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

Example 81 with BoostQuery

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

the class SolrPluginUtils method flattenBooleanQuery.

private static void flattenBooleanQuery(BooleanQuery.Builder to, BooleanQuery from, float fromBoost) {
    for (BooleanClause clause : from.clauses()) {
        Query cq = clause.getQuery();
        float boost = fromBoost;
        while (cq instanceof BoostQuery) {
            BoostQuery bq = (BoostQuery) cq;
            cq = bq.getQuery();
            boost *= bq.getBoost();
        }
        if (cq instanceof BooleanQuery && !clause.isRequired() && !clause.isProhibited()) {
            /* we can recurse */
            flattenBooleanQuery(to, (BooleanQuery) cq, boost);
        } else {
            to.add(clause);
        }
    }
}

Example 82 with BoostQuery

use of org.apache.lucene.search.BoostQuery 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()]);
}

Example 83 with BoostQuery

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

the class DisjunctionMaxQueryBuilder method getQuery.

/* (non-Javadoc)
    * @see org.apache.lucene.xmlparser.QueryObjectBuilder#process(org.w3c.dom.Element)
    */
@Override
public Query getQuery(Element e) throws ParserException {
    float tieBreaker = DOMUtils.getAttribute(e, "tieBreaker", 0.0f);
    List<Query> disjuncts = new ArrayList<>();
    NodeList nl = e.getChildNodes();
    final int nlLen = nl.getLength();
    for (int i = 0; i < nlLen; i++) {
        Node node = nl.item(i);
        if (node instanceof Element) {
            // all elements are disjuncts.
            Element queryElem = (Element) node;
            Query q = factory.getQuery(queryElem);
            disjuncts.add(q);
        }
    }
    Query q = new DisjunctionMaxQuery(disjuncts, tieBreaker);
    float boost = DOMUtils.getAttribute(e, "boost", 1.0f);
    if (boost != 1f) {
        q = new BoostQuery(q, boost);
    }
    return q;
}

Example 84 with BoostQuery

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

the class FuzzyLikeThisQueryBuilder method getQuery.

@Override
public Query getQuery(Element e) throws ParserException {
    NodeList nl = e.getElementsByTagName("Field");
    int maxNumTerms = DOMUtils.getAttribute(e, "maxNumTerms", DEFAULT_MAX_NUM_TERMS);
    FuzzyLikeThisQuery fbq = new FuzzyLikeThisQuery(maxNumTerms, analyzer);
    fbq.setIgnoreTF(DOMUtils.getAttribute(e, "ignoreTF", DEFAULT_IGNORE_TF));
    final int nlLen = nl.getLength();
    for (int i = 0; i < nlLen; i++) {
        Element fieldElem = (Element) nl.item(i);
        float minSimilarity = DOMUtils.getAttribute(fieldElem, "minSimilarity", DEFAULT_MIN_SIMILARITY);
        int prefixLength = DOMUtils.getAttribute(fieldElem, "prefixLength", DEFAULT_PREFIX_LENGTH);
        String fieldName = DOMUtils.getAttributeWithInheritance(fieldElem, "fieldName");
        String value = DOMUtils.getText(fieldElem);
        fbq.addTerms(value, fieldName, minSimilarity, prefixLength);
    }
    Query q = fbq;
    float boost = DOMUtils.getAttribute(e, "boost", 1.0f);
    if (boost != 1f) {
        q = new BoostQuery(fbq, boost);
    }
    return q;
}

Example 85 with BoostQuery

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

the class LikeThisQueryBuilder method getQuery.

/* (non-Javadoc)
    * @see org.apache.lucene.xmlparser.QueryObjectBuilder#process(org.w3c.dom.Element)
    */
@Override
public Query getQuery(Element e) throws ParserException {
    //a comma-delimited list of fields
    String fieldsList = e.getAttribute("fieldNames");
    String[] fields = defaultFieldNames;
    if ((fieldsList != null) && (fieldsList.trim().length() > 0)) {
        fields = fieldsList.trim().split(",");
        //trim the fieldnames
        for (int i = 0; i < fields.length; i++) {
            fields[i] = fields[i].trim();
        }
    }
    //Parse any "stopWords" attribute
    //TODO MoreLikeThis needs to ideally have per-field stopWords lists - until then
    //I use all analyzers/fields to generate multi-field compatible stop list
    String stopWords = e.getAttribute("stopWords");
    Set<String> stopWordsSet = null;
    if ((stopWords != null) && (fields != null)) {
        stopWordsSet = new HashSet<>();
        for (String field : fields) {
            try (TokenStream ts = analyzer.tokenStream(field, stopWords)) {
                CharTermAttribute termAtt = ts.addAttribute(CharTermAttribute.class);
                ts.reset();
                while (ts.incrementToken()) {
                    stopWordsSet.add(termAtt.toString());
                }
                ts.end();
            } catch (IOException ioe) {
                throw new ParserException("IoException parsing stop words list in " + getClass().getName() + ":" + ioe.getLocalizedMessage());
            }
        }
    }
    MoreLikeThisQuery mlt = new MoreLikeThisQuery(DOMUtils.getText(e), fields, analyzer, fields[0]);
    mlt.setMaxQueryTerms(DOMUtils.getAttribute(e, "maxQueryTerms", DEFAULT_MAX_QUERY_TERMS));
    mlt.setMinTermFrequency(DOMUtils.getAttribute(e, "minTermFrequency", DEFAULT_MIN_TERM_FREQUENCY));
    mlt.setPercentTermsToMatch(DOMUtils.getAttribute(e, "percentTermsToMatch", DEFAULT_PERCENT_TERMS_TO_MATCH) / 100);
    mlt.setStopWords(stopWordsSet);
    int minDocFreq = DOMUtils.getAttribute(e, "minDocFreq", -1);
    if (minDocFreq >= 0) {
        mlt.setMinDocFreq(minDocFreq);
    }
    Query q = mlt;
    float boost = DOMUtils.getAttribute(e, "boost", 1.0f);
    if (boost != 1f) {
        q = new BoostQuery(mlt, boost);
    }
    return q;
}