public virtual ICollection <Tree> Process(ICollection <Tree> coll)
{
IList <Tree> ret = new List <Tree>();
// Apparently TsurgeonPatterns are not thread safe
MultiWordTreeExpander expander = new MultiWordTreeExpander();
foreach (Tree t in coll)
{
// Begin with basic POS / phrasal category inference
MultiWordPreprocessor.TraverseAndFix(t, null, this._enclosing.unigramTagger, this.ner);
// Now "decompress" further the expanded trees formed by multiword token splitting
t = expander.ExpandPhrases(t, this.tn, this.tf);
t = this.tn.NormalizeWholeTree(t, this.tf);
ret.Add(t);
}
return(ret);
}
private static void ResolveDummyTags(File treeFile, TwoDimensionalCounter <string, string> unigramTagger, bool retainNER, TreeNormalizer tn)
{
ITreeFactory tf = new LabeledScoredTreeFactory();
MultiWordTreeExpander expander = new MultiWordTreeExpander();
try
{
BufferedReader br = new BufferedReader(new InputStreamReader(new FileInputStream(treeFile), "UTF-8"));
ITreeReaderFactory trf = new SpanishTreeReaderFactory();
ITreeReader tr = trf.NewTreeReader(br);
PrintWriter pw = new PrintWriter(new TextWriter(new FileOutputStream(new File(treeFile + ".fixed")), false, "UTF-8"));
int nTrees = 0;
for (Tree t; (t = tr.ReadTree()) != null; nTrees++)
{
TraverseAndFix(t, null, unigramTagger, retainNER);
// Now "decompress" further the expanded trees formed by
// multiword token splitting
t = expander.ExpandPhrases(t, tn, tf);
if (tn != null)
{
t = tn.NormalizeWholeTree(t, tf);
}
pw.Println(t.ToString());
}
pw.Close();
tr.Close();
System.Console.Out.WriteLine("Processed " + nTrees + " trees");
}
catch (UnsupportedEncodingException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
catch (FileNotFoundException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
catch (IOException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
}
