Package org.apache.lucene.analysis

Examples of org.apache.lucene.analysis.CachingTokenFilter


    Analyzer analyzer = new MockAnalyzer(random());
    IndexWriter w = new IndexWriter(dir, newIndexWriterConfig( TEST_VERSION_CURRENT, analyzer));
    Document doc = new Document();
    TokenStream stream = analyzer.tokenStream("field", "abcd   ");
    stream.reset(); // TODO: weird to reset before wrapping with CachingTokenFilter... correct?
    stream = new CachingTokenFilter(stream);
    FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
    customType.setStoreTermVectors(true);
    customType.setStoreTermVectorPositions(true);
    customType.setStoreTermVectorOffsets(true);
    Field f = new Field("field", stream, customType);
View Full Code Here


        source = this.analyzer.tokenStream(field, text);
        source.reset();
      } catch (IOException e1) {
        throw new RuntimeException(e1);
      }
      CachingTokenFilter buffer = new CachingTokenFilter(source);

      PositionIncrementAttribute posIncrAtt = null;
      int numTokens = 0;
      int positionCount = 0;
      boolean severalTokensAtSamePosition = false;

      if (buffer.hasAttribute(PositionIncrementAttribute.class)) {
        posIncrAtt = buffer.getAttribute(PositionIncrementAttribute.class);
      }

      try {

        while (buffer.incrementToken()) {
          numTokens++;
          int positionIncrement = (posIncrAtt != null) ? posIncrAtt
              .getPositionIncrement() : 1;
          if (positionIncrement != 0) {
            positionCount += positionIncrement;

          } else {
            severalTokensAtSamePosition = true;
          }

        }

      } catch (IOException e) {
        // ignore
      }

      try {
        // rewind the buffer stream
        buffer.reset();

        // close original stream - all tokens buffered
        source.close();
      } catch (IOException e) {
        // ignore
      }

      if (!buffer.hasAttribute(CharTermAttribute.class)) {
        return new NoTokenFoundQueryNode();
      }

      CharTermAttribute termAtt = buffer.getAttribute(CharTermAttribute.class);

      if (numTokens == 0) {
        return new NoTokenFoundQueryNode();

      } else if (numTokens == 1) {
        String term = null;
        try {
          boolean hasNext;
          hasNext = buffer.incrementToken();
          assert hasNext == true;
          term = termAtt.toString();

        } catch (IOException e) {
          // safe to ignore, because we know the number of tokens
        }

        fieldNode.setText(term);

        return fieldNode;

      } else if (severalTokensAtSamePosition || !(node instanceof QuotedFieldQueryNode)) {
        if (positionCount == 1 || !(node instanceof QuotedFieldQueryNode)) {
          // no phrase query:
          LinkedList<QueryNode> children = new LinkedList<QueryNode>();

          for (int i = 0; i < numTokens; i++) {
            String term = null;
            try {
              boolean hasNext = buffer.incrementToken();
              assert hasNext == true;
              term = termAtt.toString();

            } catch (IOException e) {
              // safe to ignore, because we know the number of tokens
            }

            children.add(new FieldQueryNode(field, term, -1, -1));

          }
          return new GroupQueryNode(
            new StandardBooleanQueryNode(children, positionCount==1));
        } else {
          // phrase query:
          MultiPhraseQueryNode mpq = new MultiPhraseQueryNode();

          List<FieldQueryNode> multiTerms = new ArrayList<FieldQueryNode>();
          int position = -1;
          int i = 0;
          int termGroupCount = 0;
          for (; i < numTokens; i++) {
            String term = null;
            int positionIncrement = 1;
            try {
              boolean hasNext = buffer.incrementToken();
              assert hasNext == true;
              term = termAtt.toString();
              if (posIncrAtt != null) {
                positionIncrement = posIncrAtt.getPositionIncrement();
              }

            } catch (IOException e) {
              // safe to ignore, because we know the number of tokens
            }

            if (positionIncrement > 0 && multiTerms.size() > 0) {

              for (FieldQueryNode termNode : multiTerms) {

                if (this.positionIncrementsEnabled) {
                  termNode.setPositionIncrement(position);
                } else {
                  termNode.setPositionIncrement(termGroupCount);
                }

                mpq.add(termNode);

              }

              // Only increment once for each "group" of
              // terms that were in the same position:
              termGroupCount++;

              multiTerms.clear();

            }

            position += positionIncrement;
            multiTerms.add(new FieldQueryNode(field, term, -1, -1));

          }

          for (FieldQueryNode termNode : multiTerms) {

            if (this.positionIncrementsEnabled) {
              termNode.setPositionIncrement(position);

            } else {
              termNode.setPositionIncrement(termGroupCount);
            }

            mpq.add(termNode);

          }

          return mpq;

        }

      } else {

        TokenizedPhraseQueryNode pq = new TokenizedPhraseQueryNode();

        int position = -1;

        for (int i = 0; i < numTokens; i++) {
          String term = null;
          int positionIncrement = 1;

          try {
            boolean hasNext = buffer.incrementToken();
            assert hasNext == true;
            term = termAtt.toString();

            if (posIncrAtt != null) {
              positionIncrement = posIncrAtt.getPositionIncrement();
View Full Code Here

    } catch (IOException e) {
      ParseException p = new ParseException("Unable to initialize TokenStream to analyze query text");
      p.initCause(e);
      throw p;
    }
    CachingTokenFilter buffer = new CachingTokenFilter(source);
    TermToBytesRefAttribute termAtt = null;
    PositionIncrementAttribute posIncrAtt = null;
    int numTokens = 0;

    buffer.reset();

    if (buffer.hasAttribute(TermToBytesRefAttribute.class)) {
      termAtt = buffer.getAttribute(TermToBytesRefAttribute.class);
    }
    if (buffer.hasAttribute(PositionIncrementAttribute.class)) {
      posIncrAtt = buffer.getAttribute(PositionIncrementAttribute.class);
    }

    int positionCount = 0;
    boolean severalTokensAtSamePosition = false;

    boolean hasMoreTokens = false;
    if (termAtt != null) {
      try {
        hasMoreTokens = buffer.incrementToken();
        while (hasMoreTokens) {
          numTokens++;
          int positionIncrement = (posIncrAtt != null) ? posIncrAtt.getPositionIncrement() : 1;
          if (positionIncrement != 0) {
            positionCount += positionIncrement;
          } else {
            severalTokensAtSamePosition = true;
          }
          hasMoreTokens = buffer.incrementToken();
        }
      } catch (IOException e) {
        // ignore
      }
    }
    try {
      // rewind the buffer stream
      buffer.reset();

      // close original stream - all tokens buffered
      source.close();
    }
    catch (IOException e) {
      ParseException p = new ParseException("Cannot close TokenStream analyzing query text");
      p.initCause(e);
      throw p;
    }

    BytesRef bytes = termAtt == null ? null : termAtt.getBytesRef();

    if (numTokens == 0)
      return null;
    else if (numTokens == 1) {
      try {
        boolean hasNext = buffer.incrementToken();
        assert hasNext == true;
        termAtt.fillBytesRef();
      } catch (IOException e) {
        // safe to ignore, because we know the number of tokens
      }
      return newTermQuery(new Term(field, BytesRef.deepCopyOf(bytes)));
    } else {
      if (severalTokensAtSamePosition || (!quoted && !autoGeneratePhraseQueries)) {
        if (positionCount == 1 || (!quoted && !autoGeneratePhraseQueries)) {
          // no phrase query:
         
          if (positionCount == 1) {
            // simple case: only one position, with synonyms
            BooleanQuery q = newBooleanQuery(true);
            for (int i = 0; i < numTokens; i++) {
              try {
                boolean hasNext = buffer.incrementToken();
                assert hasNext == true;
                termAtt.fillBytesRef();
              } catch (IOException e) {
                // safe to ignore, because we know the number of tokens
              }
              Query currentQuery = newTermQuery(
                  new Term(field, BytesRef.deepCopyOf(bytes)));
              q.add(currentQuery, BooleanClause.Occur.SHOULD);
            }
            return q;
          } else {
            // multiple positions
            BooleanQuery q = newBooleanQuery(false);
            final BooleanClause.Occur occur = operator == Operator.AND ? BooleanClause.Occur.MUST : BooleanClause.Occur.SHOULD;
            Query currentQuery = null;
            for (int i = 0; i < numTokens; i++) {
              try {
                boolean hasNext = buffer.incrementToken();
                assert hasNext == true;
                termAtt.fillBytesRef();
              } catch (IOException e) {
                // safe to ignore, because we know the number of tokens
              }
              if (posIncrAtt != null && posIncrAtt.getPositionIncrement() == 0) {
                if (!(currentQuery instanceof BooleanQuery)) {
                  Query t = currentQuery;
                  currentQuery = newBooleanQuery(true);
                  ((BooleanQuery)currentQuery).add(t, BooleanClause.Occur.SHOULD);
                }
                ((BooleanQuery)currentQuery).add(newTermQuery(new Term(field, BytesRef.deepCopyOf(bytes))), BooleanClause.Occur.SHOULD);
              } else {
                if (currentQuery != null) {
                  q.add(currentQuery, occur);
                }
                currentQuery = newTermQuery(new Term(field, BytesRef.deepCopyOf(bytes)));
              }
            }
            q.add(currentQuery, occur);
            return q;
          }
        }
        else {
          // phrase query:
          MultiPhraseQuery mpq = newMultiPhraseQuery();
          mpq.setSlop(phraseSlop);
          List<Term> multiTerms = new ArrayList<Term>();
          int position = -1;
          for (int i = 0; i < numTokens; i++) {
            int positionIncrement = 1;
            try {
              boolean hasNext = buffer.incrementToken();
              assert hasNext == true;
              termAtt.fillBytesRef();
              if (posIncrAtt != null) {
                positionIncrement = posIncrAtt.getPositionIncrement();
              }
            } catch (IOException e) {
              // safe to ignore, because we know the number of tokens
            }

            if (positionIncrement > 0 && multiTerms.size() > 0) {
              if (enablePositionIncrements) {
                mpq.add(multiTerms.toArray(new Term[0]),position);
              } else {
                mpq.add(multiTerms.toArray(new Term[0]));
              }
              multiTerms.clear();
            }
            position += positionIncrement;
            multiTerms.add(new Term(field, BytesRef.deepCopyOf(bytes)));
          }
          if (enablePositionIncrements) {
            mpq.add(multiTerms.toArray(new Term[0]),position);
          } else {
            mpq.add(multiTerms.toArray(new Term[0]));
          }
          return mpq;
        }
      }
      else {
        PhraseQuery pq = newPhraseQuery();
        pq.setSlop(phraseSlop);
        int position = -1;

        for (int i = 0; i < numTokens; i++) {
          int positionIncrement = 1;

          try {
            boolean hasNext = buffer.incrementToken();
            assert hasNext == true;
            termAtt.fillBytesRef();
            if (posIncrAtt != null) {
              positionIncrement = posIncrAtt.getPositionIncrement();
            }
View Full Code Here

  protected AtomicReaderContext getLeafContext() throws IOException {
    if (internalReader == null) {
      if(wrapToCaching && !(tokenStream instanceof CachingTokenFilter)) {
        assert !cachedTokenStream;
        tokenStream = new CachingTokenFilter(new OffsetLimitTokenFilter(tokenStream, maxDocCharsToAnalyze));
        cachedTokenStream = true;
      }
      final MemoryIndex indexer = new MemoryIndex(true);
      indexer.addField(DelegatingAtomicReader.FIELD_NAME, tokenStream);
      tokenStream.reset();
View Full Code Here

        source = this.analyzer.tokenStream(field, new StringReader(text));
        source.reset();
      } catch (IOException e1) {
        throw new RuntimeException(e1);
      }
      CachingTokenFilter buffer = new CachingTokenFilter(source);

      PositionIncrementAttribute posIncrAtt = null;
      int numTokens = 0;
      int positionCount = 0;
      boolean severalTokensAtSamePosition = false;

      if (buffer.hasAttribute(PositionIncrementAttribute.class)) {
        posIncrAtt = buffer.getAttribute(PositionIncrementAttribute.class);
      }

      try {

        while (buffer.incrementToken()) {
          numTokens++;
          int positionIncrement = (posIncrAtt != null) ? posIncrAtt
              .getPositionIncrement() : 1;
          if (positionIncrement != 0) {
            positionCount += positionIncrement;

          } else {
            severalTokensAtSamePosition = true;
          }

        }

      } catch (IOException e) {
        // ignore
      }

      try {
        // rewind the buffer stream
        buffer.reset();

        // close original stream - all tokens buffered
        source.close();
      } catch (IOException e) {
        // ignore
      }

      if (!buffer.hasAttribute(CharTermAttribute.class)) {
        return new NoTokenFoundQueryNode();
      }

      CharTermAttribute termAtt = buffer.getAttribute(CharTermAttribute.class);

      if (numTokens == 0) {
        return new NoTokenFoundQueryNode();

      } else if (numTokens == 1) {
        String term = null;
        try {
          boolean hasNext;
          hasNext = buffer.incrementToken();
          assert hasNext == true;
          term = termAtt.toString();

        } catch (IOException e) {
          // safe to ignore, because we know the number of tokens
        }

        fieldNode.setText(term);

        return fieldNode;

      } else if (severalTokensAtSamePosition || !(node instanceof QuotedFieldQueryNode)) {
        if (positionCount == 1 || !(node instanceof QuotedFieldQueryNode)) {
          // no phrase query:
          LinkedList<QueryNode> children = new LinkedList<QueryNode>();

          for (int i = 0; i < numTokens; i++) {
            String term = null;
            try {
              boolean hasNext = buffer.incrementToken();
              assert hasNext == true;
              term = termAtt.toString();

            } catch (IOException e) {
              // safe to ignore, because we know the number of tokens
            }

            children.add(new FieldQueryNode(field, term, -1, -1));

          }
          return new GroupQueryNode(
            new StandardBooleanQueryNode(children, positionCount==1));
        } else {
          // phrase query:
          MultiPhraseQueryNode mpq = new MultiPhraseQueryNode();

          List<FieldQueryNode> multiTerms = new ArrayList<FieldQueryNode>();
          int position = -1;
          int i = 0;
          int termGroupCount = 0;
          for (; i < numTokens; i++) {
            String term = null;
            int positionIncrement = 1;
            try {
              boolean hasNext = buffer.incrementToken();
              assert hasNext == true;
              term = termAtt.toString();
              if (posIncrAtt != null) {
                positionIncrement = posIncrAtt.getPositionIncrement();
              }

            } catch (IOException e) {
              // safe to ignore, because we know the number of tokens
            }

            if (positionIncrement > 0 && multiTerms.size() > 0) {

              for (FieldQueryNode termNode : multiTerms) {

                if (this.positionIncrementsEnabled) {
                  termNode.setPositionIncrement(position);
                } else {
                  termNode.setPositionIncrement(termGroupCount);
                }

                mpq.add(termNode);

              }

              // Only increment once for each "group" of
              // terms that were in the same position:
              termGroupCount++;

              multiTerms.clear();

            }

            position += positionIncrement;
            multiTerms.add(new FieldQueryNode(field, term, -1, -1));

          }

          for (FieldQueryNode termNode : multiTerms) {

            if (this.positionIncrementsEnabled) {
              termNode.setPositionIncrement(position);

            } else {
              termNode.setPositionIncrement(termGroupCount);
            }

            mpq.add(termNode);

          }

          return mpq;

        }

      } else {

        TokenizedPhraseQueryNode pq = new TokenizedPhraseQueryNode();

        int position = -1;

        for (int i = 0; i < numTokens; i++) {
          String term = null;
          int positionIncrement = 1;

          try {
            boolean hasNext = buffer.incrementToken();
            assert hasNext == true;
            term = termAtt.toString();

            if (posIncrAtt != null) {
              positionIncrement = posIncrAtt.getPositionIncrement();
View Full Code Here

    } catch (IOException e) {
      ParseException p = new ParseException("Unable to initialize TokenStream to analyze query text");
      p.initCause(e);
      throw p;
    }
    CachingTokenFilter buffer = new CachingTokenFilter(source);
    TermToBytesRefAttribute termAtt = null;
    PositionIncrementAttribute posIncrAtt = null;
    int numTokens = 0;

    buffer.reset();

    if (buffer.hasAttribute(TermToBytesRefAttribute.class)) {
      termAtt = buffer.getAttribute(TermToBytesRefAttribute.class);
    }
    if (buffer.hasAttribute(PositionIncrementAttribute.class)) {
      posIncrAtt = buffer.getAttribute(PositionIncrementAttribute.class);
    }

    int positionCount = 0;
    boolean severalTokensAtSamePosition = false;

    boolean hasMoreTokens = false;
    if (termAtt != null) {
      try {
        hasMoreTokens = buffer.incrementToken();
        while (hasMoreTokens) {
          numTokens++;
          int positionIncrement = (posIncrAtt != null) ? posIncrAtt.getPositionIncrement() : 1;
          if (positionIncrement != 0) {
            positionCount += positionIncrement;
          } else {
            severalTokensAtSamePosition = true;
          }
          hasMoreTokens = buffer.incrementToken();
        }
      } catch (IOException e) {
        // ignore
      }
    }
    try {
      // rewind the buffer stream
      buffer.reset();

      // close original stream - all tokens buffered
      source.close();
    }
    catch (IOException e) {
      ParseException p = new ParseException("Cannot close TokenStream analyzing query text");
      p.initCause(e);
      throw p;
    }

    BytesRef bytes = termAtt == null ? null : termAtt.getBytesRef();

    if (numTokens == 0)
      return null;
    else if (numTokens == 1) {
      try {
        boolean hasNext = buffer.incrementToken();
        assert hasNext == true;
        termAtt.fillBytesRef();
      } catch (IOException e) {
        // safe to ignore, because we know the number of tokens
      }
      return newTermQuery(new Term(field, BytesRef.deepCopyOf(bytes)));
    } else {
      if (severalTokensAtSamePosition || (!quoted && !autoGeneratePhraseQueries)) {
        if (positionCount == 1 || (!quoted && !autoGeneratePhraseQueries)) {
          // no phrase query:
          BooleanQuery q = newBooleanQuery(positionCount == 1);

          BooleanClause.Occur occur = positionCount > 1 && operator == AND_OPERATOR ?
            BooleanClause.Occur.MUST : BooleanClause.Occur.SHOULD;

          for (int i = 0; i < numTokens; i++) {
            try {
              boolean hasNext = buffer.incrementToken();
              assert hasNext == true;
              termAtt.fillBytesRef();
            } catch (IOException e) {
              // safe to ignore, because we know the number of tokens
            }
            Query currentQuery = newTermQuery(
                new Term(field, BytesRef.deepCopyOf(bytes)));
            q.add(currentQuery, occur);
          }
          return q;
        }
        else {
          // phrase query:
          MultiPhraseQuery mpq = newMultiPhraseQuery();
          mpq.setSlop(phraseSlop);
          List<Term> multiTerms = new ArrayList<Term>();
          int position = -1;
          for (int i = 0; i < numTokens; i++) {
            int positionIncrement = 1;
            try {
              boolean hasNext = buffer.incrementToken();
              assert hasNext == true;
              termAtt.fillBytesRef();
              if (posIncrAtt != null) {
                positionIncrement = posIncrAtt.getPositionIncrement();
              }
            } catch (IOException e) {
              // safe to ignore, because we know the number of tokens
            }

            if (positionIncrement > 0 && multiTerms.size() > 0) {
              if (enablePositionIncrements) {
                mpq.add(multiTerms.toArray(new Term[0]),position);
              } else {
                mpq.add(multiTerms.toArray(new Term[0]));
              }
              multiTerms.clear();
            }
            position += positionIncrement;
            multiTerms.add(new Term(field, BytesRef.deepCopyOf(bytes)));
          }
          if (enablePositionIncrements) {
            mpq.add(multiTerms.toArray(new Term[0]),position);
          } else {
            mpq.add(multiTerms.toArray(new Term[0]));
          }
          return mpq;
        }
      }
      else {
        PhraseQuery pq = newPhraseQuery();
        pq.setSlop(phraseSlop);
        int position = -1;

        for (int i = 0; i < numTokens; i++) {
          int positionIncrement = 1;

          try {
            boolean hasNext = buffer.incrementToken();
            assert hasNext == true;
            termAtt.fillBytesRef();
            if (posIncrAtt != null) {
              positionIncrement = posIncrAtt.getPositionIncrement();
            }
View Full Code Here

    Analyzer analyzer = new MockAnalyzer(random());
    IndexWriter w = new IndexWriter(dir, newIndexWriterConfig( TEST_VERSION_CURRENT, analyzer));
    Document doc = new Document();
    TokenStream stream = analyzer.tokenStream("field", new StringReader("abcd   "));
    stream.reset(); // TODO: weird to reset before wrapping with CachingTokenFilter... correct?
    stream = new CachingTokenFilter(stream);
    FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
    customType.setStoreTermVectors(true);
    customType.setStoreTermVectorPositions(true);
    customType.setStoreTermVectorOffsets(true);
    Field f = new Field("field", stream, customType);
View Full Code Here

    Document doc = new Document();
    IOException priorException = null;
    TokenStream stream = analyzer.tokenStream("field", "abcd   ");
    try {
      stream.reset(); // TODO: weird to reset before wrapping with CachingTokenFilter... correct?
      TokenStream cachedStream = new CachingTokenFilter(stream);
      FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
      customType.setStoreTermVectors(true);
      customType.setStoreTermVectorPositions(true);
      customType.setStoreTermVectorOffsets(true);
      Field f = new Field("field", cachedStream, customType);
View Full Code Here

  protected AtomicReaderContext getLeafContext() throws IOException {
    if (internalReader == null) {
      if(wrapToCaching && !(tokenStream instanceof CachingTokenFilter)) {
        assert !cachedTokenStream;
        tokenStream = new CachingTokenFilter(new OffsetLimitTokenFilter(tokenStream, maxDocCharsToAnalyze));
        cachedTokenStream = true;
      }
      final MemoryIndex indexer = new MemoryIndex(true);
      indexer.addField(DelegatingAtomicReader.FIELD_NAME, tokenStream);
      tokenStream.reset();
View Full Code Here

      FieldQueryNode fieldNode = ((FieldQueryNode) node);
      String text = fieldNode.getTextAsString();
      String field = fieldNode.getFieldAsString();

      CachingTokenFilter buffer = null;
      PositionIncrementAttribute posIncrAtt = null;
      int numTokens = 0;
      int positionCount = 0;
      boolean severalTokensAtSamePosition = false;
     
      TokenStream source = null;
      try {
        source = this.analyzer.tokenStream(field, text);
        source.reset();
        buffer = new CachingTokenFilter(source);

        if (buffer.hasAttribute(PositionIncrementAttribute.class)) {
          posIncrAtt = buffer.getAttribute(PositionIncrementAttribute.class);
        }

        try {

          while (buffer.incrementToken()) {
            numTokens++;
            int positionIncrement = (posIncrAtt != null) ? posIncrAtt
                .getPositionIncrement() : 1;
            if (positionIncrement != 0) {
              positionCount += positionIncrement;

            } else {
              severalTokensAtSamePosition = true;
            }

          }

        } catch (IOException e) {
          // ignore
        }
      } catch (IOException e) {
        throw new RuntimeException(e);
      } finally {
        IOUtils.closeWhileHandlingException(source);
      }
     
      // rewind the buffer stream
      buffer.reset();

      if (!buffer.hasAttribute(CharTermAttribute.class)) {
        return new NoTokenFoundQueryNode();
      }

      CharTermAttribute termAtt = buffer.getAttribute(CharTermAttribute.class);

      if (numTokens == 0) {
        return new NoTokenFoundQueryNode();

      } else if (numTokens == 1) {
        String term = null;
        try {
          boolean hasNext;
          hasNext = buffer.incrementToken();
          assert hasNext == true;
          term = termAtt.toString();

        } catch (IOException e) {
          // safe to ignore, because we know the number of tokens
        }

        fieldNode.setText(term);

        return fieldNode;

      } else if (severalTokensAtSamePosition || !(node instanceof QuotedFieldQueryNode)) {
        if (positionCount == 1 || !(node instanceof QuotedFieldQueryNode)) {
          // no phrase query:
         
          if (positionCount == 1) {
            // simple case: only one position, with synonyms
            LinkedList<QueryNode> children = new LinkedList<>();
           
            for (int i = 0; i < numTokens; i++) {
              String term = null;
              try {
                boolean hasNext = buffer.incrementToken();
                assert hasNext == true;
                term = termAtt.toString();
               
              } catch (IOException e) {
                // safe to ignore, because we know the number of tokens
              }
             
              children.add(new FieldQueryNode(field, term, -1, -1));
             
            }
            return new GroupQueryNode(
                new StandardBooleanQueryNode(children, positionCount==1));
           
          } else {
            // multiple positions
            QueryNode q = new StandardBooleanQueryNode(Collections.<QueryNode>emptyList(),false);
            QueryNode currentQuery = null;
            for (int i = 0; i < numTokens; i++) {
              String term = null;
              try {
                boolean hasNext = buffer.incrementToken();
                assert hasNext == true;
                term = termAtt.toString();
              } catch (IOException e) {
                // safe to ignore, because we know the number of tokens
              }
              if (posIncrAtt != null && posIncrAtt.getPositionIncrement() == 0) {
                if (!(currentQuery instanceof BooleanQueryNode)) {
                  QueryNode t = currentQuery;
                  currentQuery = new StandardBooleanQueryNode(Collections.<QueryNode>emptyList(), true);
                  ((BooleanQueryNode)currentQuery).add(t);
                }
                ((BooleanQueryNode)currentQuery).add(new FieldQueryNode(field, term, -1, -1));
              } else {
                if (currentQuery != null) {
                  if (this.defaultOperator == Operator.OR) {
                    q.add(currentQuery);
                  } else {
                    q.add(new ModifierQueryNode(currentQuery, Modifier.MOD_REQ));
                  }
                }
                currentQuery = new FieldQueryNode(field, term, -1, -1);
              }
            }
            if (this.defaultOperator == Operator.OR) {
              q.add(currentQuery);
            } else {
              q.add(new ModifierQueryNode(currentQuery, Modifier.MOD_REQ));
            }
           
            if (q instanceof BooleanQueryNode) {
              q = new GroupQueryNode(q);
            }
            return q;
          }
        } else {
          // phrase query:
          MultiPhraseQueryNode mpq = new MultiPhraseQueryNode();

          List<FieldQueryNode> multiTerms = new ArrayList<>();
          int position = -1;
          int i = 0;
          int termGroupCount = 0;
          for (; i < numTokens; i++) {
            String term = null;
            int positionIncrement = 1;
            try {
              boolean hasNext = buffer.incrementToken();
              assert hasNext == true;
              term = termAtt.toString();
              if (posIncrAtt != null) {
                positionIncrement = posIncrAtt.getPositionIncrement();
              }

            } catch (IOException e) {
              // safe to ignore, because we know the number of tokens
            }

            if (positionIncrement > 0 && multiTerms.size() > 0) {

              for (FieldQueryNode termNode : multiTerms) {

                if (this.positionIncrementsEnabled) {
                  termNode.setPositionIncrement(position);
                } else {
                  termNode.setPositionIncrement(termGroupCount);
                }

                mpq.add(termNode);

              }

              // Only increment once for each "group" of
              // terms that were in the same position:
              termGroupCount++;

              multiTerms.clear();

            }

            position += positionIncrement;
            multiTerms.add(new FieldQueryNode(field, term, -1, -1));

          }

          for (FieldQueryNode termNode : multiTerms) {

            if (this.positionIncrementsEnabled) {
              termNode.setPositionIncrement(position);

            } else {
              termNode.setPositionIncrement(termGroupCount);
            }

            mpq.add(termNode);

          }

          return mpq;

        }

      } else {

        TokenizedPhraseQueryNode pq = new TokenizedPhraseQueryNode();

        int position = -1;

        for (int i = 0; i < numTokens; i++) {
          String term = null;
          int positionIncrement = 1;

          try {
            boolean hasNext = buffer.incrementToken();
            assert hasNext == true;
            term = termAtt.toString();

            if (posIncrAtt != null) {
              positionIncrement = posIncrAtt.getPositionIncrement();
View Full Code Here

TOP

Related Classes of org.apache.lucene.analysis.CachingTokenFilter

Copyright © 2018 www.massapicom. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.