Java Code Examples for org.apache.lucene.index.Terms#getDocCount()

/**
 * count the number of documents in the index having at least a value for the 'class' field
 *
 * @return the no. of documents having a value for the 'class' field
 * @throws IOException if accessing to term vectors or search fails
 */
protected int countDocsWithClass() throws IOException {
  Terms terms = MultiTerms.getTerms(this.indexReader, this.classFieldName);
  int docCount;
  if (terms == null || terms.getDocCount() == -1) { // in case codec doesn't support getDocCount
    TotalHitCountCollector classQueryCountCollector = new TotalHitCountCollector();
    BooleanQuery.Builder q = new BooleanQuery.Builder();
    q.add(new BooleanClause(new WildcardQuery(new Term(classFieldName, String.valueOf(WildcardQuery.WILDCARD_STRING))), BooleanClause.Occur.MUST));
    if (query != null) {
      q.add(query, BooleanClause.Occur.MUST);
    }
    indexSearcher.search(q.build(),
        classQueryCountCollector);
    docCount = classQueryCountCollector.getTotalHits();
  } else {
    docCount = terms.getDocCount();
  }
  return docCount;
}
 

@Override
public FieldStats stats(Terms terms, int maxDoc) throws IOException {
    long minValue = NumericUtils.getMinLong(terms);
    long maxValue = NumericUtils.getMaxLong(terms);
    return new FieldStats.Date(
        maxDoc, terms.getDocCount(), terms.getSumDocFreq(), terms.getSumTotalTermFreq(), minValue, maxValue, dateTimeFormatter()
    );
}
 

@Override
public FieldStats stats(Terms terms, int maxDoc) throws IOException {
    long minValue = NumericUtils.getMinInt(terms);
    long maxValue = NumericUtils.getMaxInt(terms);
    return new FieldStats.Long(
        maxDoc, terms.getDocCount(), terms.getSumDocFreq(), terms.getSumTotalTermFreq(), minValue, maxValue
    );
}
 

@Override
public FieldStats stats(Terms terms, int maxDoc) throws IOException {
    float minValue = NumericUtils.sortableIntToFloat(NumericUtils.getMinInt(terms));
    float maxValue = NumericUtils.sortableIntToFloat(NumericUtils.getMaxInt(terms));
    return new FieldStats.Float(
        maxDoc, terms.getDocCount(), terms.getSumDocFreq(), terms.getSumTotalTermFreq(), minValue, maxValue
    );
}
 

@Override
public FieldStats stats(Terms terms, int maxDoc) throws IOException {
    double minValue = NumericUtils.sortableLongToDouble(NumericUtils.getMinLong(terms));
    double maxValue = NumericUtils.sortableLongToDouble(NumericUtils.getMaxLong(terms));
    return new FieldStats.Double(
        maxDoc, terms.getDocCount(), terms.getSumDocFreq(), terms.getSumTotalTermFreq(), minValue, maxValue
    );
}
 

@Override
public FieldStats stats(Terms terms, int maxDoc) throws IOException {
    long minValue = NumericUtils.getMinInt(terms);
    long maxValue = NumericUtils.getMaxInt(terms);
    return new FieldStats.Long(
        maxDoc, terms.getDocCount(), terms.getSumDocFreq(), terms.getSumTotalTermFreq(), minValue, maxValue
    );
}
 

@Override
public FieldStats stats(Terms terms, int maxDoc) throws IOException {
    long minValue = NumericUtils.getMinInt(terms);
    long maxValue = NumericUtils.getMaxInt(terms);
    return new FieldStats.Long(
        maxDoc, terms.getDocCount(), terms.getSumDocFreq(), terms.getSumTotalTermFreq(), minValue, maxValue
    );
}
 

@Override
public FieldStats stats(Terms terms, int maxDoc) throws IOException {
    long minValue = NumericUtils.getMinLong(terms);
    long maxValue = NumericUtils.getMaxLong(terms);
    return new FieldStats.Long(
        maxDoc, terms.getDocCount(), terms.getSumDocFreq(), terms.getSumTotalTermFreq(), minValue, maxValue
    );
}
 

/**
 * Returns the average number of unique terms times the number of docs belonging to the input class
 *
 * @param  term the class term
 * @return the average number of unique terms
 * @throws java.io.IOException If there is a low-level I/O error
 */
private double getTextTermFreqForClass(Term term, String fieldName) throws IOException {
  double avgNumberOfUniqueTerms;
  Terms terms = MultiTerms.getTerms(indexReader, fieldName);
  long numPostings = terms.getSumDocFreq(); // number of term/doc pairs
  avgNumberOfUniqueTerms = numPostings / (double) terms.getDocCount(); // avg # of unique terms per doc
  int docsWithC = indexReader.docFreq(term);
  return avgNumberOfUniqueTerms * docsWithC; // avg # of unique terms in text fields per doc * # docs with c
}
 

/**
 * Returns the average number of unique terms times the number of docs belonging to the input class
 * @param term the term representing the class
 * @return the average number of unique terms
 * @throws IOException if a low level I/O problem happens
 */
private double getTextTermFreqForClass(Term term) throws IOException {
  double avgNumberOfUniqueTerms = 0;
  for (String textFieldName : textFieldNames) {
    Terms terms = MultiTerms.getTerms(indexReader, textFieldName);
    long numPostings = terms.getSumDocFreq(); // number of term/doc pairs
    avgNumberOfUniqueTerms += numPostings / (double) terms.getDocCount(); // avg # of unique terms per doc
  }
  int docsWithC = indexReader.docFreq(term);
  return avgNumberOfUniqueTerms * docsWithC; // avg # of unique terms in text fields per doc * # docs with c
}
 

/**
 * @return a {@link FieldStats} instance that maps to the type of this field based on the provided {@link Terms} instance.
 */
public FieldStats stats(Terms terms, int maxDoc) throws IOException {
    return new FieldStats.Text(
        maxDoc, terms.getDocCount(), terms.getSumDocFreq(), terms.getSumTotalTermFreq(), terms.getMin(), terms.getMax()
    );
}
 

/** Create a {@link DocIdSetBuilder} instance that is optimized for
 *  accumulating docs that match the given {@link Terms}. */
public DocIdSetBuilder(int maxDoc, Terms terms) throws IOException {
  this(maxDoc, terms.getDocCount(), terms.getSumDocFreq());
}