/// <summary>
/// Reads a single tree in standard Penn Treebank format from the
/// input stream.
/// </summary>
/// <remarks>
/// Reads a single tree in standard Penn Treebank format from the
/// input stream. The method supports additional parentheses around the
/// tree (an unnamed ROOT node) so long as they are balanced. If the token stream
/// ends before the current tree is complete, then the method will throw an
/// <code>IOException</code>.
/// <p>
/// Note that the method will skip malformed trees and attempt to
/// read additional trees from the input stream. It is possible, however,
/// that a malformed tree will corrupt the token stream. In this case,
/// an <code>IOException</code> will eventually be thrown.
/// </remarks>
/// <returns>A single tree, or <code>null</code> at end of token stream.</returns>
/// <exception cref="System.IO.IOException"/>
public virtual Tree ReadTree()
{
Tree t = null;
while (tokenizer.MoveNext() && t == null)
{
//Setup PDA
this.currentTree = null;
this.stack = new List <Tree>();
try
{
t = GetTreeFromInputStream();
}
catch (NoSuchElementException)
{
throw new IOException("End of token stream encountered before parsing could complete.");
}
if (t != null)
{
// cdm 20100618: Don't do this! This was never the historical behavior!!!
// Escape empty trees e.g. (())
// while(t != null && (t.value() == null || t.value().equals("")) && t.numChildren() <= 1)
// t = t.firstChild();
if (treeNormalizer != null && treeFactory != null)
{
t = treeNormalizer.NormalizeWholeTree(t, treeFactory);
}
if (t != null)
{
t.IndexLeaves(true);
}
}
}
return(t);
}
public static Tree NormalizeTree(Tree tree, TreeNormalizer tn, ITreeFactory tf)
{
foreach (Tree node in tree)
{
if (node.IsLeaf())
{
node.Label().SetValue(tn.NormalizeTerminal(node.Label().Value()));
}
else
{
node.Label().SetValue(tn.NormalizeNonterminal(node.Label().Value()));
}
}
return(tn.NormalizeWholeTree(tree, tf));
}