private void Compact(CharVector kx, TernaryTree map, char p)
{
int k;
if (p == 0)
{
return;
}
if (m_sc[p] == 0xFFFF)
{
k = map.Find(m_kv.Array, m_lo[p]);
if (k < 0)
{
k = kx.Alloc(StrLen(m_kv.Array, m_lo[p]) + 1);
StrCpy(kx.Array, k, m_kv.Array, m_lo[p]);
map.Insert(kx.Array, k, (char)k);
}
m_lo[p] = (char)k;
}
else
{
Compact(kx, map, m_lo[p]);
if (m_sc[p] != 0)
{
Compact(kx, map, m_eq[p]);
}
Compact(kx, map, m_hi[p]);
}
}
/// <summary>
/// Each node stores a character (splitchar) which is part of some key(s). In a
/// compressed branch (one that only contain a single string key) the trailer
/// of the key which is not already in nodes is stored externally in the kv
/// array. As items are inserted, key substrings decrease. Some substrings may
/// completely disappear when the whole branch is totally decompressed. The
/// tree is traversed to find the key substrings actually used. In addition,
/// duplicate substrings are removed using a map (implemented with a
/// TernaryTree!).
///
/// </summary>
public virtual void TrimToSize()
{
// first balance the tree for best performance
Balance();
// redimension the node arrays
RedimNodeArrays(m_freenode);
// ok, compact kv array
CharVector kx = new CharVector();
kx.Alloc(1);
TernaryTree map = new TernaryTree();
Compact(kx, map, m_root);
m_kv = kx;
m_kv.TrimToSize();
}
/// <summary>
/// The actual insertion function, recursive version.
/// </summary>
private char Insert(char p, char[] key, int start, char val)
{
int len = StrLen(key, start);
if (p == 0)
{
// this means there is no branch, this node will start a new branch.
// Instead of doing that, we store the key somewhere else and create
// only one node with a pointer to the key
p = m_freenode++;
m_eq[p] = val; // holds data
m_length++;
m_hi[p] = (char)0;
if (len > 0)
{
m_sc[p] = (char)0xFFFF; // indicates branch is compressed
m_lo[p] = (char)m_kv.Alloc(len + 1); // use 'lo' to hold pointer to key
StrCpy(m_kv.Array, m_lo[p], key, start);
}
else
{
m_sc[p] = (char)0;
m_lo[p] = (char)0;
}
return(p);
}
if (m_sc[p] == 0xFFFF)
{
// branch is compressed: need to decompress
// this will generate garbage in the external key array
// but we can do some garbage collection later
char pp = m_freenode++;
m_lo[pp] = m_lo[p]; // previous pointer to key
m_eq[pp] = m_eq[p]; // previous pointer to data
m_lo[p] = (char)0;
if (len > 0)
{
m_sc[p] = m_kv[m_lo[pp]];
m_eq[p] = pp;
m_lo[pp]++;
if (m_kv[m_lo[pp]] == 0)
{
// key completly decompressed leaving garbage in key array
m_lo[pp] = (char)0;
m_sc[pp] = (char)0;
m_hi[pp] = (char)0;
}
else
{
// we only got first char of key, rest is still there
m_sc[pp] = (char)0xFFFF;
}
}
else
{
// In this case we can save a node by swapping the new node
// with the compressed node
m_sc[pp] = (char)0xFFFF;
m_hi[p] = pp;
m_sc[p] = (char)0;
m_eq[p] = val;
m_length++;
return(p);
}
}
char s = key[start];
if (s < m_sc[p])
{
m_lo[p] = Insert(m_lo[p], key, start, val);
}
else if (s == m_sc[p])
{
if (s != 0)
{
m_eq[p] = Insert(m_eq[p], key, start + 1, val);
}
else
{
// key already in tree, overwrite data
m_eq[p] = val;
}
}
else
{
m_hi[p] = Insert(m_hi[p], key, start, val);
}
return(p);
}
