Java PhraseQuery.setBoost方法代码示例

import org.apache.lucene.search.PhraseQuery; //导入方法依赖的package包/类
private Query applySlop(Query q, int slop) {
    if (q instanceof PhraseQuery) {
        PhraseQuery pq = (PhraseQuery) q;
        PhraseQuery.Builder builder = new PhraseQuery.Builder();
        builder.setSlop(slop);
        final Term[] terms = pq.getTerms();
        final int[] positions = pq.getPositions();
        for (int i = 0; i < terms.length; ++i) {
            builder.add(terms[i], positions[i]);
        }
        pq = builder.build();
        pq.setBoost(q.getBoost());
        return pq;
    } else if (q instanceof MultiPhraseQuery) {
        ((MultiPhraseQuery) q).setSlop(slop);
        return q;
    } else {
        return q;
    }
}
 

import org.apache.lucene.search.PhraseQuery; //导入方法依赖的package包/类
private void checkOverlap( Collection<Query> expandQueries, Term[] src, Term[] dest, int slop, float boost ){
  // beginning from 1 (not 0) is safe because that the PhraseQuery has multiple terms
  // is guaranteed in flatten() method (if PhraseQuery has only one term, flatten()
  // converts PhraseQuery to TermQuery)
  for( int i = 1; i < src.length; i++ ){
    boolean overlap = true;
    for( int j = i; j < src.length; j++ ){
      if( ( j - i ) < dest.length && !src[j].text().equals( dest[j-i].text() ) ){
        overlap = false;
        break;
      }
    }
    if( overlap && src.length - i < dest.length ){
      PhraseQuery pq = new PhraseQuery();
      for( Term srcTerm : src )
        pq.add( srcTerm );
      for( int k = src.length - i; k < dest.length; k++ ){
        pq.add( new Term( src[0].field(), dest[k].text() ) );
      }
      pq.setSlop( slop );
      pq.setBoost( boost );
      if(!expandQueries.contains( pq ) )
        expandQueries.add( pq );
    }
  }
}
 

import org.apache.lucene.search.PhraseQuery; //导入方法依赖的package包/类
protected Query pq( float boost, int slop, String field, String... texts ){
  PhraseQuery query = new PhraseQuery();
  for( String text : texts ){
    query.add( new Term( field, text ) );
  }
  query.setBoost( boost );
  query.setSlop( slop );
  return query;
}
 

import org.apache.lucene.search.PhraseQuery; //导入方法依赖的package包/类
public void testCJKBoostedPhrase() throws Exception {
  // individual CJK chars as terms
  SimpleCJKAnalyzer analyzer = new SimpleCJKAnalyzer();
  
  PhraseQuery expected = new PhraseQuery();
  expected.setBoost(0.5f);
  expected.add(new Term("field", "中"));
  expected.add(new Term("field", "国"));
  
  assertEquals(expected, getQuery("\"中国\"^0.5", analyzer));
}
 

import org.apache.lucene.search.PhraseQuery; //导入方法依赖的package包/类
/**
 * Rewrite a term query. The base class rewrites the term itself.
 *
 * @param q  The query to rewrite
 * @return   Rewritten version, or 'q' unchanged if no change needed.
 */
protected Query rewrite(TermQuery q) 
{
  Term oldTerm = q.getTerm();
  Term newTerm = rewrite(oldTerm);
  
  // If the term is unchanged, don't change the query.
  if (oldTerm == newTerm)
    return q;
  
  // If the term is going away, inform the caller
  if (newTerm == null)
    return null;
  
  // If the term is splitting, make it into two terms.
  int spacePos = newTerm.text().indexOf(' ');
  if (oldTerm.text().indexOf(' ') < 0 && spacePos >= 0) {
    PhraseQuery pq = new PhraseQuery();
    pq.add(new Term(newTerm.field(), newTerm.text().substring(0, spacePos)));
    pq.add(new Term(newTerm.field(), newTerm.text().substring(spacePos+1)));
    pq.setBoost(q.getBoost());
    return pq;
  }
  
  // Make a new query for the new term.
  TermQuery newQuery = new TermQuery(newTerm);
  newQuery.setBoost(q.getBoost());
  return newQuery;
}
 

import org.apache.lucene.search.PhraseQuery; //导入方法依赖的package包/类
public static BooleanQuery buildQuery(String text, String field) throws IOException {
    BooleanQuery fullQuery = new BooleanQuery();
    TokenStream tokStream = ANALYZER.tokenStream(field, text);
    try {
        tokStream.reset();
        Term lastTerm = null;
        while (tokStream.incrementToken()) {
            String termStr = tokStream.getAttribute(CharTermAttribute.class).toString();

            Term term = new Term(field, termStr);
            TermQuery termQuery = new TermQuery(term);
            termQuery.setBoost(0.8f);
            fullQuery.add(termQuery, BooleanClause.Occur.SHOULD);

            List<String> expansionWords = WordNetExpansion.getStrictExpansion(termStr);
            if (expansionWords != null) {
                for (String expTermStr : expansionWords) {
                    TermQuery expTermQuery = new TermQuery(new Term(field, expTermStr));
                    expTermQuery.setBoost(0.4f);
                    fullQuery.add(expTermQuery, BooleanClause.Occur.SHOULD);
                }
            }

            if (lastTerm != null) {
                PhraseQuery bigram = new PhraseQuery();
                bigram.add(lastTerm);
                bigram.add(term);
                bigram.setBoost(0.1f);
                fullQuery.add(bigram, BooleanClause.Occur.SHOULD);

                if (USE_PROXIMITY_QUERY) {
                    SpanNearQuery spanNearQuery = new SpanNearQuery(new SpanQuery[]{
                            new SpanTermQuery(lastTerm),
                            new SpanTermQuery(term)},
                            6,
                            false);
                    spanNearQuery.setBoost(0.1f);
                    fullQuery.add(spanNearQuery, BooleanClause.Occur.SHOULD);
                }
            }
            lastTerm = term;
        }
        tokStream.end();
    } finally {
        tokStream.close();
    }
    return fullQuery;
}