/// <param name="args"/>
public static void Main(string[] args)
{
Properties options = StringUtils.ArgsToProperties(args, argOptionDefs);
if (!options.Contains(string.Empty) || options.Contains("help"))
{
log.Info(Usage());
return;
}
bool retainNER = PropertiesUtils.GetBool(options, "ner", false);
bool normalize = PropertiesUtils.GetBool(options, "normalize", true);
File treeFile = new File(options.GetProperty(string.Empty));
TwoDimensionalCounter <string, string> labelTerm = new TwoDimensionalCounter <string, string>();
TwoDimensionalCounter <string, string> termLabel = new TwoDimensionalCounter <string, string>();
TwoDimensionalCounter <string, string> labelPreterm = new TwoDimensionalCounter <string, string>();
TwoDimensionalCounter <string, string> pretermLabel = new TwoDimensionalCounter <string, string>();
TwoDimensionalCounter <string, string> unigramTagger = new TwoDimensionalCounter <string, string>();
try
{
BufferedReader br = new BufferedReader(new InputStreamReader(new FileInputStream(treeFile), "UTF-8"));
ITreeReaderFactory trf = new SpanishTreeReaderFactory();
ITreeReader tr = trf.NewTreeReader(br);
for (Tree t; (t = tr.ReadTree()) != null;)
{
UpdateTagger(unigramTagger, t);
}
tr.Close();
//Closes the underlying reader
System.Console.Out.WriteLine("Resolving DUMMY tags");
ResolveDummyTags(treeFile, unigramTagger, retainNER, normalize ? new SpanishTreeNormalizer(true, false, false) : null);
System.Console.Out.WriteLine("#Unknown Word Types: " + MultiWordPreprocessor.ManualUWModel.nUnknownWordTypes);
System.Console.Out.WriteLine(string.Format("#Missing POS: %d (fixed: %d, %.2f%%)", nMissingPOS, nFixedPOS, (double)nFixedPOS / nMissingPOS * 100));
System.Console.Out.WriteLine(string.Format("#Missing Phrasal: %d (fixed: %d, %.2f%%)", nMissingPhrasal, nFixedPhrasal, (double)nFixedPhrasal / nMissingPhrasal * 100));
System.Console.Out.WriteLine("Done!");
}
catch (UnsupportedEncodingException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
catch (FileNotFoundException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
catch (IOException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
}
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);
}
}
public static void Main(string[] args)
{
if (args.Length < 1)
{
System.Console.Error.Printf("Usage: java %s tree_file%n", typeof(TreeToTSV).FullName);
System.Environment.Exit(-1);
}
string treeFile = args[0];
try
{
BufferedReader br = new BufferedReader(new InputStreamReader(new FileInputStream(treeFile), "UTF-8"));
ITreeReaderFactory trf = new SpanishTreeReaderFactory();
ITreeReader tr = trf.NewTreeReader(br);
StringBuilder sb = new StringBuilder();
string nl = Runtime.GetProperty("line.separator");
Pattern nePattern = Pattern.Compile("^grup\\.nom\\.");
Pattern npPattern = Pattern.Compile("^np0000.$");
for (Tree tree; (tree = tr.ReadTree()) != null;)
{
foreach (Tree t in tree)
{
if (!t.IsPreTerminal())
{
continue;
}
char type = 'O';
Tree grandma = t.Ancestor(1, tree);
string grandmaValue = ((CoreLabel)grandma.Label()).Value();
// grup.nom.x
if (nePattern.Matcher(grandmaValue).Find())
{
type = grandmaValue[9];
}
else
{
// else check the pos for np0000x or not
string pos = ((CoreLabel)t.Label()).Value();
if (npPattern.Matcher(pos).Find())
{
type = pos[6];
}
}
Tree wordNode = t.FirstChild();
string word = ((CoreLabel)wordNode.Label()).Value();
sb.Append(word).Append("\t");
switch (type)
{
case 'p':
{
sb.Append("PERS");
break;
}
case 'l':
{
sb.Append("LUG");
break;
}
case 'o':
{
sb.Append("ORG");
break;
}
case '0':
{
sb.Append("OTROS");
break;
}
default:
{
sb.Append("O");
break;
}
}
sb.Append(nl);
}
sb.Append(nl);
}
System.Console.Out.Write(sb.ToString());
tr.Close();
}
catch (UnsupportedEncodingException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
catch (FileNotFoundException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
catch (IOException e)
{
Sharpen.Runtime.PrintStackTrace(e);
}
}