Examples with SpanTermQuery org.apache.lucene.search.spans.SpanTermQuery used on opensource projects

Example 91 with SpanTermQuery

use of org.apache.lucene.search.spans.SpanTermQuery in project Anserini by castorini.

the class SdmQueryGenerator method buildQuery.

/*
  * Sequential Dependency Model
  */
@Override
public Query buildQuery(String field, Analyzer analyzer, String queryText) {
    List<String> tokens = AnalyzerUtils.analyze(analyzer, queryText);
    BooleanQuery.Builder termsBuilder = new BooleanQuery.Builder();
    if (tokens.size() == 1) {
        termsBuilder.add(new TermQuery(new Term(field, tokens.get(0))), BooleanClause.Occur.SHOULD);
        return termsBuilder.build();
    }
    BooleanQuery.Builder orderedWindowBuilder = new BooleanQuery.Builder();
    BooleanQuery.Builder unorderedWindowBuilder = new BooleanQuery.Builder();
    for (int i = 0; i < tokens.size() - 1; i++) {
        termsBuilder.add(new TermQuery(new Term(field, tokens.get(i))), BooleanClause.Occur.SHOULD);
        SpanTermQuery t1 = new SpanTermQuery(new Term(field, tokens.get(i)));
        SpanTermQuery t2 = new SpanTermQuery(new Term(field, tokens.get(i + 1)));
        SpanNearQuery orderedQ = new SpanNearQuery(new SpanQuery[] { t1, t2 }, 1, true);
        SpanNearQuery unorderedQ = new SpanNearQuery(new SpanQuery[] { t1, t2 }, 8, false);
        orderedWindowBuilder.add(orderedQ, BooleanClause.Occur.SHOULD);
        unorderedWindowBuilder.add(unorderedQ, BooleanClause.Occur.SHOULD);
    }
    termsBuilder.add(new TermQuery(new Term(field, tokens.get(tokens.size() - 1))), BooleanClause.Occur.SHOULD);
    BooleanQuery.Builder builder = new BooleanQuery.Builder();
    builder.add(new BoostQuery(termsBuilder.build(), termWeight), BooleanClause.Occur.SHOULD);
    builder.add(new BoostQuery(orderedWindowBuilder.build(), orderWindowWeight), BooleanClause.Occur.SHOULD);
    builder.add(new BoostQuery(unorderedWindowBuilder.build(), unorderWindowWeight), BooleanClause.Occur.SHOULD);
    return builder.build();
}

Example 92 with SpanTermQuery

use of org.apache.lucene.search.spans.SpanTermQuery in project elasticsearch by elastic.

the class CustomFieldQuery method flatten.

@Override
void flatten(Query sourceQuery, IndexReader reader, Collection<Query> flatQueries, float boost) throws IOException {
    if (sourceQuery instanceof BoostQuery) {
        BoostQuery bq = (BoostQuery) sourceQuery;
        sourceQuery = bq.getQuery();
        boost *= bq.getBoost();
        flatten(sourceQuery, reader, flatQueries, boost);
    } else if (sourceQuery instanceof SpanTermQuery) {
        super.flatten(new TermQuery(((SpanTermQuery) sourceQuery).getTerm()), reader, flatQueries, boost);
    } else if (sourceQuery instanceof ConstantScoreQuery) {
        flatten(((ConstantScoreQuery) sourceQuery).getQuery(), reader, flatQueries, boost);
    } else if (sourceQuery instanceof FunctionScoreQuery) {
        flatten(((FunctionScoreQuery) sourceQuery).getSubQuery(), reader, flatQueries, boost);
    } else if (sourceQuery instanceof MultiPhrasePrefixQuery) {
        flatten(sourceQuery.rewrite(reader), reader, flatQueries, boost);
    } else if (sourceQuery instanceof FiltersFunctionScoreQuery) {
        flatten(((FiltersFunctionScoreQuery) sourceQuery).getSubQuery(), reader, flatQueries, boost);
    } else if (sourceQuery instanceof MultiPhraseQuery) {
        MultiPhraseQuery q = ((MultiPhraseQuery) sourceQuery);
        convertMultiPhraseQuery(0, new int[q.getTermArrays().length], q, q.getTermArrays(), q.getPositions(), reader, flatQueries);
    } else if (sourceQuery instanceof BlendedTermQuery) {
        final BlendedTermQuery blendedTermQuery = (BlendedTermQuery) sourceQuery;
        flatten(blendedTermQuery.rewrite(reader), reader, flatQueries, boost);
    } else if (sourceQuery instanceof ESToParentBlockJoinQuery) {
        ESToParentBlockJoinQuery blockJoinQuery = (ESToParentBlockJoinQuery) sourceQuery;
        flatten(blockJoinQuery.getChildQuery(), reader, flatQueries, boost);
    } else if (sourceQuery instanceof BoostingQuery) {
        BoostingQuery boostingQuery = (BoostingQuery) sourceQuery;
        //flatten positive query with query boost
        flatten(boostingQuery.getMatch(), reader, flatQueries, boost);
        //flatten negative query with negative boost
        flatten(boostingQuery.getContext(), reader, flatQueries, boostingQuery.getBoost());
    } else if (sourceQuery instanceof SynonymQuery) {
        // SynonymQuery should be handled by the parent class directly.
        // This statement should be removed when https://issues.apache.org/jira/browse/LUCENE-7484 is merged.
        SynonymQuery synQuery = (SynonymQuery) sourceQuery;
        for (Term term : synQuery.getTerms()) {
            flatten(new TermQuery(term), reader, flatQueries, boost);
        }
    } else {
        super.flatten(sourceQuery, reader, flatQueries, boost);
    }
}

Example 93 with SpanTermQuery

use of org.apache.lucene.search.spans.SpanTermQuery in project elasticsearch by elastic.

the class QueryStringQueryBuilderTests method testToQueryWithGraph.

public void testToQueryWithGraph() throws Exception {
    assumeTrue("test runs only when at least a type is registered", getCurrentTypes().length > 0);
    for (Operator op : Operator.values()) {
        BooleanClause.Occur defaultOp = op.toBooleanClauseOccur();
        MapperQueryParser queryParser = new MapperQueryParser(createShardContext());
        QueryParserSettings settings = new QueryParserSettings("");
        settings.defaultField(STRING_FIELD_NAME);
        settings.fieldsAndWeights(Collections.emptyMap());
        settings.fuzziness(Fuzziness.AUTO);
        settings.analyzeWildcard(true);
        settings.rewriteMethod(MultiTermQuery.CONSTANT_SCORE_REWRITE);
        settings.defaultOperator(op.toQueryParserOperator());
        settings.forceAnalyzer(new MockSynonymAnalyzer());
        settings.forceQuoteAnalyzer(new MockSynonymAnalyzer());
        queryParser.reset(settings);
        // simple multi-term
        Query query = queryParser.parse("guinea pig");
        Query expectedQuery = new BooleanQuery.Builder().add(new BooleanQuery.Builder().add(new TermQuery(new Term(STRING_FIELD_NAME, "guinea")), Occur.MUST).add(new TermQuery(new Term(STRING_FIELD_NAME, "pig")), Occur.MUST).build(), defaultOp).add(new TermQuery(new Term(STRING_FIELD_NAME, "cavy")), defaultOp).build();
        assertThat(query, Matchers.equalTo(expectedQuery));
        // simple with additional tokens
        query = queryParser.parse("that guinea pig smells");
        expectedQuery = new BooleanQuery.Builder().add(new TermQuery(new Term(STRING_FIELD_NAME, "that")), defaultOp).add(new BooleanQuery.Builder().add(new BooleanQuery.Builder().add(new BooleanClause(new TermQuery(new Term(STRING_FIELD_NAME, "guinea")), Occur.MUST)).add(new BooleanClause(new TermQuery(new Term(STRING_FIELD_NAME, "pig")), Occur.MUST)).build(), Occur.SHOULD).add(new TermQuery(new Term(STRING_FIELD_NAME, "cavy")), Occur.SHOULD).build(), defaultOp).add(new TermQuery(new Term(STRING_FIELD_NAME, "smells")), defaultOp).build();
        assertThat(query, Matchers.equalTo(expectedQuery));
        // complex
        query = queryParser.parse("+that -(guinea pig) +smells");
        expectedQuery = new BooleanQuery.Builder().add(new TermQuery(new Term(STRING_FIELD_NAME, "that")), Occur.MUST).add(new BooleanQuery.Builder().add(new BooleanQuery.Builder().add(new TermQuery(new Term(STRING_FIELD_NAME, "guinea")), Occur.MUST).add(new TermQuery(new Term(STRING_FIELD_NAME, "pig")), Occur.MUST).build(), defaultOp).add(new TermQuery(new Term(STRING_FIELD_NAME, "cavy")), defaultOp).build(), Occur.MUST_NOT).add(new TermQuery(new Term(STRING_FIELD_NAME, "smells")), Occur.MUST).build();
        assertThat(query, Matchers.equalTo(expectedQuery));
        // no parent should cause guinea and pig to be treated as separate tokens
        query = queryParser.parse("+that -guinea pig +smells");
        expectedQuery = new BooleanQuery.Builder().add(new TermQuery(new Term(STRING_FIELD_NAME, "that")), BooleanClause.Occur.MUST).add(new TermQuery(new Term(STRING_FIELD_NAME, "guinea")), BooleanClause.Occur.MUST_NOT).add(new TermQuery(new Term(STRING_FIELD_NAME, "pig")), defaultOp).add(new TermQuery(new Term(STRING_FIELD_NAME, "smells")), BooleanClause.Occur.MUST).build();
        assertThat(query, Matchers.equalTo(expectedQuery));
        // span query
        query = queryParser.parse("\"that guinea pig smells\"");
        expectedQuery = new BooleanQuery.Builder().add(new SpanNearQuery.Builder(STRING_FIELD_NAME, true).addClause(new SpanTermQuery(new Term(STRING_FIELD_NAME, "that"))).addClause(new SpanOrQuery(new SpanNearQuery.Builder(STRING_FIELD_NAME, true).addClause(new SpanTermQuery(new Term(STRING_FIELD_NAME, "guinea"))).addClause(new SpanTermQuery(new Term(STRING_FIELD_NAME, "pig"))).build(), new SpanTermQuery(new Term(STRING_FIELD_NAME, "cavy")))).addClause(new SpanTermQuery(new Term(STRING_FIELD_NAME, "smells"))).build(), Occur.SHOULD).setDisableCoord(true).build();
        assertThat(query, Matchers.equalTo(expectedQuery));
        // span query with slop
        query = queryParser.parse("\"that guinea pig smells\"~2");
        expectedQuery = new BooleanQuery.Builder().add(new SpanNearQuery.Builder(STRING_FIELD_NAME, true).addClause(new SpanTermQuery(new Term(STRING_FIELD_NAME, "that"))).addClause(new SpanOrQuery(new SpanNearQuery.Builder(STRING_FIELD_NAME, true).addClause(new SpanTermQuery(new Term(STRING_FIELD_NAME, "guinea"))).addClause(new SpanTermQuery(new Term(STRING_FIELD_NAME, "pig"))).build(), new SpanTermQuery(new Term(STRING_FIELD_NAME, "cavy")))).addClause(new SpanTermQuery(new Term(STRING_FIELD_NAME, "smells"))).setSlop(2).build(), Occur.SHOULD).setDisableCoord(true).build();
        assertThat(query, Matchers.equalTo(expectedQuery));
    }
}

Example 94 with SpanTermQuery

use of org.apache.lucene.search.spans.SpanTermQuery in project elasticsearch by elastic.

the class SimpleQueryParserTests method testAnalyzerWithGraph.

public void testAnalyzerWithGraph() {
    SimpleQueryParser.Settings settings = new SimpleQueryParser.Settings();
    settings.analyzeWildcard(true);
    Map<String, Float> weights = new HashMap<>();
    weights.put("field1", 1.0f);
    SimpleQueryParser parser = new MockSimpleQueryParser(new MockSynonymAnalyzer(), weights, -1, settings);
    for (Operator op : Operator.values()) {
        BooleanClause.Occur defaultOp = op.toBooleanClauseOccur();
        parser.setDefaultOperator(defaultOp);
        // non-phrase won't detect multi-word synonym because of whitespace splitting
        Query query = parser.parse("guinea pig");
        Query expectedQuery = new BooleanQuery.Builder().add(new BooleanClause(new TermQuery(new Term("field1", "guinea")), defaultOp)).add(new BooleanClause(new TermQuery(new Term("field1", "pig")), defaultOp)).build();
        assertThat(query, equalTo(expectedQuery));
        // phrase will pick it up
        query = parser.parse("\"guinea pig\"");
        SpanTermQuery span1 = new SpanTermQuery(new Term("field1", "guinea"));
        SpanTermQuery span2 = new SpanTermQuery(new Term("field1", "pig"));
        expectedQuery = new SpanOrQuery(new SpanNearQuery(new SpanQuery[] { span1, span2 }, 0, true), new SpanTermQuery(new Term("field1", "cavy")));
        assertThat(query, equalTo(expectedQuery));
        // phrase with slop
        query = parser.parse("big \"tiny guinea pig\"~2");
        expectedQuery = new BooleanQuery.Builder().add(new TermQuery(new Term("field1", "big")), defaultOp).add(new SpanNearQuery(new SpanQuery[] { new SpanTermQuery(new Term("field1", "tiny")), new SpanOrQuery(new SpanNearQuery(new SpanQuery[] { span1, span2 }, 0, true), new SpanTermQuery(new Term("field1", "cavy"))) }, 2, true), defaultOp).build();
        assertThat(query, equalTo(expectedQuery));
    }
}

Example 95 with SpanTermQuery

use of org.apache.lucene.search.spans.SpanTermQuery in project elasticsearch by elastic.

the class CandidateQueryTests method testDuelSpecificQueries.

public void testDuelSpecificQueries() throws Exception {
    List<ParseContext.Document> documents = new ArrayList<>();
    CommonTermsQuery commonTermsQuery = new CommonTermsQuery(BooleanClause.Occur.SHOULD, BooleanClause.Occur.SHOULD, 128);
    commonTermsQuery.add(new Term("field", "quick"));
    commonTermsQuery.add(new Term("field", "brown"));
    commonTermsQuery.add(new Term("field", "fox"));
    addQuery(commonTermsQuery, documents);
    BlendedTermQuery blendedTermQuery = BlendedTermQuery.booleanBlendedQuery(new Term[] { new Term("field", "quick"), new Term("field", "brown"), new Term("field", "fox") }, false);
    addQuery(blendedTermQuery, documents);
    SpanNearQuery spanNearQuery = new SpanNearQuery.Builder("field", true).addClause(new SpanTermQuery(new Term("field", "quick"))).addClause(new SpanTermQuery(new Term("field", "brown"))).addClause(new SpanTermQuery(new Term("field", "fox"))).build();
    addQuery(spanNearQuery, documents);
    SpanNearQuery spanNearQuery2 = new SpanNearQuery.Builder("field", true).addClause(new SpanTermQuery(new Term("field", "the"))).addClause(new SpanTermQuery(new Term("field", "lazy"))).addClause(new SpanTermQuery(new Term("field", "doc"))).build();
    SpanOrQuery spanOrQuery = new SpanOrQuery(spanNearQuery, spanNearQuery2);
    addQuery(spanOrQuery, documents);
    SpanNotQuery spanNotQuery = new SpanNotQuery(spanNearQuery, spanNearQuery);
    addQuery(spanNotQuery, documents);
    long lowerLong = randomIntBetween(0, 256);
    long upperLong = lowerLong + randomIntBetween(0, 32);
    addQuery(LongPoint.newRangeQuery("long_field", lowerLong, upperLong), documents);
    indexWriter.addDocuments(documents);
    indexWriter.close();
    directoryReader = DirectoryReader.open(directory);
    IndexSearcher shardSearcher = newSearcher(directoryReader);
    // Disable query cache, because ControlQuery cannot be cached...
    shardSearcher.setQueryCache(null);
    Document document = new Document();
    document.add(new TextField("field", "the quick brown fox jumps over the lazy dog", Field.Store.NO));
    long randomLong = randomIntBetween((int) lowerLong, (int) upperLong);
    document.add(new LongPoint("long_field", randomLong));
    MemoryIndex memoryIndex = MemoryIndex.fromDocument(document, new WhitespaceAnalyzer());
    duelRun(queryStore, memoryIndex, shardSearcher);
}