binary) search for it */
 570     unsigned int i = 0;
 571     while (j - i > 1) {
 572       unsigned int m = (i+j)/2;
 573       if (ttisnil(&t->array[m - 1])) j = m;
 574       else i = m;
 575     }
 576     return i;
 577   }
 578   /* else must find a boundary in hash part */
 579   else if (t->node == dummynode)  /* hash part is empty? */
 580     return j;  /* that is easy... */
 581   else return unbound_search(t, j);
 582 }
 583
 584
 585
 586 #if defined(LUA_DEBUG)
 587
 588 Node *luaH_mainposition (const Table *t, const TValue *key) {
 589   return mainposition(t, key);
 590 }
 591
 592 int luaH_isdummy (Node *n) { return n == dummynode; }
 593
 594 #endif