Table 的相关结构的定义在 lobject.h 里面。在 Lua 里面 Table 有两个结构,一个是 aray,另一个是 hash。array 是对应正整数作为 key 的情况。

typedef union TKey {
  struct {
    TValuefields;
    int next;  /* for chaining (offset for next node) */
  } nk;
  TValue tvk;
} TKey;


/* copy a value into a key without messing up field 'next' */
#define setnodekey(L,key,obj) \
  { TKey *k_=(key); const TValue *io_=(obj); \
    k_->nk.value_ = io_->value_; k_->nk.tt_ = io_->tt_; \
    (void)L; checkliveness(L,io_); }


typedef struct Node {
  TValue i_val;
  TKey i_key;
} Node;


typedef struct Table {
  CommonHeader;
  lu_byte flags;  /* 1<<p means tagmethod(p) is not present */
  lu_byte lsizenode;  /* log2 of size of 'node' array */
  unsigned int sizearray;  /* size of 'array' array */
  TValue *array;  /* array part */
  Node *node;
  Node *lastfree;  /* any free position is before this position */
  struct Table *metatable;
  GCObject *gclist;
} Table;

ltable.h 里面定义是一些辅助宏。

#ifndef ltable_h
#define ltable_h

#include "lobject.h"


#define gnode(t,i)  (&(t)->node[i])
#define gval(n)   (&(n)->i_val)
#define gnext(n)  ((n)->i_key.nk.next)


/* 'const' to avoid wrong writings that can mess up field 'next' */
#define gkey(n)   cast(const TValue*, (&(n)->i_key.tvk))

/*
** writable version of 'gkey'; allows updates to individual fields,
** but not to the whole (which has incompatible type)
*/
#define wgkey(n)    (&(n)->i_key.nk)

#define invalidateTMcache(t)  ((t)->flags = 0)


/* true when 't' is using 'dummynode' as its hash part */
#define isdummy(t)    ((t)->lastfree == NULL)


/* allocated size for hash nodes */
#define allocsizenode(t)  (isdummy(t) ? 0 : sizenode(t))


/* returns the key, given the value of a table entry */
#define keyfromval(v) \
  (gkey(cast(Node *, cast(char *, (v)) - offsetof(Node, i_val))))


LUAI_FUNC const TValue *luaH_getint (Table *t, lua_Integer key);
LUAI_FUNC void luaH_setint (lua_State *L, Table *t, lua_Integer key,
                                                    TValue *value);
LUAI_FUNC const TValue *luaH_getshortstr (Table *t, TString *key);
LUAI_FUNC const TValue *luaH_getstr (Table *t, TString *key);
LUAI_FUNC const TValue *luaH_get (Table *t, const TValue *key);
LUAI_FUNC TValue *luaH_newkey (lua_State *L, Table *t, const TValue *key);
LUAI_FUNC TValue *luaH_set (lua_State *L, Table *t, const TValue *key);
LUAI_FUNC Table *luaH_new (lua_State *L);
LUAI_FUNC void luaH_resize (lua_State *L, Table *t, unsigned int nasize,
                                                    unsigned int nhsize);
LUAI_FUNC void luaH_resizearray (lua_State *L, Table *t, unsigned int nasize);
LUAI_FUNC void luaH_free (lua_State *L, Table *t);
LUAI_FUNC int luaH_next (lua_State *L, Table *t, StkId key);
LUAI_FUNC int luaH_getn (Table *t);


#if defined(LUA_DEBUG)
LUAI_FUNC Node *luaH_mainposition (const Table *t, const TValue *key);
LUAI_FUNC int luaH_isdummy (const Table *t);
#endif


#endif

ltable.c 的开头定义了一些常量,宏。

/*
** Maximum size of array part (MAXASIZE) is 2^MAXABITS. MAXABITS is
** the largest integer such that MAXASIZE fits in an unsigned int.
*/
#define MAXABITS  cast_int(sizeof(int) * CHAR_BIT - 1)
#define MAXASIZE  (1u << MAXABITS)

/*
** Maximum size of hash part is 2^MAXHBITS. MAXHBITS is the largest
** integer such that 2^MAXHBITS fits in a signed int. (Note that the
** maximum number of elements in a table, 2^MAXABITS + 2^MAXHBITS, still
** fits comfortably in an unsigned int.)
*/
#define MAXHBITS  (MAXABITS - 1)


#define hashpow2(t,n)   (gnode(t, lmod((n), sizenode(t))))

#define hashstr(t,str)    hashpow2(t, (str)->hash)
#define hashboolean(t,p)  hashpow2(t, p)
#define hashint(t,i)    hashpow2(t, i)


/*
** for some types, it is better to avoid modulus by power of 2, as
** they tend to have many 2 factors.
*/
#define hashmod(t,n)  (gnode(t, ((n) % ((sizenode(t)-1)|1))))


#define hashpointer(t,p)  hashmod(t, point2uint(p))


#define dummynode   (&dummynode_)

static const Node dummynode_ = {
  {NILCONSTANT},  /* value */
  {
    {NILCONSTANT, 0}
  }  /* key */
};

我们先来看 luaH_new 方法。

Table *luaH_new (lua_State *L) {
  GCObject *o = luaC_newobj(L, LUA_TTABLE, sizeof(Table));
  Table *t = gco2t(o);
  t->metatable = NULL;
  t->flags = cast_byte(~0);
  t->array = NULL;
  t->sizearray = 0;
  setnodevector(L, t, 0);
  return t;
}

这里设置了一些默认的属性,然后调用 setnodevector

static void setnodevector (lua_State *L, Table *t, unsigned int size) {
  if (size == 0) {  /* no elements to hash part? */
    t->node = cast(Node *, dummynode);  /* use common 'dummynode' */
    t->lsizenode = 0;
    t->lastfree = NULL;  /* signal that it is using dummy node */
  }
  else {
    int i;
    int lsize = luaO_ceillog2(size);
    if (lsize > MAXHBITS)
      luaG_runerror(L, "table overflow");
    size = twoto(lsize);
    t->node = luaM_newvector(L, size, Node);
    for (i = 0; i < (int)size; i++) {
      Node *n = gnode(t, i);
      gnext(n) = 0;
      setnilvalue(wgkey(n));
      setnilvalue(gval(n));
    }
    t->lsizenode = cast_byte(lsize);
    t->lastfree = gnode(t, size);  /* all positions are free */
  }
}

新建的 table 传进来的 size 是 0,那样的话,就是用 dummynode 来初始化 table。别的情况就是在 resize 的时候调用这个方法,如果是此种情况下的话,则会分配了 2 ** luaO_ceillog2(size) 长度的空间来保存 node。

/*
** search function for integers
*/
const TValue *luaH_getint (Table *t, lua_Integer key) {
  /* (1 <= key && key <= t->sizearray) */
  if (l_castS2U(key) - 1 < t->sizearray)
    return &t->array[key - 1];
  else {
    Node *n = hashint(t, key);
    for (;;) {  /* check whether 'key' is somewhere in the chain */
      if (ttisinteger(gkey(n)) && ivalue(gkey(n)) == key)
        return gval(n);  /* that's it */
      else {
        int nx = gnext(n);
        if (nx == 0) break;
        n += nx;
      }
    }
    return luaO_nilobject;
  }
}


/*
** search function for short strings
*/
const TValue *luaH_getshortstr (Table *t, TString *key) {
  Node *n = hashstr(t, key);
  lua_assert(key->tt == LUA_TSHRSTR);
  for (;;) {  /* check whether 'key' is somewhere in the chain */
    const TValue *k = gkey(n);
    if (ttisshrstring(k) && eqshrstr(tsvalue(k), key))
      return gval(n);  /* that's it */
    else {
      int nx = gnext(n);
      if (nx == 0)
        return luaO_nilobject;  /* not found */
      n += nx;
    }
  }
}


/*
** "Generic" get version. (Not that generic: not valid for integers,
** which may be in array part, nor for floats with integral values.)
*/
static const TValue *getgeneric (Table *t, const TValue *key) {
  Node *n = mainposition(t, key);
  for (;;) {  /* check whether 'key' is somewhere in the chain */
    if (luaV_rawequalobj(gkey(n), key))
      return gval(n);  /* that's it */
    else {
      int nx = gnext(n);
      if (nx == 0)
        return luaO_nilobject;  /* not found */
      n += nx;
    }
  }
}


const TValue *luaH_getstr (Table *t, TString *key) {
  if (key->tt == LUA_TSHRSTR)
    return luaH_getshortstr(t, key);
  else {  /* for long strings, use generic case */
    TValue ko;
    setsvalue(cast(lua_State *, NULL), &ko, key);
    return getgeneric(t, &ko);
  }
}


/*
** main search function
*/
const TValue *luaH_get (Table *t, const TValue *key) {
  switch (ttype(key)) {
    case LUA_TSHRSTR: return luaH_getshortstr(t, tsvalue(key));
    case LUA_TNUMINT: return luaH_getint(t, ivalue(key));
    case LUA_TNIL: return luaO_nilobject;
    case LUA_TNUMFLT: {
      lua_Integer k;
      if (luaV_tointeger(key, &k, 0)) /* index is int? */
        return luaH_getint(t, k);  /* use specialized version */
      /* else... */
    }  /* FALLTHROUGH */
    default:
      return getgeneric(t, key);
  }
}

这里的入口应该是 luaH_get,根据 key 来用不同的方法来用不同的方法获取数据。

/*
** returns the index for 'key' if 'key' is an appropriate key to live in
** the array part of the table, 0 otherwise.
*/
static unsigned int arrayindex (const TValue *key) {
  if (ttisinteger(key)) {
    lua_Integer k = ivalue(key);
    if (0 < k && (lua_Unsigned)k <= MAXASIZE)
      return cast(unsigned int, k);  /* 'key' is an appropriate array index */
  }
  return 0;  /* 'key' did not match some condition */
}


/*
** returns the index of a 'key' for table traversals. First goes all
** elements in the array part, then elements in the hash part. The
** beginning of a traversal is signaled by 0.
*/
static unsigned int findindex (lua_State *L, Table *t, StkId key) {
  unsigned int i;
  if (ttisnil(key)) return 0;  /* first iteration */
  i = arrayindex(key);
  if (i != 0 && i <= t->sizearray)  /* is 'key' inside array part? */
    return i;  /* yes; that's the index */
  else {
    int nx;
    Node *n = mainposition(t, key);
    for (;;) {  /* check whether 'key' is somewhere in the chain */
      /* key may be dead already, but it is ok to use it in 'next' */
      if (luaV_rawequalobj(gkey(n), key) ||
            (ttisdeadkey(gkey(n)) && iscollectable(key) &&
             deadvalue(gkey(n)) == gcvalue(key))) {
        i = cast_int(n - gnode(t, 0));  /* key index in hash table */
        /* hash elements are numbered after array ones */
        return (i + 1) + t->sizearray;
      }
      nx = gnext(n);
      if (nx == 0)
        luaG_runerror(L, "invalid key to 'next'");  /* key not found */
      else n += nx;
    }
  }
}


int luaH_next (lua_State *L, Table *t, StkId key) {
  unsigned int i = findindex(L, t, key);  /* find original element */
  for (; i < t->sizearray; i++) {  /* try first array part */
    if (!ttisnil(&t->array[i])) {  /* a non-nil value? */
      setivalue(key, i + 1);
      setobj2s(L, key+1, &t->array[i]);
      return 1;
    }
  }
  for (i -= t->sizearray; cast_int(i) < sizenode(t); i++) {  /* hash part */
    if (!ttisnil(gval(gnode(t, i)))) {  /* a non-nil value? */
      setobj2s(L, key, gkey(gnode(t, i)));
      setobj2s(L, key+1, gval(gnode(t, i)));
      return 1;
    }
  }
  return 0;  /* no more elements */
}

这个应该是迭代的时候回用到。

/*
** {=============================================================
** Rehash
** ==============================================================
*/

/*
** Compute the optimal size for the array part of table 't'. 'nums' is a
** "count array" where 'nums[i]' is the number of integers in the table
** between 2^(i - 1) + 1 and 2^i. 'pna' enters with the total number of
** integer keys in the table and leaves with the number of keys that
** will go to the array part; return the optimal size.
*/
static unsigned int computesizes (unsigned int nums[], unsigned int *pna) {
  int i;
  unsigned int twotoi;  /* 2^i (candidate for optimal size) */
  unsigned int a = 0;  /* number of elements smaller than 2^i */
  unsigned int na = 0;  /* number of elements to go to array part */
  unsigned int optimal = 0;  /* optimal size for array part */
  /* loop while keys can fill more than half of total size */
  for (i = 0, twotoi = 1; *pna > twotoi / 2; i++, twotoi *= 2) {
    if (nums[i] > 0) {
      a += nums[i];
      if (a > twotoi/2) {  /* more than half elements present? */
        optimal = twotoi;  /* optimal size (till now) */
        na = a;  /* all elements up to 'optimal' will go to array part */
      }
    }
  }
  lua_assert((optimal == 0 || optimal / 2 < na) && na <= optimal);
  *pna = na;
  return optimal;
}


static int countint (const TValue *key, unsigned int *nums) {
  unsigned int k = arrayindex(key);
  if (k != 0) {  /* is 'key' an appropriate array index? */
    nums[luaO_ceillog2(k)]++;  /* count as such */
    return 1;
  }
  else
    return 0;
}


/*
** Count keys in array part of table 't': Fill 'nums[i]' with
** number of keys that will go into corresponding slice and return
** total number of non-nil keys.
*/
static unsigned int numusearray (const Table *t, unsigned int *nums) {
  int lg;
  unsigned int ttlg;  /* 2^lg */
  unsigned int ause = 0;  /* summation of 'nums' */
  unsigned int i = 1;  /* count to traverse all array keys */
  /* traverse each slice */
  for (lg = 0, ttlg = 1; lg <= MAXABITS; lg++, ttlg *= 2) {
    unsigned int lc = 0;  /* counter */
    unsigned int lim = ttlg;
    if (lim > t->sizearray) {
      lim = t->sizearray;  /* adjust upper limit */
      if (i > lim)
        break;  /* no more elements to count */
    }
    /* count elements in range (2^(lg - 1), 2^lg] */
    for (; i <= lim; i++) {
      if (!ttisnil(&t->array[i-1]))
        lc++;
    }
    nums[lg] += lc;
    ause += lc;
  }
  return ause;
}


static int numusehash (const Table *t, unsigned int *nums, unsigned int *pna) {
  int totaluse = 0;  /* total number of elements */
  int ause = 0;  /* elements added to 'nums' (can go to array part) */
  int i = sizenode(t);
  while (i--) {
    Node *n = &t->node[i];
    if (!ttisnil(gval(n))) {
      ause += countint(gkey(n), nums);
      totaluse++;
    }
  }
  *pna += ause;
  return totaluse;
}


static void setarrayvector (lua_State *L, Table *t, unsigned int size) {
  unsigned int i;
  luaM_reallocvector(L, t->array, t->sizearray, size, TValue);
  for (i=t->sizearray; i<size; i++)
     setnilvalue(&t->array[i]);
  t->sizearray = size;
}


static void setnodevector (lua_State *L, Table *t, unsigned int size) {
  if (size == 0) {  /* no elements to hash part? */
    t->node = cast(Node *, dummynode);  /* use common 'dummynode' */
    t->lsizenode = 0;
    t->lastfree = NULL;  /* signal that it is using dummy node */
  }
  else {
    int i;
    int lsize = luaO_ceillog2(size);
    if (lsize > MAXHBITS)
      luaG_runerror(L, "table overflow");
    size = twoto(lsize);
    t->node = luaM_newvector(L, size, Node);
    for (i = 0; i < (int)size; i++) {
      Node *n = gnode(t, i);
      gnext(n) = 0;
      setnilvalue(wgkey(n));
      setnilvalue(gval(n));
    }
    t->lsizenode = cast_byte(lsize);
    t->lastfree = gnode(t, size);  /* all positions are free */
  }
}


void luaH_resize (lua_State *L, Table *t, unsigned int nasize,
                                          unsigned int nhsize) {
  unsigned int i;
  int j;
  unsigned int oldasize = t->sizearray;
  int oldhsize = allocsizenode(t);
  Node *nold = t->node;  /* save old hash ... */
  if (nasize > oldasize)  /* array part must grow? */
    setarrayvector(L, t, nasize);
  /* create new hash part with appropriate size */
  setnodevector(L, t, nhsize);
  if (nasize < oldasize) {  /* array part must shrink? */
    t->sizearray = nasize;
    /* re-insert elements from vanishing slice */
    for (i=nasize; i<oldasize; i++) {
      if (!ttisnil(&t->array[i]))
        luaH_setint(L, t, i + 1, &t->array[i]);
    }
    /* shrink array */
    luaM_reallocvector(L, t->array, oldasize, nasize, TValue);
  }
  /* re-insert elements from hash part */
  for (j = oldhsize - 1; j >= 0; j--) {
    Node *old = nold + j;
    if (!ttisnil(gval(old))) {
      /* doesn't need barrier/invalidate cache, as entry was
         already present in the table */
      setobjt2t(L, luaH_set(L, t, gkey(old)), gval(old));
    }
  }
  if (oldhsize > 0)  /* not the dummy node? */
    luaM_freearray(L, nold, cast(size_t, oldhsize)); /* free old hash */
}


void luaH_resizearray (lua_State *L, Table *t, unsigned int nasize) {
  int nsize = allocsizenode(t);
  luaH_resize(L, t, nasize, nsize);
}

/*
** nums[i] = number of keys 'k' where 2^(i - 1) < k <= 2^i
*/
static void rehash (lua_State *L, Table *t, const TValue *ek) {
  unsigned int asize;  /* optimal size for array part */
  unsigned int na;  /* number of keys in the array part */
  unsigned int nums[MAXABITS + 1];
  int i;
  int totaluse;
  for (i = 0; i <= MAXABITS; i++) nums[i] = 0;  /* reset counts */
  na = numusearray(t, nums);  /* count keys in array part */
  totaluse = na;  /* all those keys are integer keys */
  totaluse += numusehash(t, nums, &na);  /* count keys in hash part */
  /* count extra key */
  na += countint(ek, nums);
  totaluse++;
  /* compute new size for array part */
  asize = computesizes(nums, &na);
  /* resize the table to new computed sizes */
  luaH_resize(L, t, asize, totaluse - na);
}



/*
** }=============================================================
*/

这部分考虑的是当大小变化时,需要重新 hash。

void luaH_free (lua_State *L, Table *t) {
  if (!isdummy(t))
    luaM_freearray(L, t->node, cast(size_t, sizenode(t)));
  luaM_freearray(L, t->array, t->sizearray);
  luaM_free(L, t);
}

这个方法用来释放 table 的空间。

static Node *getfreepos (Table *t) {
  if (!isdummy(t)) {
    while (t->lastfree > t->node) {
      t->lastfree--;
      if (ttisnil(gkey(t->lastfree)))
        return t->lastfree;
    }
  }
  return NULL;  /* could not find a free place */
}



/*
** inserts a new key into a hash table; first, check whether key's main
** position is free. If not, check whether colliding node is in its main
** position or not: if it is not, move colliding node to an empty place and
** put new key in its main position; otherwise (colliding node is in its main
** position), new key goes to an empty position.
*/
TValue *luaH_newkey (lua_State *L, Table *t, const TValue *key) {
  Node *mp;
  TValue aux;
  if (ttisnil(key)) luaG_runerror(L, "table index is nil");
  else if (ttisfloat(key)) {
    lua_Integer k;
    if (luaV_tointeger(key, &k, 0)) {  /* does index fit in an integer? */
      setivalue(&aux, k);
      key = &aux;  /* insert it as an integer */
    }
    else if (luai_numisnan(fltvalue(key)))
      luaG_runerror(L, "table index is NaN");
  }
  mp = mainposition(t, key);
  if (!ttisnil(gval(mp)) || isdummy(t)) {  /* main position is taken? */
    Node *othern;
    Node *f = getfreepos(t);  /* get a free place */
    if (f == NULL) {  /* cannot find a free place? */
      rehash(L, t, key);  /* grow table */
      /* whatever called 'newkey' takes care of TM cache */
      return luaH_set(L, t, key);  /* insert key into grown table */
    }
    lua_assert(!isdummy(t));
    othern = mainposition(t, gkey(mp));
    if (othern != mp) {  /* is colliding node out of its main position? */
      /* yes; move colliding node into free position */
      while (othern + gnext(othern) != mp)  /* find previous */
        othern += gnext(othern);
      gnext(othern) = cast_int(f - othern);  /* rechain to point to 'f' */
      *f = *mp;  /* copy colliding node into free pos. (mp->next also goes) */
      if (gnext(mp) != 0) {
        gnext(f) += cast_int(mp - f);  /* correct 'next' */
        gnext(mp) = 0;  /* now 'mp' is free */
      }
      setnilvalue(gval(mp));
    }
    else {  /* colliding node is in its own main position */
      /* new node will go into free position */
      if (gnext(mp) != 0)
        gnext(f) = cast_int((mp + gnext(mp)) - f);  /* chain new position */
      else lua_assert(gnext(f) == 0);
      gnext(mp) = cast_int(f - mp);
      mp = f;
    }
  }
  setnodekey(L, &mp->i_key, key);
  luaC_barrierback(L, t, key);
  lua_assert(ttisnil(gval(mp)));
  return gval(mp);
}

这部分是为了产生新的 key 的方法,主要是发生碰撞的处理。

/*
** beware: when using this function you probably need to check a GC
** barrier and invalidate the TM cache.
*/
TValue *luaH_set (lua_State *L, Table *t, const TValue *key) {
  const TValue *p = luaH_get(t, key);
  if (p != luaO_nilobject)
    return cast(TValue *, p);
  else return luaH_newkey(L, t, key);
}


void luaH_setint (lua_State *L, Table *t, lua_Integer key, TValue *value) {
  const TValue *p = luaH_getint(t, key);
  TValue *cell;
  if (p != luaO_nilobject)
    cell = cast(TValue *, p);
  else {
    TValue k;
    setivalue(&k, key);
    cell = luaH_newkey(L, t, &k);
  }
  setobj2t(L, cell, value);
}

这个是用 key 来保存值。

static int unbound_search (Table *t, unsigned int j) {
  unsigned int i = j;  /* i is zero or a present index */
  j++;
  /* find 'i' and 'j' such that i is present and j is not */
  while (!ttisnil(luaH_getint(t, j))) {
    i = j;
    if (j > cast(unsigned int, MAX_INT)/2) {  /* overflow? */
      /* table was built with bad purposes: resort to linear search */
      i = 1;
      while (!ttisnil(luaH_getint(t, i))) i++;
      return i - 1;
    }
    j *= 2;
  }
  /* now do a binary search between them */
  while (j - i > 1) {
    unsigned int m = (i+j)/2;
    if (ttisnil(luaH_getint(t, m))) j = m;
    else i = m;
  }
  return i;
}


/*
** Try to find a boundary in table 't'. A 'boundary' is an integer index
** such that t[i] is non-nil and t[i+1] is nil (and 0 if t[1] is nil).
*/
int luaH_getn (Table *t) {
  unsigned int j = t->sizearray;
  if (j > 0 && ttisnil(&t->array[j - 1])) {
    /* there is a boundary in the array part: (binary) search for it */
    unsigned int i = 0;
    while (j - i > 1) {
      unsigned int m = (i+j)/2;
      if (ttisnil(&t->array[m - 1])) j = m;
      else i = m;
    }
    return i;
  }
  /* else must find a boundary in hash part */
  else if (isdummy(t))  /* hash part is empty? */
    return j;  /* that is easy... */
  else return unbound_search(t, j);
}

这个看起来的作用就是注释描述的,但是具体作用我还不清楚。