Java Code Examples for edu.stanford.nlp.util.PropertiesUtils#getInt()

The following examples show how to use edu.stanford.nlp.util.PropertiesUtils#getInt() . You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may check out the related API usage on the sidebar.
Example 1
Source File: GermanPostprocessor.java    From phrasal with GNU General Public License v3.0 6 votes vote down vote up
/**
 * A main method for training and evaluating the postprocessor.
 * 
 * @param args
 */
public static void main(String[] args) {
  // Strips off hyphens
  Properties options = StringUtils.argsToProperties(args, optionArgDefs());
  if (options.containsKey("help") || args.length == 0) {
    System.err.println(usage(GermanPostprocessor.class.getName()));
    System.exit(-1);
  }

  int nThreads = PropertiesUtils.getInt(options, "nthreads", 1);
  GermanPreprocessor preProcessor = new GermanPreprocessor();
  GermanPostprocessor postProcessor = new GermanPostprocessor(options);
  
  CRFPostprocessor.setup(postProcessor, preProcessor, options);
  CRFPostprocessor.execute(nThreads, preProcessor, postProcessor);    
}
 
Example 2
Source File: FrenchPostprocessor.java    From phrasal with GNU General Public License v3.0 6 votes vote down vote up
/**
 * A main method for training and evaluating the postprocessor.
 * 
 * @param args
 */
public static void main(String[] args) {
  // Strips off hyphens
  Properties options = StringUtils.argsToProperties(args, optionArgDefs());
  if (options.containsKey("help") || args.length == 0) {
    System.err.println(usage(FrenchPostprocessor.class.getName()));
    System.exit(-1);
  }

  int nThreads = PropertiesUtils.getInt(options, "nthreads", 1);
  FrenchPreprocessor preProcessor = new FrenchPreprocessor();
  FrenchPostprocessor postProcessor = new FrenchPostprocessor(options);
  
  CRFPostprocessor.setup(postProcessor, preProcessor, options);
  CRFPostprocessor.execute(nThreads, preProcessor, postProcessor);    
}
 
Example 3
Source File: EnglishPostprocessor.java    From phrasal with GNU General Public License v3.0 6 votes vote down vote up
/**
 * A main method for training and evaluating the postprocessor.
 * 
 * @param args
 */
public static void main(String[] args) {
  // Strips off hyphens
  Properties options = StringUtils.argsToProperties(args, optionArgDefs());
  if (options.containsKey("help") || args.length == 0) {
    System.err.println(usage(EnglishPostprocessor.class.getName()));
    System.exit(-1);
  }

  int nThreads = PropertiesUtils.getInt(options, "nthreads", 1);
  EnglishPreprocessor preProcessor = new EnglishPreprocessor();
  EnglishPostprocessor postProcessor = new EnglishPostprocessor(options);
  
  CRFPostprocessor.setup(postProcessor, preProcessor, options);
  CRFPostprocessor.execute(nThreads, preProcessor, postProcessor);    
}
 
Example 4
Source File: SpanishPostprocessor.java    From phrasal with GNU General Public License v3.0 6 votes vote down vote up
/**
 * A main method for training and evaluating the postprocessor.
 * 
 * @param args
 */
public static void main(String[] args) {
  // Strips off hyphens
  Properties options = StringUtils.argsToProperties(args, optionArgDefs());
  if (options.containsKey("help") || args.length == 0) {
    System.err.println(usage(SpanishPostprocessor.class.getName()));
    System.exit(-1);
  }

  int nThreads = PropertiesUtils.getInt(options, "nthreads", 1);
  SpanishPreprocessor preProcessor = new SpanishPreprocessor();
  SpanishPostprocessor postProcessor = new SpanishPostprocessor(options);
  
  CRFPostprocessor.setup(postProcessor, preProcessor, options);
  CRFPostprocessor.execute(nThreads, preProcessor, postProcessor);    
}
 
Example 5
Source File: TargetFunctionWordInsertion.java    From phrasal with GNU General Public License v3.0 6 votes vote down vote up
/**
 * Constructor.
 * 
 * @param args
 */
public TargetFunctionWordInsertion(String...args) {
  Properties options = FeatureUtils.argsToProperties(args);
  if (args.length < 2) {
    throw new RuntimeException("Must specify source and target unigram counts files");
  }
  System.err.println("Loading TargetFunctionWordInsertion template...");
  String sourceFilename = options.getProperty("sourceFile");
  String targetFilename = options.getProperty("targetFile");
  this.rankCutoff = PropertiesUtils.getInt(options, "rankCutoff", DEFAULT_RANK_CUTOFF);
  
  System.err.println("Source words:");
  sourceFunctionWordSet = loadCountsFile(sourceFilename);
  System.err.println("Target words:");
  targetFunctionWordSet = loadCountsFile(targetFilename);
}
 
Example 6
Source File: SentenceLevelEvaluation.java    From phrasal with GNU General Public License v3.0 5 votes vote down vote up
/**
 * 
 * @param args
 * @throws IOException
 */
public static void main(String[] args) throws IOException {
  if (args.length < 1) {
    System.err.print(usage());
    System.exit(-1);
  }
  
  Properties options = StringUtils.argsToProperties(args, argDefs());
  int ngramOrder = PropertiesUtils.getInt(options, "order", BLEUMetric.DEFAULT_MAX_NGRAM_ORDER);
  boolean disableTokenization = PropertiesUtils.getBool(options, "no-nist", false);
  String metric = options.getProperty("metric", "bleu");

  String[] refs = options.getProperty("").split("\\s+");
  List<List<Sequence<IString>>> referencesList = MetricUtils.readReferences(refs, ! disableTokenization);
  System.err.printf("Metric: %s with %d references%n", metric, referencesList.get(0).size());
  
  LineNumberReader reader = new LineNumberReader(new InputStreamReader(
      System.in));
  int sourceInputId = 0;
  for (String line; (line = reader.readLine()) != null; ++sourceInputId) {
    line = disableTokenization ? line : NISTTokenizer.tokenize(line);
    Sequence<IString> translation = IStrings.tokenize(line);
    double score = getScore(translation, referencesList.get(sourceInputId), ngramOrder, metric);
    System.out.printf("%.4f%n", score);
  }
  System.err.printf("Scored %d input segments%n", sourceInputId);
}
 
Example 7
Source File: LexicalReorderingFeaturizer.java    From phrasal with GNU General Public License v3.0 5 votes vote down vote up
/**
 * Constructor for reflection loading discriminative lexicalized reordering.
 * 
 * @param args
 */
public LexicalReorderingFeaturizer(String...args) {
  Properties options = FeatureUtils.argsToProperties(args);
  this.dynamic = PropertiesUtils.getBool(options, "dynamic", false);
  if (dynamic) {
    this.discriminativeSet = null;
    this.mlrt = null;
    this.featureTags = Arrays.stream(LexicalReorderingTable.msdBidirectionalPositionMapping).map(m -> 
    String.format("%s:%s", FEATURE_PREFIX, m)).toArray(String[]::new);
    this.useAlignmentConstellations = false;
    this.useClasses = false;
    this.countFeatureIndex = -1;
    this.lexicalCutoff = 0;

  } else {
    this.discriminativeSet = new ArrayList<>(Arrays.asList(LexicalReorderingTable.ReorderingTypes.values()));
    this.useAlignmentConstellations = options.containsKey("conditionOnConstellations");
    this.countFeatureIndex = PropertiesUtils.getInt(options, "countFeatureIndex", -1);
    // Which reordering classes to extract
    if (options.containsKey("classes")) {
      String[] typeStrings = options.getProperty("classes").split("-");
      discriminativeSet = new ArrayList<>();
      for (String type : typeStrings) {
        discriminativeSet.add(LexicalReorderingTable.ReorderingTypes.valueOf(type));
      }
    }
    // Use class-based feature representations
    this.useClasses = options.containsKey("useClasses");
    if (useClasses) {
      sourceMap = SourceClassMap.getInstance();
      targetMap = TargetClassMap.getInstance();
    }
    this.mlrt = null;
    this.featureTags = null;
    this.lexicalCutoff = PropertiesUtils.getInt(options, "lexicalCutoff", 0);
  }
}
 
Example 8
Source File: RuleIndicator.java    From phrasal with GNU General Public License v3.0 5 votes vote down vote up
/**
 * Constructor for reflection loading.
 * 
 * @param args
 */
public RuleIndicator(String... args) {
  Properties options = FeatureUtils.argsToProperties(args);
  this.addLexicalizedRule = options.containsKey("addLexicalized");
  this.addClassBasedRule = options.containsKey("addClassBased");

  this.countFeatureIndex = PropertiesUtils.getInt(options, "countFeatureIndex", -1);
  if (addClassBasedRule) {
    sourceMap = SourceClassMap.getInstance();
    targetMap = TargetClassMap.getInstance();
  }
  this.lexicalCutoff = PropertiesUtils.getInt(options, "lexicalCutoff", 0);
}
 
Example 9
Source File: PhraseViewer.java    From phrasal with GNU General Public License v3.0 5 votes vote down vote up
private static boolean validateCommandLine(String[] args) {
  // Command line parsing
  Properties options = StringUtils.argsToProperties(args, argDefs());

  VERBOSE = options.containsKey("v");
  SRC_FILE = options.getProperty("s", null);
  OPTS_FILE = options.getProperty("o", null);
  XSD_FILE = options.getProperty("x", null);
  FIRST_ID = PropertiesUtils.getInt(options, "f", Integer.MIN_VALUE);
  LAST_ID = PropertiesUtils.getInt(options,"l",Integer.MAX_VALUE);

  return true;
}
 
Example 10
Source File: BLEUMetric.java    From phrasal with GNU General Public License v3.0 4 votes vote down vote up
public static void main(String[] args) throws IOException {
  if (args.length < 1) {
    System.err.print(usage());
    System.exit(-1);
  }

  Properties options = StringUtils.argsToProperties(args, argDefs());
  int BLEUOrder = PropertiesUtils.getInt(options, "order", BLEUMetric.DEFAULT_MAX_NGRAM_ORDER);
  boolean doSmooth = PropertiesUtils.getBool(options, "smooth", false);
  boolean disableTokenization = PropertiesUtils.getBool(options, "no-nist", false);
  boolean doCased = PropertiesUtils.getBool(options, "cased", false);

  // Setup the metric tokenization scheme. Applies to both the references and
  // hypotheses
  if (doCased) NISTTokenizer.lowercase(false);
  NISTTokenizer.normalize( ! disableTokenization);

  // Load the references
  String[] refs = options.getProperty("").split("\\s+");
  System.out.printf("Metric: BLEU-%d with %d references%n", BLEUOrder, refs.length);
  List<List<Sequence<IString>>> referencesList = MetricUtils.readReferences(refs, true);

  // For backwards compatibility
  doSmooth |= System.getProperty("smoothBLEU") != null;

  BLEUMetric<IString, String> bleu = new BLEUMetric<IString, String>(referencesList, BLEUOrder,
        doSmooth);
  BLEUMetric<IString, String>.BLEUIncrementalMetric incMetric = bleu
      .getIncrementalMetric();

  LineNumberReader reader = new LineNumberReader(new InputStreamReader(
      System.in));
  for (String line; (line = reader.readLine()) != null; ) {
    line = NISTTokenizer.tokenize(line);
    Sequence<IString> translation = IStrings.tokenize(line);
    ScoredFeaturizedTranslation<IString, String> tran = new ScoredFeaturizedTranslation<IString, String>(
        translation, null, 0);
    incMetric.add(tran);
  }
  // Check for an incomplete set of translations
  if (reader.getLineNumber() < referencesList.size()) {
    System.err.printf("WARNING: Translation candidate file is shorter than references (%d/%d)%n", 
        reader.getLineNumber(), referencesList.size());
  }
  reader.close();

  double[] ngramPrecisions = incMetric.ngramPrecisions();
  System.out.printf("BLEU = %.3f, ", 100 * incMetric.score());
  for (int i = 0; i < ngramPrecisions.length; i++) {
    if (i != 0) {
      System.out.print("/");
    }
    System.out.printf("%.3f", ngramPrecisions[i] * 100);
  }
  System.out.printf(" (BP=%.3f, ratio=%.3f %d/%d)%n", incMetric
      .brevityPenalty(), ((1.0 * incMetric.candidateLength()) / incMetric
      .effectiveReferenceLength()), incMetric.candidateLength(), incMetric
      .effectiveReferenceLength());

  System.out.printf("%nPrecision Details:%n");
  double[][] precCounts = incMetric.ngramPrecisionCounts();
  for (int i = 0; i < ngramPrecisions.length; i++) {
    System.out.printf("\t%d:%d/%d%n", i, (int) precCounts[i][0], (int) precCounts[i][1]);
  }
}
 
Example 11
Source File: DependencyProjector.java    From phrasal with GNU General Public License v3.0 4 votes vote down vote up
public static void main(String[] args) throws IOException {
  Properties options = StringUtils.argsToProperties(args, optionArgDefs());
  String sourceTokens = PropertiesUtils.get(options, "sourceTokens", null, String.class);
  String targetTokens = PropertiesUtils.get(options, "targetTokens", null, String.class);
  String alignments = PropertiesUtils.get(options, "alignment", null, String.class);
  String annotations = PropertiesUtils.get(options, "annotations", null, String.class);
  String outdirPath = PropertiesUtils.get(options, "outdir", ".", String.class);
  String leftDepLMFilename = outdirPath + File.separator + "left.deplm";
  String rightDepLMFilename = outdirPath + File.separator + "right.deplm";
  String headDepLMFilename = outdirPath + File.separator + "head.deplm";


  File leftDepLMFile = new File(leftDepLMFilename);
  if (!leftDepLMFile.exists())
    leftDepLMFile.createNewFile();

  
  File rightDepLMFile = new File(rightDepLMFilename);
  if (!rightDepLMFile.exists())
    rightDepLMFile.createNewFile();
  
  File headDepLMFile = new File(headDepLMFilename);
  if (!headDepLMFile.exists())
    headDepLMFile.createNewFile();

  FileWriter leftFW = new FileWriter(leftDepLMFile.getAbsoluteFile());
  FileWriter rightFW = new FileWriter(rightDepLMFile.getAbsoluteFile());
  FileWriter headFW = new FileWriter(headDepLMFile.getAbsoluteFile());

  leftDepLMWriter = new BufferedWriter(leftFW);
  rightDepLMWriter = new BufferedWriter(rightFW);
  headDepLMWriter = new BufferedWriter(headFW);

  
  boolean annotationsSplit = PropertiesUtils.getBool(options, "annotationsSplit", false);
  boolean transitive = PropertiesUtils.getBool(options, "transitive", false);

  int maxDepth = PropertiesUtils.getInt(options, "maxDepth", 2);

  
  File sourceSentences = new File(sourceTokens);
  File targetSentences = new File(targetTokens);
  File alignmentFile = new File(alignments);
  BufferedReader sourceReader = new BufferedReader(new FileReader(sourceSentences));
  BufferedReader targetReader = new BufferedReader(new FileReader(targetSentences));
  BufferedReader alignmentReader = new BufferedReader(new FileReader(alignmentFile));
  String sourceSentence;
  int i = 0;
  while ((sourceSentence = sourceReader.readLine()) != null) {
    //try {
      CoreMap sentence = getParsedSentence(annotations, i, annotationsSplit);
      String targetSentence = targetReader.readLine();
      String alignmentString = alignmentReader.readLine();
      //System.err.println("---------------------------");
      //System.err.println("alignment = \"" + alignmentString + "\";");
      SymmetricalWordAlignment alignment = new SymmetricalWordAlignment(sourceSentence, targetSentence, alignmentString);
      //projectSentence(sentence, alignment);
      Map<Integer, NavigableSet<Integer>> dependencies = projectDependencies(sentence, alignment, transitive, maxDepth);
      //if (i == 0) {
      //  System.err.println(dependencies.get(-1));
      //  System.err.println(dependencies.get(1));

      //}
      //printDependencyString(dependencies, -1, alignment.e(), "");
      //System.out.println(dependencies);
      printLeftAndRightDependencies(dependencies, alignment.e());
      //System.err.println("---------------------------");
    //} catch (Exception e) {
    //  e.printStackTrace();
    //}
    i++;
  }

  
  sourceReader.close();
  targetReader.close();
  alignmentReader.close();

  leftDepLMWriter.close();
  rightDepLMWriter.close();
  headDepLMWriter.close();
  
}
 
Example 12
Source File: BLEUGenreEvaluator.java    From phrasal with GNU General Public License v3.0 4 votes vote down vote up
/**
 * 
 * @param args
 * @throws IOException
 */
public static void main(String[] args) throws IOException {
  if (args.length < 2) {
    System.err.print(usage());
    System.exit(-1);
  }

  Properties options = StringUtils.argsToProperties(args, argDefs());
  int BLEUOrder = PropertiesUtils.getInt(options, "order", BLEUMetric.DEFAULT_MAX_NGRAM_ORDER);
  boolean doCased = PropertiesUtils.getBool(options, "cased", false);

  // Setup the metric tokenization scheme. Applies to both the references and
  // hypotheses
  if (doCased) NISTTokenizer.lowercase(false);

  // Load the references
  String[] parameters = options.getProperty("").split("\\s+");
  String[] refs = new String[parameters.length - 1];
  System.arraycopy(parameters, 1, refs, 0, refs.length);
  List<InputProperties> inputProperties = InputProperties.parse(new File(parameters[0]));
  List<List<Sequence<IString>>> referencesList = MetricUtils.readReferences(refs, true);
  
  Map<String,BLEUMetric<IString, String>.BLEUIncrementalMetric> metrics = 
      BLEUGenreEvaluator.run(referencesList, inputProperties, BLEUOrder, System.in);

  for (Map.Entry<String,BLEUMetric<IString, String>.BLEUIncrementalMetric> entry : metrics.entrySet()) {
    String genre = entry.getKey();
    BLEUMetric<IString, String>.BLEUIncrementalMetric incMetric = entry.getValue();
    System.out.printf("Genre: %s%n", genre);
    double[] ngramPrecisions = incMetric.ngramPrecisions();
    System.out.printf("BLEU = %.3f, ", 100 * incMetric.score());
    for (int i = 0; i < ngramPrecisions.length; i++) {
      if (i != 0) {
        System.out.print("/");
      }
      System.out.printf("%.3f", ngramPrecisions[i] * 100);
    }
    System.out.printf(" (BP=%.3f, ratio=%.3f %d/%d)%n", incMetric
        .brevityPenalty(), ((1.0 * incMetric.candidateLength()) / incMetric
            .effectiveReferenceLength()), incMetric.candidateLength(), incMetric
            .effectiveReferenceLength());

    System.out.printf("%nPrecision Details:%n");
    double[][] precCounts = incMetric.ngramPrecisionCounts();
    for (int i = 0; i < ngramPrecisions.length; i++) {
      System.out.printf("\t%d:%d/%d%n", i, (int) precCounts[i][0], (int) precCounts[i][1]);
    }
    System.out.println();
  }
}
 
Example 13
Source File: PhrasalService.java    From phrasal with GNU General Public License v3.0 4 votes vote down vote up
/**
   * Start the service.
   * 
   * @param args
   */
  public static void main(String[] args) {
    Properties options = StringUtils.argsToProperties(args, optionArgDefs());
    int port = PropertiesUtils.getInt(options, "p", DEFAULT_HTTP_PORT);
    boolean loadMockServlet = PropertiesUtils.getBool(options, "m", false);
    boolean localHost = PropertiesUtils.getBool(options, "l", false);
    String uiFile = options.getProperty("u", "debug.html");
    String resourcePath = options.getProperty("r", ".");

    // Parse arguments
    String argList = options.getProperty("",null);
    String[] parsedArgs = argList == null ? null : argList.split("\\s+");
    if (parsedArgs == null || parsedArgs.length != 1) {
      System.out.println(usage());
      System.exit(-1);
    }
    String phrasalIniFile = parsedArgs[0];
    
    // Setup the jetty server
    Server server = new Server();

    // Jetty 8 way of configuring the server
//    Connector connector = new SelectChannelConnector();
//    connector.setPort(port);
//    server.addConnector(connector);

//  Jetty9 way of configuring the server
    ServerConnector connector = new ServerConnector(server);
    connector.setPort(port);
    server.addConnector(connector);

    if (localHost) {
      connector.setHost(DEBUG_URL);
    }
    
    // Setup the servlet context
    ServletContextHandler context = new ServletContextHandler(ServletContextHandler.SESSIONS);
    context.setContextPath("/");
 
    // Add Phrasal servlet
    PhrasalServlet servlet = loadMockServlet ? new PhrasalServlet() : new PhrasalServlet(phrasalIniFile);
    context.addServlet(new ServletHolder(servlet), SERVLET_ROOT);

    // TODO(spenceg): gzip compression causes an encoding problem for unicode characters
    // on the client. Not sure if the compression or decompression is the problem.
//    EnumSet<DispatcherType> dispatches = EnumSet.of(DispatcherType.REQUEST, DispatcherType.ASYNC);
//    context.addFilter(new FilterHolder(new IncludableGzipFilter()), "/t", dispatches);

    // Add debugging web-page
    ResourceHandler resourceHandler = new ResourceHandler();
    resourceHandler.setWelcomeFiles(new String[]{ uiFile });
    resourceHandler.setResourceBase(resourcePath);

    HandlerList handlers = new HandlerList();
    handlers.setHandlers(new Handler[] { resourceHandler, context });
    server.setHandler(handlers);
    
    // Start the service
    try {
      logger.info("Starting PhrasalService on port: " + String.valueOf(port));
      server.start();
      server.join();
    } catch (Exception e) {
      logger.error("Servlet crashed. Service shutting down.");
      e.printStackTrace();
    }
  }