Examples with BooleanClause org.apache.lucene.search.BooleanClause used on opensource projects

Example 91 with BooleanClause

use of org.apache.lucene.search.BooleanClause in project lucene-solr by apache.

the class SimpleNaiveBayesClassifier method getWordFreqForClass.

/**
   * Returns the number of documents of the input class ( from the whole index or from a subset)
   * that contains the word ( in a specific field or in all the fields if no one selected)
   * @param word the token produced by the analyzer
   * @param term the term representing the class
   * @return the number of documents of the input class
   * @throws IOException if a low level I/O problem happens
   */
private int getWordFreqForClass(String word, Term term) throws IOException {
    BooleanQuery.Builder booleanQuery = new BooleanQuery.Builder();
    BooleanQuery.Builder subQuery = new BooleanQuery.Builder();
    for (String textFieldName : textFieldNames) {
        subQuery.add(new BooleanClause(new TermQuery(new Term(textFieldName, word)), BooleanClause.Occur.SHOULD));
    }
    booleanQuery.add(new BooleanClause(subQuery.build(), BooleanClause.Occur.MUST));
    booleanQuery.add(new BooleanClause(new TermQuery(term), BooleanClause.Occur.MUST));
    if (query != null) {
        booleanQuery.add(query, BooleanClause.Occur.MUST);
    }
    TotalHitCountCollector totalHitCountCollector = new TotalHitCountCollector();
    indexSearcher.search(booleanQuery.build(), totalHitCountCollector);
    return totalHitCountCollector.getTotalHits();
}

Example 92 with BooleanClause

use of org.apache.lucene.search.BooleanClause in project lucene-solr by apache.

the class SimpleNaiveBayesClassifier method countDocsWithClass.

/**
   * 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 = MultiFields.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;
}

Example 93 with BooleanClause

use of org.apache.lucene.search.BooleanClause in project lucene-solr by apache.

the class KNearestNeighborDocumentClassifier method knnSearch.

/**
   * Returns the top k results from a More Like This query based on the input document
   *
   * @param document the document to use for More Like This search
   * @return the top results for the MLT query
   * @throws IOException If there is a low-level I/O error
   */
private TopDocs knnSearch(Document document) throws IOException {
    BooleanQuery.Builder mltQuery = new BooleanQuery.Builder();
    for (String fieldName : textFieldNames) {
        String boost = null;
        if (fieldName.contains("^")) {
            String[] field2boost = fieldName.split("\\^");
            fieldName = field2boost[0];
            boost = field2boost[1];
        }
        String[] fieldValues = document.getValues(fieldName);
        // we want always to use the boost coming from TF * IDF of the term
        mlt.setBoost(true);
        if (boost != null) {
            // this is an additional multiplicative boost coming from the field boost
            mlt.setBoostFactor(Float.parseFloat(boost));
        }
        mlt.setAnalyzer(field2analyzer.get(fieldName));
        for (String fieldContent : fieldValues) {
            mltQuery.add(new BooleanClause(mlt.like(fieldName, new StringReader(fieldContent)), BooleanClause.Occur.SHOULD));
        }
    }
    Query classFieldQuery = new WildcardQuery(new Term(classFieldName, "*"));
    mltQuery.add(new BooleanClause(classFieldQuery, BooleanClause.Occur.MUST));
    if (query != null) {
        mltQuery.add(query, BooleanClause.Occur.MUST);
    }
    return indexSearcher.search(mltQuery.build(), k);
}

Example 94 with BooleanClause

use of org.apache.lucene.search.BooleanClause in project lucene-solr by apache.

the class QueryBuilder method addMinShouldMatchToBoolean.

/**
   * Rebuilds a boolean query and sets a new minimum number should match value.
   */
private BooleanQuery addMinShouldMatchToBoolean(BooleanQuery query, float fraction) {
    BooleanQuery.Builder builder = new BooleanQuery.Builder();
    builder.setMinimumNumberShouldMatch((int) (fraction * query.clauses().size()));
    for (BooleanClause clause : query) {
        builder.add(clause);
    }
    return builder.build();
}

Example 95 with BooleanClause

use of org.apache.lucene.search.BooleanClause in project lucene-solr by apache.

the class MultiTermHighlighting method extractAutomata.

/**
   * Extracts MultiTermQueries that match the provided field predicate.
   * Returns equivalent automata that will match terms.
   */
public static CharacterRunAutomaton[] extractAutomata(Query query, Predicate<String> fieldMatcher, boolean lookInSpan, Function<Query, Collection<Query>> preRewriteFunc) {
    // TODO Lucene needs a Query visitor API!  LUCENE-3041
    List<CharacterRunAutomaton> list = new ArrayList<>();
    Collection<Query> customSubQueries = preRewriteFunc.apply(query);
    if (customSubQueries != null) {
        for (Query sub : customSubQueries) {
            list.addAll(Arrays.asList(extractAutomata(sub, fieldMatcher, lookInSpan, preRewriteFunc)));
        }
    } else if (query instanceof BooleanQuery) {
        for (BooleanClause clause : (BooleanQuery) query) {
            if (!clause.isProhibited()) {
                list.addAll(Arrays.asList(extractAutomata(clause.getQuery(), fieldMatcher, lookInSpan, preRewriteFunc)));
            }
        }
    } else if (query instanceof ConstantScoreQuery) {
        list.addAll(Arrays.asList(extractAutomata(((ConstantScoreQuery) query).getQuery(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (query instanceof BoostQuery) {
        list.addAll(Arrays.asList(extractAutomata(((BoostQuery) query).getQuery(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (query instanceof DisjunctionMaxQuery) {
        for (Query sub : ((DisjunctionMaxQuery) query).getDisjuncts()) {
            list.addAll(Arrays.asList(extractAutomata(sub, fieldMatcher, lookInSpan, preRewriteFunc)));
        }
    } else if (lookInSpan && query instanceof SpanOrQuery) {
        for (Query sub : ((SpanOrQuery) query).getClauses()) {
            list.addAll(Arrays.asList(extractAutomata(sub, fieldMatcher, lookInSpan, preRewriteFunc)));
        }
    } else if (lookInSpan && query instanceof SpanNearQuery) {
        for (Query sub : ((SpanNearQuery) query).getClauses()) {
            list.addAll(Arrays.asList(extractAutomata(sub, fieldMatcher, lookInSpan, preRewriteFunc)));
        }
    } else if (lookInSpan && query instanceof SpanNotQuery) {
        list.addAll(Arrays.asList(extractAutomata(((SpanNotQuery) query).getInclude(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (lookInSpan && query instanceof SpanPositionCheckQuery) {
        list.addAll(Arrays.asList(extractAutomata(((SpanPositionCheckQuery) query).getMatch(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (lookInSpan && query instanceof SpanBoostQuery) {
        list.addAll(Arrays.asList(extractAutomata(((SpanBoostQuery) query).getQuery(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (lookInSpan && query instanceof SpanMultiTermQueryWrapper) {
        list.addAll(Arrays.asList(extractAutomata(((SpanMultiTermQueryWrapper<?>) query).getWrappedQuery(), fieldMatcher, lookInSpan, preRewriteFunc)));
    } else if (query instanceof PrefixQuery) {
        final PrefixQuery pq = (PrefixQuery) query;
        Term prefix = pq.getPrefix();
        if (fieldMatcher.test(prefix.field())) {
            list.add(new CharacterRunAutomaton(Operations.concatenate(Automata.makeString(prefix.text()), Automata.makeAnyString())) {

                @Override
                public String toString() {
                    return pq.toString();
                }
            });
        }
    } else if (query instanceof FuzzyQuery) {
        final FuzzyQuery fq = (FuzzyQuery) query;
        if (fieldMatcher.test(fq.getField())) {
            String utf16 = fq.getTerm().text();
            int[] termText = new int[utf16.codePointCount(0, utf16.length())];
            for (int cp, i = 0, j = 0; i < utf16.length(); i += Character.charCount(cp)) {
                termText[j++] = cp = utf16.codePointAt(i);
            }
            int termLength = termText.length;
            int prefixLength = Math.min(fq.getPrefixLength(), termLength);
            String suffix = UnicodeUtil.newString(termText, prefixLength, termText.length - prefixLength);
            LevenshteinAutomata builder = new LevenshteinAutomata(suffix, fq.getTranspositions());
            String prefix = UnicodeUtil.newString(termText, 0, prefixLength);
            Automaton automaton = builder.toAutomaton(fq.getMaxEdits(), prefix);
            list.add(new CharacterRunAutomaton(automaton) {

                @Override
                public String toString() {
                    return fq.toString();
                }
            });
        }
    } else if (query instanceof TermRangeQuery) {
        final TermRangeQuery tq = (TermRangeQuery) query;
        if (fieldMatcher.test(tq.getField())) {
            final CharsRef lowerBound;
            if (tq.getLowerTerm() == null) {
                lowerBound = null;
            } else {
                lowerBound = new CharsRef(tq.getLowerTerm().utf8ToString());
            }
            final CharsRef upperBound;
            if (tq.getUpperTerm() == null) {
                upperBound = null;
            } else {
                upperBound = new CharsRef(tq.getUpperTerm().utf8ToString());
            }
            final boolean includeLower = tq.includesLower();
            final boolean includeUpper = tq.includesUpper();
            final CharsRef scratch = new CharsRef();
            @SuppressWarnings("deprecation") final Comparator<CharsRef> comparator = CharsRef.getUTF16SortedAsUTF8Comparator();
            // this is *not* an automaton, but its very simple
            list.add(new CharacterRunAutomaton(Automata.makeEmpty()) {

                @Override
                public boolean run(char[] s, int offset, int length) {
                    scratch.chars = s;
                    scratch.offset = offset;
                    scratch.length = length;
                    if (lowerBound != null) {
                        int cmp = comparator.compare(scratch, lowerBound);
                        if (cmp < 0 || (!includeLower && cmp == 0)) {
                            return false;
                        }
                    }
                    if (upperBound != null) {
                        int cmp = comparator.compare(scratch, upperBound);
                        if (cmp > 0 || (!includeUpper && cmp == 0)) {
                            return false;
                        }
                    }
                    return true;
                }

                @Override
                public String toString() {
                    return tq.toString();
                }
            });
        }
    } else if (query instanceof AutomatonQuery) {
        final AutomatonQuery aq = (AutomatonQuery) query;
        if (fieldMatcher.test(aq.getField())) {
            list.add(new CharacterRunAutomaton(aq.getAutomaton()) {

                @Override
                public String toString() {
                    return aq.toString();
                }
            });
        }
    }
    return list.toArray(new CharacterRunAutomaton[list.size()]);
}