public override void Build(InputIterator tfit)
{
if (tfit.HasPayloads)
{
throw new ArgumentException("this suggester doesn't support payloads");
}
if (tfit.Comparator != null)
{
// make sure it's unsorted
// WTF - this could result in yet another sorted iteration....
tfit = new UnsortedInputIterator(tfit);
}
if (tfit.HasContexts)
{
throw new System.ArgumentException("this suggester doesn't support contexts");
}
count = 0;
trie = new JaspellTernarySearchTrie();
trie.MatchAlmostDiff = editDistance;
BytesRef spare;
CharsRef charsSpare = new CharsRef();
while ((spare = tfit.Next()) != null)
{
long weight = tfit.Weight;
if (spare.Length == 0)
{
continue;
}
charsSpare.Grow(spare.Length);
UnicodeUtil.UTF8toUTF16(spare.Bytes, spare.Offset, spare.Length, charsSpare);
trie.Put(charsSpare.ToString(), Convert.ToInt64(weight));
}
}
public override void Build(IInputEnumerator enumerator)
{
// LUCENENET: Added guard clause for null
if (enumerator is null)
{
throw new ArgumentNullException(nameof(enumerator));
}
if (enumerator.HasPayloads)
{
throw new ArgumentException("this suggester doesn't support payloads");
}
if (enumerator.Comparer != null)
{
// make sure it's unsorted
// WTF - this could result in yet another sorted iteration....
enumerator = new UnsortedInputEnumerator(enumerator);
}
if (enumerator.HasContexts)
{
throw new ArgumentException("this suggester doesn't support contexts");
}
count = 0;
trie = new JaspellTernarySearchTrie {
MatchAlmostDiff = editDistance
};
BytesRef spare;
var charsSpare = new CharsRef();
while (enumerator.MoveNext())
{
spare = enumerator.Current;
long weight = enumerator.Weight;
if (spare.Length == 0)
{
continue;
}
charsSpare.Grow(spare.Length);
UnicodeUtil.UTF8toUTF16(spare.Bytes, spare.Offset, spare.Length, charsSpare);
trie.Put(charsSpare.ToString(), weight);
}
}
/// <summary>
/// Constructor method.
/// </summary>
/// <param name="outerInstance">The containing <see cref="JaspellTernarySearchTrie"/></param>
/// <param name="splitchar">
/// The char used in the split. </param>
/// <param name="parent">
/// The parent node. </param>
internal TSTNode(JaspellTernarySearchTrie outerInstance, char splitchar, TSTNode parent)
{
this.outerInstance = outerInstance;
this.splitchar = splitchar;
relatives[PARENT] = parent;
}
public override void Build(IInputIterator tfit)
{
if (tfit.HasPayloads)
{
throw new ArgumentException("this suggester doesn't support payloads");
}
if (tfit.Comparator != null)
{
// make sure it's unsorted
// WTF - this could result in yet another sorted iteration....
tfit = new UnsortedInputIterator(tfit);
}
if (tfit.HasContexts)
{
throw new System.ArgumentException("this suggester doesn't support contexts");
}
count = 0;
trie = new JaspellTernarySearchTrie { MatchAlmostDiff = editDistance };
BytesRef spare;
var charsSpare = new CharsRef();
while ((spare = tfit.Next()) != null)
{
long weight = tfit.Weight;
if (spare.Length == 0)
{
continue;
}
charsSpare.Grow(spare.Length);
UnicodeUtil.UTF8toUTF16(spare.Bytes, spare.Offset, spare.Length, charsSpare);
trie.Put(charsSpare.ToString(), Convert.ToInt64(weight));
}
}
private void WriteRecursively(DataOutput @out, JaspellTernarySearchTrie.TSTNode node)
{
if (node == null)
{
return;
}
@out.WriteString(new string(new char[] { node.splitchar }, 0, 1));
sbyte mask = 0;
if (node.relatives[JaspellTernarySearchTrie.TSTNode.LOKID] != null)
{
mask |= LO_KID;
}
if (node.relatives[JaspellTernarySearchTrie.TSTNode.EQKID] != null)
{
mask |= EQ_KID;
}
if (node.relatives[JaspellTernarySearchTrie.TSTNode.HIKID] != null)
{
mask |= HI_KID;
}
if (node.data != null)
{
mask |= HAS_VALUE;
}
@out.WriteByte((byte)mask);
if (node.data != null)
{
@out.WriteLong((long)(node.data));
}
WriteRecursively(@out, node.relatives[JaspellTernarySearchTrie.TSTNode.LOKID]);
WriteRecursively(@out, node.relatives[JaspellTernarySearchTrie.TSTNode.EQKID]);
WriteRecursively(@out, node.relatives[JaspellTernarySearchTrie.TSTNode.HIKID]);
}
private void ReadRecursively(DataInput @in, JaspellTernarySearchTrie.TSTNode node)
{
node.splitchar = @in.ReadString()[0];
sbyte mask = (sbyte)@in.ReadByte();
if ((mask & HAS_VALUE) != 0)
{
node.data = Convert.ToInt64(@in.ReadLong());
}
if ((mask & LO_KID) != 0)
{
var kid = new JaspellTernarySearchTrie.TSTNode(trie, '\0', node);
node.relatives[JaspellTernarySearchTrie.TSTNode.LOKID] = kid;
ReadRecursively(@in, kid);
}
if ((mask & EQ_KID) != 0)
{
var kid = new JaspellTernarySearchTrie.TSTNode(trie, '\0', node);
node.relatives[JaspellTernarySearchTrie.TSTNode.EQKID] = kid;
ReadRecursively(@in, kid);
}
if ((mask & HI_KID) != 0)
{
var kid = new JaspellTernarySearchTrie.TSTNode(trie, '\0', node);
node.relatives[JaspellTernarySearchTrie.TSTNode.HIKID] = kid;
ReadRecursively(@in, kid);
}
}
/// <summary>
/// Constructor method.
/// </summary>
/// <param name="splitchar">
/// The char used in the split. </param>
/// <param name="parent">
/// The parent node. </param>
protected internal TSTNode(JaspellTernarySearchTrie outerInstance, char splitchar, TSTNode parent)
{
this.outerInstance = outerInstance;
this.splitchar = splitchar;
relatives[PARENT] = parent;
}