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

Example 11 with TopFieldDocs

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

the class DrillSideways method search.

/**
   * Search, sorting by {@link Sort}, and computing
   * drill down and sideways counts.
   */
public DrillSidewaysResult search(DrillDownQuery query, Query filter, FieldDoc after, int topN, Sort sort, boolean doDocScores, boolean doMaxScore) throws IOException {
    if (filter != null) {
        query = new DrillDownQuery(config, filter, query);
    }
    if (sort != null) {
        int limit = searcher.getIndexReader().maxDoc();
        if (limit == 0) {
            // the collector does not alow numHits = 0
            limit = 1;
        }
        final int fTopN = Math.min(topN, limit);
        if (executor != null) {
            // We have an executor, let use the multi-threaded version
            final CollectorManager<TopFieldCollector, TopFieldDocs> collectorManager = new CollectorManager<TopFieldCollector, TopFieldDocs>() {

                @Override
                public TopFieldCollector newCollector() throws IOException {
                    return TopFieldCollector.create(sort, fTopN, after, true, doDocScores, doMaxScore);
                }

                @Override
                public TopFieldDocs reduce(Collection<TopFieldCollector> collectors) throws IOException {
                    final TopFieldDocs[] topFieldDocs = new TopFieldDocs[collectors.size()];
                    int pos = 0;
                    for (TopFieldCollector collector : collectors) topFieldDocs[pos++] = collector.topDocs();
                    return TopDocs.merge(sort, topN, topFieldDocs);
                }
            };
            ConcurrentDrillSidewaysResult<TopFieldDocs> r = search(query, collectorManager);
            return new DrillSidewaysResult(r.facets, r.collectorResult);
        } else {
            final TopFieldCollector hitCollector = TopFieldCollector.create(sort, fTopN, after, true, doDocScores, doMaxScore);
            DrillSidewaysResult r = search(query, hitCollector);
            return new DrillSidewaysResult(r.facets, hitCollector.topDocs());
        }
    } else {
        return search(after, query, topN);
    }
}

Example 12 with TopFieldDocs

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

the class LatLonPoint method nearest.

/**
   * Finds the {@code n} nearest indexed points to the provided point, according to Haversine distance.
   * <p>
   * This is functionally equivalent to running {@link MatchAllDocsQuery} with a {@link LatLonDocValuesField#newDistanceSort},
   * but is far more efficient since it takes advantage of properties the indexed BKD tree.  Currently this
   * only works with {@link Lucene60PointsFormat} (used by the default codec).  Multi-valued fields are
   * currently not de-duplicated, so if a document had multiple instances of the specified field that
   * make it into the top n, that document will appear more than once.
   * <p>
   * Documents are ordered by ascending distance from the location. The value returned in {@link FieldDoc} for
   * the hits contains a Double instance with the distance in meters.
   * 
   * @param searcher IndexSearcher to find nearest points from.
   * @param field field name. must not be null.
   * @param latitude latitude at the center: must be within standard +/-90 coordinate bounds.
   * @param longitude longitude at the center: must be within standard +/-180 coordinate bounds.
   * @param n the number of nearest neighbors to retrieve.
   * @return TopFieldDocs containing documents ordered by distance, where the field value for each {@link FieldDoc} is the distance in meters
   * @throws IllegalArgumentException if the underlying PointValues is not a {@code Lucene60PointsReader} (this is a current limitation), or
   *         if {@code field} or {@code searcher} is null, or if {@code latitude}, {@code longitude} or {@code n} are out-of-bounds
   * @throws IOException if an IOException occurs while finding the points.
   */
// TODO: what about multi-valued documents? what happens?
public static TopFieldDocs nearest(IndexSearcher searcher, String field, double latitude, double longitude, int n) throws IOException {
    GeoUtils.checkLatitude(latitude);
    GeoUtils.checkLongitude(longitude);
    if (n < 1) {
        throw new IllegalArgumentException("n must be at least 1; got " + n);
    }
    if (field == null) {
        throw new IllegalArgumentException("field must not be null");
    }
    if (searcher == null) {
        throw new IllegalArgumentException("searcher must not be null");
    }
    List<BKDReader> readers = new ArrayList<>();
    List<Integer> docBases = new ArrayList<>();
    List<Bits> liveDocs = new ArrayList<>();
    int totalHits = 0;
    for (LeafReaderContext leaf : searcher.getIndexReader().leaves()) {
        PointValues points = leaf.reader().getPointValues(field);
        if (points != null) {
            if (points instanceof BKDReader == false) {
                throw new IllegalArgumentException("can only run on Lucene60PointsReader points implementation, but got " + points);
            }
            totalHits += points.getDocCount();
            BKDReader reader = (BKDReader) points;
            if (reader != null) {
                readers.add(reader);
                docBases.add(leaf.docBase);
                liveDocs.add(leaf.reader().getLiveDocs());
            }
        }
    }
    NearestNeighbor.NearestHit[] hits = NearestNeighbor.nearest(latitude, longitude, readers, liveDocs, docBases, n);
    // Convert to TopFieldDocs:
    ScoreDoc[] scoreDocs = new ScoreDoc[hits.length];
    for (int i = 0; i < hits.length; i++) {
        NearestNeighbor.NearestHit hit = hits[i];
        scoreDocs[i] = new FieldDoc(hit.docID, 0.0f, new Object[] { Double.valueOf(hit.distanceMeters) });
    }
    return new TopFieldDocs(totalHits, scoreDocs, null, 0.0f);
}

Example 13 with TopFieldDocs

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

the class AnalyzingInfixSuggester method lookup.

/**
   * This is an advanced method providing the capability to send down to the suggester any 
   * arbitrary lucene query to be used to filter the result of the suggester
   * 
   * @param key the keyword being looked for
   * @param contextQuery an arbitrary Lucene query to be used to filter the result of the suggester. {@link #addContextToQuery} could be used to build this contextQuery.
   * @param num number of items to return
   * @param allTermsRequired all searched terms must match or not
   * @param doHighlight if true, the matching term will be highlighted in the search result
   * @return the result of the suggester
   * @throws IOException f the is IO exception while reading data from the index
   */
public List<LookupResult> lookup(CharSequence key, BooleanQuery contextQuery, int num, boolean allTermsRequired, boolean doHighlight) throws IOException {
    if (searcherMgr == null) {
        throw new IllegalStateException("suggester was not built");
    }
    final BooleanClause.Occur occur;
    if (allTermsRequired) {
        occur = BooleanClause.Occur.MUST;
    } else {
        occur = BooleanClause.Occur.SHOULD;
    }
    BooleanQuery.Builder query;
    Set<String> matchedTokens;
    String prefixToken = null;
    try (TokenStream ts = queryAnalyzer.tokenStream("", new StringReader(key.toString()))) {
        //long t0 = System.currentTimeMillis();
        ts.reset();
        final CharTermAttribute termAtt = ts.addAttribute(CharTermAttribute.class);
        final OffsetAttribute offsetAtt = ts.addAttribute(OffsetAttribute.class);
        String lastToken = null;
        query = new BooleanQuery.Builder();
        int maxEndOffset = -1;
        matchedTokens = new HashSet<>();
        while (ts.incrementToken()) {
            if (lastToken != null) {
                matchedTokens.add(lastToken);
                query.add(new TermQuery(new Term(TEXT_FIELD_NAME, lastToken)), occur);
            }
            lastToken = termAtt.toString();
            if (lastToken != null) {
                maxEndOffset = Math.max(maxEndOffset, offsetAtt.endOffset());
            }
        }
        ts.end();
        if (lastToken != null) {
            Query lastQuery;
            if (maxEndOffset == offsetAtt.endOffset()) {
                // Use PrefixQuery (or the ngram equivalent) when
                // there was no trailing discarded chars in the
                // string (e.g. whitespace), so that if query does
                // not end with a space we show prefix matches for
                // that token:
                lastQuery = getLastTokenQuery(lastToken);
                prefixToken = lastToken;
            } else {
                // Use TermQuery for an exact match if there were
                // trailing discarded chars (e.g. whitespace), so
                // that if query ends with a space we only show
                // exact matches for that term:
                matchedTokens.add(lastToken);
                lastQuery = new TermQuery(new Term(TEXT_FIELD_NAME, lastToken));
            }
            if (lastQuery != null) {
                query.add(lastQuery, occur);
            }
        }
        if (contextQuery != null) {
            boolean allMustNot = true;
            for (BooleanClause clause : contextQuery.clauses()) {
                if (clause.getOccur() != BooleanClause.Occur.MUST_NOT) {
                    allMustNot = false;
                    break;
                }
            }
            if (allMustNot) {
                // All are MUST_NOT: add the contextQuery to the main query instead (not as sub-query)
                for (BooleanClause clause : contextQuery.clauses()) {
                    query.add(clause);
                }
            } else if (allTermsRequired == false) {
                // We must carefully upgrade the query clauses to MUST:
                BooleanQuery.Builder newQuery = new BooleanQuery.Builder();
                newQuery.add(query.build(), BooleanClause.Occur.MUST);
                newQuery.add(contextQuery, BooleanClause.Occur.MUST);
                query = newQuery;
            } else {
                // Add contextQuery as sub-query
                query.add(contextQuery, BooleanClause.Occur.MUST);
            }
        }
    }
    // TODO: we could allow blended sort here, combining
    // weight w/ score.  Now we ignore score and sort only
    // by weight:
    Query finalQuery = finishQuery(query, allTermsRequired);
    //System.out.println("finalQuery=" + finalQuery);
    // Sort by weight, descending:
    TopFieldCollector c = TopFieldCollector.create(SORT, num, true, false, false);
    // We sorted postings by weight during indexing, so we
    // only retrieve the first num hits now:
    Collector c2 = new EarlyTerminatingSortingCollector(c, SORT, num);
    List<LookupResult> results = null;
    SearcherManager mgr;
    IndexSearcher searcher;
    synchronized (searcherMgrLock) {
        // acquire & release on same SearcherManager, via local reference
        mgr = searcherMgr;
        searcher = mgr.acquire();
    }
    try {
        //System.out.println("got searcher=" + searcher);
        searcher.search(finalQuery, c2);
        TopFieldDocs hits = c.topDocs();
        // Slower way if postings are not pre-sorted by weight:
        // hits = searcher.search(query, null, num, SORT);
        results = createResults(searcher, hits, num, key, doHighlight, matchedTokens, prefixToken);
    } finally {
        mgr.release(searcher);
    }
    return results;
}

Example 14 with TopFieldDocs

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

the class BlobHandler method handleRequestBody.

@Override
public void handleRequestBody(final SolrQueryRequest req, SolrQueryResponse rsp) throws Exception {
    String httpMethod = req.getHttpMethod();
    String path = (String) req.getContext().get("path");
    SolrConfigHandler.setWt(req, JSON);
    List<String> pieces = StrUtils.splitSmart(path, '/');
    String blobName = null;
    if (pieces.size() >= 3)
        blobName = pieces.get(2);
    if ("POST".equals(httpMethod)) {
        if (blobName == null || blobName.isEmpty()) {
            rsp.add("error", "Name not found");
            return;
        }
        String err = SolrConfigHandler.validateName(blobName);
        if (err != null) {
            log.warn("no blob name");
            rsp.add("error", err);
            return;
        }
        if (req.getContentStreams() == null) {
            log.warn("no content stream");
            rsp.add("error", "No stream");
            return;
        }
        for (ContentStream stream : req.getContentStreams()) {
            ByteBuffer payload = SimplePostTool.inputStreamToByteArray(stream.getStream(), maxSize);
            MessageDigest m = MessageDigest.getInstance("MD5");
            m.update(payload.array(), payload.position(), payload.limit());
            String md5 = new BigInteger(1, m.digest()).toString(16);
            TopDocs duplicate = req.getSearcher().search(new TermQuery(new Term("md5", md5)), 1);
            if (duplicate.totalHits > 0) {
                rsp.add("error", "duplicate entry");
                forward(req, null, new MapSolrParams((Map) makeMap("q", "md5:" + md5, "fl", "id,size,version,timestamp,blobName")), rsp);
                log.warn("duplicate entry for blob :" + blobName);
                return;
            }
            TopFieldDocs docs = req.getSearcher().search(new TermQuery(new Term("blobName", blobName)), 1, new Sort(new SortField("version", SortField.Type.LONG, true)));
            long version = 0;
            if (docs.totalHits > 0) {
                Document doc = req.getSearcher().doc(docs.scoreDocs[0].doc);
                Number n = doc.getField("version").numericValue();
                version = n.longValue();
            }
            version++;
            String id = blobName + "/" + version;
            Map<String, Object> doc = makeMap(ID, id, "md5", md5, "blobName", blobName, VERSION, version, "timestamp", new Date(), "size", payload.limit(), "blob", payload);
            verifyWithRealtimeGet(blobName, version, req, doc);
            log.info(StrUtils.formatString("inserting new blob {0} ,size {1}, md5 {2}", doc.get(ID), String.valueOf(payload.limit()), md5));
            indexMap(req, rsp, doc);
            log.info(" Successfully Added and committed a blob with id {} and size {} ", id, payload.limit());
            break;
        }
    } else {
        int version = -1;
        if (pieces.size() > 3) {
            try {
                version = Integer.parseInt(pieces.get(3));
            } catch (NumberFormatException e) {
                rsp.add("error", "Invalid version" + pieces.get(3));
                return;
            }
        }
        if (ReplicationHandler.FILE_STREAM.equals(req.getParams().get(CommonParams.WT))) {
            if (blobName == null) {
                throw new SolrException(SolrException.ErrorCode.NOT_FOUND, "Please send the request in the format /blob/<blobName>/<version>");
            } else {
                String q = "blobName:{0}";
                if (version != -1)
                    q = "id:{0}/{1}";
                QParser qparser = QParser.getParser(StrUtils.formatString(q, blobName, version), req);
                final TopDocs docs = req.getSearcher().search(qparser.parse(), 1, new Sort(new SortField("version", SortField.Type.LONG, true)));
                if (docs.totalHits > 0) {
                    rsp.add(ReplicationHandler.FILE_STREAM, new SolrCore.RawWriter() {

                        @Override
                        public void write(OutputStream os) throws IOException {
                            Document doc = req.getSearcher().doc(docs.scoreDocs[0].doc);
                            IndexableField sf = doc.getField("blob");
                            FieldType fieldType = req.getSchema().getField("blob").getType();
                            ByteBuffer buf = (ByteBuffer) fieldType.toObject(sf);
                            if (buf == null) {
                                //should never happen unless a user wrote this document directly
                                throw new SolrException(SolrException.ErrorCode.NOT_FOUND, "Invalid document . No field called blob");
                            } else {
                                os.write(buf.array(), 0, buf.limit());
                            }
                        }
                    });
                } else {
                    throw new SolrException(SolrException.ErrorCode.NOT_FOUND, StrUtils.formatString("Invalid combination of blobName {0} and version {1}", blobName, version));
                }
            }
        } else {
            String q = "*:*";
            if (blobName != null) {
                q = "blobName:{0}";
                if (version != -1) {
                    q = "id:{0}/{1}";
                }
            }
            forward(req, null, new MapSolrParams((Map) makeMap("q", StrUtils.formatString(q, blobName, version), "fl", "id,size,version,timestamp,blobName,md5", SORT, "version desc")), rsp);
        }
    }
}

Example 15 with TopFieldDocs

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

the class TestDemoExpressions method testDollarVariable.

/** Uses variables with $ */
public void testDollarVariable() throws Exception {
    Expression expr = JavascriptCompiler.compile("$0+$score");
    SimpleBindings bindings = new SimpleBindings();
    bindings.add(new SortField("$0", SortField.Type.SCORE));
    bindings.add(new SortField("$score", SortField.Type.SCORE));
    Sort sort = new Sort(expr.getSortField(bindings, true));
    Query query = new TermQuery(new Term("body", "contents"));
    TopFieldDocs td = searcher.search(query, 3, sort, true, true);
    for (int i = 0; i < 3; i++) {
        FieldDoc d = (FieldDoc) td.scoreDocs[i];
        float expected = 2 * d.score;
        float actual = ((Double) d.fields[0]).floatValue();
        assertEquals(expected, actual, CheckHits.explainToleranceDelta(expected, actual));
    }
}