public virtual void TestEnglishDependenciesByTree()
{
// the trees to test
string[] testTrees = new string[] { "((S (NP (NNP Sam)) (VP (VBD died) (NP-TMP (NN today)))))", "(ROOT (S (NP (PRP I)) (VP (VBD saw) (NP (NP (DT the) (NN book)) (SBAR (WHNP (WDT which)) (S (NP (PRP you)) (VP (VBD bought)))))) (. .)))" };
// the expected dependency answers (basic)
string[] testAnswers = new string[] { "root(ROOT-0, died-2) nsubj(died-2, Sam-1) tmod(died-2, today-3)", "nsubj(saw-2, I-1) root(ROOT-0, saw-2) det(book-4, the-3) dobj(saw-2, book-4) dobj(bought-7, which-5) ref(book-4, which-5) dobj(bought-7, which-5) nsubj(bought-7, you-6) rcmod(book-4, bought-7)" };
// the expected dependency answers (collapsed dependencies)
string[] testAnswersCollapsed = new string[] { "root(ROOT-0, died-2) nsubj(died-2, Sam-1) tmod(died-2, today-3)", "nsubj(saw-2, I-1) root(ROOT-0, saw-2) det(book-4, the-3) dobj(saw-2, book-4) dobj(bought-7, book-4) nsubj(bought-7, you-6) rcmod(book-4, bought-7)" };
// the expected dependency answers (conjunctions processed)
string[] testAnswersCCProcessed = new string[] { "root(ROOT-0, died-2) nsubj(died-2, Sam-1) tmod(died-2, today-3)", "nsubj(saw-2, I-1) root(ROOT-0, saw-2) det(book-4, the-3) dobj(saw-2, book-4) dobj(bought-7, book-4) nsubj(bought-7, you-6) rcmod(book-4, bought-7)" };
for (int i = 0; i < testTrees.Length; i++)
{
string testTree = testTrees[i];
string testAnswer = testAnswers[i];
string testAnswerCollapsed = testAnswersCollapsed[i];
string testAnswerCCProcessed = testAnswersCCProcessed[i];
HashSet <string> testAnswerTokens = TokenSet(testAnswer);
HashSet <string> testAnswerCollapsedTokens = TokenSet(testAnswerCollapsed);
HashSet <string> testAnswerCCProcessedTokens = TokenSet(testAnswerCCProcessed);
Tree tree;
try
{
tree = new PennTreeReader(new StringReader(testTree), new LabeledScoredTreeFactory()).ReadTree();
}
catch (IOException e)
{
// these trees should all parse correctly
throw new Exception(e);
}
GrammaticalStructure gs = new EnglishGrammaticalStructure(tree);
NUnit.Framework.Assert.AreEqual("Unexpected basic dependencies for tree " + testTree, testAnswerTokens, TokenSet(gs.TypedDependencies(GrammaticalStructure.Extras.Maximal)));
NUnit.Framework.Assert.AreEqual("Unexpected collapsed dependencies for tree " + testTree, testAnswerCollapsedTokens, TokenSet(gs.TypedDependenciesCollapsed(GrammaticalStructure.Extras.Maximal)));
NUnit.Framework.Assert.AreEqual("Unexpected cc-processed dependencies for tree " + testTree, testAnswerCCProcessedTokens, TokenSet(gs.TypedDependenciesCCprocessed(GrammaticalStructure.Extras.Maximal)));
}
}
public static void Main(string[] args)
{
Edu.Stanford.Nlp.Trees.CoordinationTransformer transformer = new Edu.Stanford.Nlp.Trees.CoordinationTransformer(null);
Treebank tb = new MemoryTreebank();
Properties props = StringUtils.ArgsToProperties(args);
string treeFileName = props.GetProperty("treeFile");
if (treeFileName != null)
{
try
{
ITreeReader tr = new PennTreeReader(new BufferedReader(new InputStreamReader(new FileInputStream(treeFileName))), new LabeledScoredTreeFactory());
for (Tree t; (t = tr.ReadTree()) != null;)
{
tb.Add(t);
}
}
catch (IOException e)
{
throw new Exception("File problem: " + e);
}
}
foreach (Tree t_1 in tb)
{
System.Console.Out.WriteLine("Original tree");
t_1.PennPrint();
System.Console.Out.WriteLine();
System.Console.Out.WriteLine("Tree transformed");
Tree tree = transformer.TransformTree(t_1);
tree.PennPrint();
System.Console.Out.WriteLine();
System.Console.Out.WriteLine("----------------------------");
}
}
// todo [cdm 2015]: These next two methods duplicate the Tree.valueOf methods!
/// <summary>Simple tree reading utility method.</summary>
/// <remarks>Simple tree reading utility method. Given a tree formatted as a PTB string, returns a Tree made by a specific TreeFactory.</remarks>
public static Tree ReadTree(string ptbTreeString, ITreeFactory treeFactory)
{
try
{
PennTreeReader ptr = new PennTreeReader(new StringReader(ptbTreeString), treeFactory);
return(ptr.ReadTree());
}
catch (IOException ex)
{
throw new Exception(ex);
}
}
public virtual void TestRead()
{
string treeText = "(1 (2 This)) (3 (4 is) (5 a)) (6 (\\* small) (7 \\/test))";
StringReader reader = new StringReader(treeText);
PennTreeReader treeReader = new PennTreeReader(reader);
string[] expected = new string[] { "(1 (2 This))", "(3 (4 is) (5 a))", "(6 (* small) (7 /test))" };
for (int i = 0; i < expected.Length; ++i)
{
Tree tree = treeReader.ReadTree();
NUnit.Framework.Assert.IsTrue(tree != null);
NUnit.Framework.Assert.AreEqual(expected[i], tree.ToString());
}
Tree tree_1 = treeReader.ReadTree();
NUnit.Framework.Assert.IsFalse(tree_1 != null);
}
/// <summary>Just for testing.</summary>
public static void Main(string[] args)
{
try
{
ITreeReader tr = new PennTreeReader(new StringReader("(S (NP (NNP Sam)) (VP (VBD died) (NP (NN today))))"), new LabeledScoredTreeFactory());
Tree t = tr.ReadTree();
System.Console.Out.WriteLine(t);
TreeGraphNode tgn = new TreeGraphNode(t, (TreeGraphNode)null);
System.Console.Out.WriteLine(tgn.ToPrettyString(0));
EnglishGrammaticalStructure gs = new EnglishGrammaticalStructure(tgn);
System.Console.Out.WriteLine(tgn.ToPrettyString(0));
tgn.PercolateHeads(new SemanticHeadFinder());
System.Console.Out.WriteLine(tgn.ToPrettyString(0));
}
catch (Exception e)
{
log.Error("Horrible error: " + e);
log.Error(e);
}
}