edu.stanford.nlp.objectbank.DelimitRegExIterator Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of stanford-parser Show documentation

Stanford Parser processes raw text in English, Chinese, German, Arabic, and French, and extracts constituency parse trees.

The newest version!

package edu.stanford.nlp.objectbank;

import java.io.BufferedReader;
import java.io.IOException;
import java.io.Reader;
import java.util.ArrayList;
import java.util.Iterator;
import java.util.NoSuchElementException;
import java.util.function.Function;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

import edu.stanford.nlp.io.RuntimeIOException;
import edu.stanford.nlp.util.AbstractIterator;


/**
 * An Iterator that reads the contents of a Reader, delimited by the specified
 * delimiter, and then subsequently processed by an Function to produce
 * Objects of type T.
 *
 * @author Jenny Finkel  getFactory(String delim) {
    return DelimitRegExIteratorFactory.defaultDelimitRegExIteratorFactory(delim);
  }

  /**
   * Returns a factory that vends DelimitRegExIterators that reads the contents of the
   * given Reader, splits on the specified delimiter, applies op, then returns the result.
   */
  public static  IteratorFromReaderFactory getFactory(String delim, Function op) {
    return new DelimitRegExIteratorFactory<>(delim, op);
  }

  public static class DelimitRegExIteratorFactory implements IteratorFromReaderFactory /*, Serializable */ {

    private static final long serialVersionUID = 6846060575832573082L;

    private final String delim;
    @SuppressWarnings("serial")
    private final Function op;

    public static DelimitRegExIteratorFactory defaultDelimitRegExIteratorFactory(String delim) {
      return new DelimitRegExIteratorFactory<>(delim, new IdentityFunction<>());
    }

    public DelimitRegExIteratorFactory(String delim, Function op) {
      this.delim = delim;
      this.op = op;
    }

    @Override
    public Iterator getIterator(Reader r) {
      return new DelimitRegExIterator<>(r, delim, op);
    }

  }

}