Example usage for org.apache.solr.client.solrj SolrQuery getCopy

List of usage examples for org.apache.solr.client.solrj SolrQuery getCopy

Introduction

In this page you can find the example usage for org.apache.solr.client.solrj SolrQuery getCopy.

Prototype

public SolrQuery getCopy() 

Source Link

Document

get a deep copy of this object

Usage

From source file:au.org.ala.biocache.dao.SearchDAOImpl.java

License:Open Source License

/**
 * Writes the index fields to the supplied output stream in CSV format.
 * <p>/*  w w  w  .  ja  v  a 2s.  c  om*/
 * DM: refactored to split the query by month to improve performance.
 * Further enhancements possible:
 * 1) Multi threaded
 * 2) More filtering, by year or decade..
 *
 * @param downloadParams
 * @param out
 * @param includeSensitive
 * @param dd               The details of the download
 * @param checkLimit
 * @param nextExecutor     The ExecutorService to use to process results on different threads
 * @throws Exception
 */
@Override
public ConcurrentMap<String, AtomicInteger> writeResultsFromIndexToStream(
        final DownloadRequestParams downloadParams, final OutputStream out, final boolean includeSensitive,
        final DownloadDetailsDTO dd, boolean checkLimit, final ExecutorService nextExecutor) throws Exception {
    expandRequestedFields(downloadParams, true);

    if (dd != null) {
        dd.resetCounts();
    }

    long start = System.currentTimeMillis();
    final ConcurrentMap<String, AtomicInteger> uidStats = new ConcurrentHashMap<>();
    getServer();

    try {
        SolrQuery solrQuery = new SolrQuery();
        queryFormatUtils.formatSearchQuery(downloadParams);

        String dFields = downloadParams.getFields();

        if (includeSensitive) {
            //include raw latitude and longitudes
            if (dFields.contains("decimalLatitude.p")) {
                dFields = dFields.replaceFirst("decimalLatitude.p",
                        "sensitive_latitude,sensitive_longitude,decimalLatitude.p");
            } else if (dFields.contains("decimalLatitude")) {
                dFields = dFields.replaceFirst("decimalLatitude",
                        "sensitive_latitude,sensitive_longitude,decimalLatitude");
            }
            if (dFields.contains(",locality,")) {
                dFields = dFields.replaceFirst(",locality,", ",locality,sensitive_locality,");
            }
            if (dFields.contains(",locality.p,")) {
                dFields = dFields.replaceFirst(",locality.p,", ",locality.p,sensitive_locality,");
            }
        }

        StringBuilder sb = new StringBuilder(dFields);
        if (!downloadParams.getExtra().isEmpty()) {
            sb.append(",").append(downloadParams.getExtra());
        }

        String[] requestedFields = sb.toString().split(",");
        List<String>[] indexedFields;
        if (downloadFields == null) {
            //default to include everything
            java.util.List<String> mappedNames = new java.util.LinkedList<String>();
            for (int i = 0; i < requestedFields.length; i++)
                mappedNames.add(requestedFields[i]);

            indexedFields = new List[] { mappedNames, new java.util.LinkedList<String>(), mappedNames,
                    mappedNames, new ArrayList(), new ArrayList() };
        } else {
            indexedFields = downloadFields.getIndexFields(requestedFields, downloadParams.getDwcHeaders(),
                    downloadParams.getLayersServiceUrl());
        }
        //apply custom header
        String[] customHeader = dd.getRequestParams().getCustomHeader().split(",");
        for (int i = 0; i + 1 < customHeader.length; i += 2) {
            for (int j = 0; j < indexedFields[0].size(); j++) {
                if (customHeader[i].equals(indexedFields[0].get(j))) {
                    indexedFields[2].set(j, customHeader[i + 1]);
                }
            }
            for (int j = 0; j < indexedFields[4].size(); j++) {
                if (customHeader[i].equals(indexedFields[5].get(j))) {
                    indexedFields[4].set(j, customHeader[i + 1]);
                }
            }
        }
        if (logger.isDebugEnabled()) {
            logger.debug("Fields included in download: " + indexedFields[0]);
            logger.debug("Fields excluded from download: " + indexedFields[1]);
            logger.debug("The headers in downloads: " + indexedFields[2]);
            logger.debug("Analysis headers: " + indexedFields[4]);
            logger.debug("Analysis fields: " + indexedFields[5]);
        }

        //set the fields to the ones that are available in the index
        String[] fields = indexedFields[0].toArray(new String[] {});
        solrQuery.setFields(fields);
        StringBuilder qasb = new StringBuilder();
        if (!"none".equals(downloadParams.getQa())) {
            solrQuery.addField("assertions");
            if (!"all".equals(downloadParams.getQa()) && !"includeall".equals(downloadParams.getQa())) {
                //add all the qa fields
                qasb.append(downloadParams.getQa());
            }
        }
        solrQuery.addField("institution_uid").addField("collection_uid").addField("data_resource_uid")
                .addField("data_provider_uid");

        solrQuery.setQuery(downloadParams.getFormattedQuery());
        solrQuery.setFacetMinCount(1);
        solrQuery.setFacetLimit(-1);

        //get the assertion facets to add them to the download fields
        boolean getAssertionsFromFacets = "all".equals(downloadParams.getQa())
                || "includeall".equals(downloadParams.getQa());
        SolrQuery monthAssertionsQuery = getAssertionsFromFacets
                ? solrQuery.getCopy().addFacetField("month", "assertions")
                : solrQuery.getCopy().addFacetField("month");
        if (getAssertionsFromFacets) {
            //set the order for the facet to be based on the index - this will force the assertions to be returned in the same order each time
            //based on alphabetical sort.  The number of QA's may change between searches so we can't guarantee that the order won't change
            monthAssertionsQuery.add("f.assertions.facet.sort", "index");
        }
        QueryResponse facetQuery = runSolrQuery(monthAssertionsQuery, downloadParams.getFormattedFq(), 0, 0,
                "score", "asc");

        //set the totalrecords for the download details
        dd.setTotalRecords(facetQuery.getResults().getNumFound());

        //use a separately configured and smaller limit when output will be unzipped
        final long maxDownloadSize;
        if (MAX_DOWNLOAD_SIZE > unzippedLimit && out instanceof OptionalZipOutputStream
                && ((OptionalZipOutputStream) out).getType() == OptionalZipOutputStream.Type.unzipped) {
            maxDownloadSize = unzippedLimit;
        } else {
            maxDownloadSize = MAX_DOWNLOAD_SIZE;
        }

        if (checkLimit && dd.getTotalRecords() < maxDownloadSize) {
            checkLimit = false;
        }

        //get the month facets to add them to the download fields get the assertion facets.
        List<Count> splitByFacet = null;

        for (FacetField facet : facetQuery.getFacetFields()) {
            if (facet.getName().equals("assertions") && facet.getValueCount() > 0) {
                qasb.append(getQAFromFacet(facet));
            }
            if (facet.getName().equals("month") && facet.getValueCount() > 0) {
                splitByFacet = facet.getValues();
            }
        }

        if ("includeall".equals(downloadParams.getQa())) {
            qasb = getAllQAFields();
        }

        String qas = qasb.toString();

        //include sensitive fields in the header when the output will be partially sensitive
        final String[] sensitiveFields;
        final String[] notSensitiveFields;
        if (dd.getSensitiveFq() != null) {
            List<String>[] sensitiveHdr = downloadFields.getIndexFields(sensitiveSOLRHdr,
                    downloadParams.getDwcHeaders(), downloadParams.getLayersServiceUrl());

            //header for the output file
            indexedFields[2].addAll(sensitiveHdr[2]);

            //lookup for fields from sensitive queries
            sensitiveFields = org.apache.commons.lang3.ArrayUtils.addAll(
                    indexedFields[0].toArray(new String[] {}), sensitiveHdr[0].toArray(new String[] {}));

            //use general fields when sensitive data is not permitted
            notSensitiveFields = org.apache.commons.lang3.ArrayUtils
                    .addAll(indexedFields[0].toArray(new String[] {}), notSensitiveSOLRHdr);
        } else {
            sensitiveFields = new String[0];
            notSensitiveFields = fields;
        }

        //add analysis headers
        indexedFields[2].addAll(indexedFields[4]);
        final String[] analysisFields = indexedFields[5].toArray(new String[0]);

        final String[] qaFields = qas.equals("") ? new String[] {} : qas.split(",");
        String[] qaTitles = downloadFields.getHeader(qaFields, false, false);

        String[] header = org.apache.commons.lang3.ArrayUtils.addAll(indexedFields[2].toArray(new String[] {}),
                qaTitles);

        //retain output header fields and field names for inclusion of header info in the download
        StringBuilder infoFields = new StringBuilder("infoFields");
        for (String h : indexedFields[3])
            infoFields.append(",").append(h);
        for (String h : qaFields)
            infoFields.append(",").append(h);

        StringBuilder infoHeader = new StringBuilder("infoHeaders");
        for (String h : header)
            infoHeader.append(",").append(h);

        String info = infoFields.toString();
        while (info.contains(",,"))
            info = info.replace(",,", ",");
        uidStats.put(info, new AtomicInteger(-1));
        String hdr = infoHeader.toString();
        while (hdr.contains(",,"))
            hdr = hdr.replace(",,", ",");
        uidStats.put(hdr, new AtomicInteger(-2));

        //construct correct RecordWriter based on the supplied fileType
        final RecordWriterError rw = downloadParams.getFileType().equals("csv")
                ? new CSVRecordWriter(out, header, downloadParams.getSep(), downloadParams.getEsc())
                : (downloadParams.getFileType().equals("tsv") ? new TSVRecordWriter(out, header)
                        : new ShapeFileRecordWriter(tmpShapefileDir, downloadParams.getFile(), out,
                                (String[]) ArrayUtils.addAll(fields, qaFields)));

        // Requirement to be able to propagate interruptions to all other threads for this execution
        // Doing this via this variable
        final AtomicBoolean interruptFound = dd != null ? dd.getInterrupt() : new AtomicBoolean(false);

        // Create a fixed length blocking queue for buffering results before they are written
        // This also creates a push-back effect to throttle the results generating threads
        // when it fills and offers to it are delayed until the writer consumes elements from the queue
        final BlockingQueue<String[]> queue = new ArrayBlockingQueue<>(resultsQueueLength);
        // Create a sentinel that we can check for reference equality to signal the end of the queue
        final String[] sentinel = new String[0];
        // An implementation of RecordWriter that adds to an in-memory queue
        final RecordWriter concurrentWrapper = new RecordWriter() {
            private AtomicBoolean finalised = new AtomicBoolean(false);
            private AtomicBoolean finalisedComplete = new AtomicBoolean(false);

            @Override
            public void write(String[] nextLine) {
                try {
                    if (Thread.currentThread().isInterrupted() || interruptFound.get() || finalised.get()) {
                        finalise();
                        return;
                    }
                    while (!queue.offer(nextLine, writerTimeoutWaitMillis, TimeUnit.MILLISECONDS)) {
                        if (Thread.currentThread().isInterrupted() || interruptFound.get() || finalised.get()) {
                            finalise();
                            break;
                        }
                    }
                } catch (InterruptedException e) {
                    Thread.currentThread().interrupt();
                    interruptFound.set(true);
                    if (logger.isDebugEnabled()) {
                        logger.debug(
                                "Queue failed to accept the next record due to a thread interrupt, calling finalise the cleanup: ",
                                e);
                    }
                    // If we were interrupted then we should call finalise to cleanup
                    finalise();
                }
            }

            @Override
            public void finalise() {
                if (finalised.compareAndSet(false, true)) {
                    try {
                        // Offer the sentinel at least once, even when the thread is interrupted
                        while (!queue.offer(sentinel, writerTimeoutWaitMillis, TimeUnit.MILLISECONDS)) {
                            // If the thread is interrupted then the queue may not have any active consumers,
                            // so don't loop forever waiting for capacity in this case
                            // The hard shutdown phase will use queue.clear to ensure that the
                            // sentinel gets onto the queue at least once
                            if (Thread.currentThread().isInterrupted() || interruptFound.get()) {
                                break;
                            }
                        }
                    } catch (InterruptedException e) {
                        Thread.currentThread().interrupt();
                        interruptFound.set(true);
                        if (logger.isDebugEnabled()) {
                            logger.debug(
                                    "Queue failed to accept the sentinel in finalise due to a thread interrupt: ",
                                    e);
                        }
                    } finally {
                        finalisedComplete.set(true);
                    }
                }
            }

            @Override
            public boolean finalised() {
                return finalisedComplete.get();
            }

        };

        // A single thread that consumes elements put onto the queue until it sees the sentinel, finalising after the sentinel or an interrupt
        Runnable writerRunnable = new Runnable() {
            @Override
            public void run() {
                try {
                    long counter = 0;
                    while (true) {
                        counter = counter + 1;

                        if (Thread.currentThread().isInterrupted() || interruptFound.get()) {
                            break;
                        }

                        String[] take = queue.take();
                        // Sentinel object equality check to see if we are done
                        if (take == sentinel || Thread.currentThread().isInterrupted()
                                || interruptFound.get()) {
                            break;
                        }
                        // Otherwise write to the wrapped record writer
                        rw.write(take);

                        //test for errors. This can contain a flush so only test occasionally
                        if (counter % resultsQueueLength == 0 && rw.hasError()) {
                            throw RecordWriterException.newRecordWriterException(dd, downloadParams, true, rw);
                        }

                    }
                } catch (RecordWriterException e) {
                    //no trace information is available to print for these errors
                    logger.error(e.getMessage());
                    interruptFound.set(true);
                } catch (InterruptedException e) {
                    Thread.currentThread().interrupt();
                    interruptFound.set(true);
                } catch (Exception e) {
                    // Reuse interruptFound variable to signal that the writer had issues
                    interruptFound.set(true);
                    logger.error("Download writer failed.", e);
                } finally {
                    rw.finalise();
                }
            }
        };
        Thread writerThread = new Thread(writerRunnable);
        writerThread.start();

        try {
            if (rw instanceof ShapeFileRecordWriter) {
                dd.setHeaderMap(((ShapeFileRecordWriter) rw).getHeaderMappings());
            }

            //order the query by _docid_ for faster paging
            solrQuery.addSortField("_docid_", ORDER.asc);

            //for each month create a separate query that pages through 500 records per page
            List<SolrQuery> queries = new ArrayList<SolrQuery>();
            if (splitByFacet != null) {
                for (Count facet : splitByFacet) {
                    if (facet.getCount() > 0) {
                        SolrQuery splitByFacetQuery;
                        //do not add remainderQuery here
                        if (facet.getName() != null) {
                            splitByFacetQuery = solrQuery.getCopy()
                                    .addFilterQuery(facet.getFacetField().getName() + ":" + facet.getName());
                            splitByFacetQuery.setFacet(false);
                            queries.add(splitByFacetQuery);
                        }

                    }
                }
                if (splitByFacet.size() > 0) {
                    SolrQuery remainderQuery = solrQuery.getCopy()
                            .addFilterQuery("-" + splitByFacet.get(0).getFacetField().getName() + ":[* TO *]");
                    queries.add(0, remainderQuery);
                }
            } else {
                queries.add(0, solrQuery);
            }

            //split into sensitive and non-sensitive queries when
            // - not including all sensitive values
            // - there is a sensitive fq
            final List<SolrQuery> sensitiveQ = new ArrayList<SolrQuery>();
            if (!includeSensitive && dd.getSensitiveFq() != null) {
                sensitiveQ.addAll(
                        splitQueries(queries, dd.getSensitiveFq(), sensitiveSOLRHdr, notSensitiveSOLRHdr));
            }

            //Set<Future<Integer>> futures = new HashSet<Future<Integer>>();
            final AtomicInteger resultsCount = new AtomicInteger(0);
            final boolean threadCheckLimit = checkLimit;

            List<Callable<Integer>> solrCallables = new ArrayList<>(queries.size());
            // execute each query, writing the results to stream
            for (final SolrQuery splitByFacetQuery : queries) {
                // define a thread
                Callable<Integer> solrCallable = new Callable<Integer>() {
                    @Override
                    public Integer call() throws Exception {
                        int startIndex = 0;
                        // Randomise the wakeup time so they don't all wakeup on a periodic cycle
                        long localThrottle = throttle + Math.round(Math.random() * throttle);

                        String[] fq = downloadParams.getFormattedFq();
                        if (splitByFacetQuery.getFilterQueries() != null
                                && splitByFacetQuery.getFilterQueries().length > 0) {
                            if (fq == null) {
                                fq = new String[0];
                            }
                            fq = org.apache.commons.lang3.ArrayUtils.addAll(fq,
                                    splitByFacetQuery.getFilterQueries());
                        }

                        QueryResponse qr = runSolrQuery(splitByFacetQuery, fq, downloadBatchSize, startIndex,
                                "_docid_", "asc");
                        AtomicInteger recordsForThread = new AtomicInteger(0);
                        if (logger.isDebugEnabled()) {
                            logger.debug(
                                    splitByFacetQuery.getQuery() + " - results: " + qr.getResults().size());
                        }

                        while (qr != null && !qr.getResults().isEmpty() && !interruptFound.get()) {
                            if (logger.isDebugEnabled()) {
                                logger.debug(
                                        "Start index: " + startIndex + ", " + splitByFacetQuery.getQuery());
                            }
                            int count = 0;
                            if (sensitiveQ.contains(splitByFacetQuery)) {
                                count = processQueryResults(uidStats, sensitiveFields, qaFields,
                                        concurrentWrapper, qr, dd, threadCheckLimit, resultsCount,
                                        maxDownloadSize, analysisFields);
                            } else {
                                // write non-sensitive values into sensitive fields when not authorised for their sensitive values
                                count = processQueryResults(uidStats, notSensitiveFields, qaFields,
                                        concurrentWrapper, qr, dd, threadCheckLimit, resultsCount,
                                        maxDownloadSize, analysisFields);
                            }
                            recordsForThread.addAndGet(count);
                            startIndex += downloadBatchSize;
                            // we have already set the Filter query the first time the query was constructed
                            // rerun with the same params but different startIndex
                            if (!threadCheckLimit || resultsCount.get() < maxDownloadSize) {
                                if (!threadCheckLimit) {
                                    // throttle the download by sleeping
                                    Thread.sleep(localThrottle);
                                }
                                qr = runSolrQuery(splitByFacetQuery, null, downloadBatchSize, startIndex,
                                        "_docid_", "asc");
                            } else {
                                qr = null;
                            }
                        }
                        return recordsForThread.get();
                    }
                };
                solrCallables.add(solrCallable);
            }

            List<Future<Integer>> futures = new ArrayList<>(solrCallables.size());
            for (Callable<Integer> nextCallable : solrCallables) {
                futures.add(nextExecutor.submit(nextCallable));
            }

            // Busy wait because we need to be able to respond to an interrupt on any callable
            // and propagate it to all of the others for this particular query
            // Because the executor service is shared to prevent too many concurrent threads being run,
            // this requires a busy wait loop on the main thread to monitor state
            boolean waitAgain = false;
            do {
                waitAgain = false;
                for (Future<Integer> future : futures) {
                    if (!future.isDone()) {
                        // Wait again even if an interrupt flag is set, as it may have been set partway through the iteration
                        // The calls to future.cancel will occur next time if the interrupt is setup partway through an iteration
                        waitAgain = true;
                        // If one thread finds an interrupt it is propagated to others using the interruptFound AtomicBoolean
                        if (interruptFound.get()) {
                            future.cancel(true);
                        }
                    }
                }
                // Don't trigger the timeout interrupt if we don't have to wait again as we are already done at this point
                if (waitAgain && (System.currentTimeMillis() - start) > downloadMaxTime) {
                    interruptFound.set(true);
                    break;
                }

                if (waitAgain) {
                    Thread.sleep(downloadCheckBusyWaitSleep);
                }
            } while (waitAgain);

            AtomicInteger totalDownload = new AtomicInteger(0);
            for (Future<Integer> future : futures) {
                if (future.isDone()) {
                    totalDownload.addAndGet(future.get());
                } else {
                    // All incomplete futures that survived the loop above are cancelled here
                    future.cancel(true);
                }
            }

            long finish = System.currentTimeMillis();
            long timeTakenInSecs = (finish - start) / 1000;
            if (timeTakenInSecs <= 0)
                timeTakenInSecs = 1;
            if (logger.isInfoEnabled()) {
                logger.info("Download of " + resultsCount + " records in " + timeTakenInSecs
                        + " seconds. Record/sec: " + resultsCount.intValue() / timeTakenInSecs);
            }

        } finally {
            try {
                // Once we get here, we need to finalise starting at the concurrent wrapper,
                // as there are no more non-sentinel records to be added to the queue
                // This eventually triggers finalisation of the underlying writer when the queue empties
                // This is a soft shutdown, and hence we wait below for this stage to complete in normal circumstances
                // Note, this blocks for writerTimeoutWaitMillis trying to legitimately add the sentinel to the end of the queue
                // We force the sentinel to be added in the hard shutdown phase below
                concurrentWrapper.finalise();
            } finally {
                try {
                    // Track the current time right now so we can abort after downloadMaxCompletionTime milliseconds in this phase
                    final long completionStartTime = System.currentTimeMillis();
                    // Busy wait check for finalised to be called in the RecordWriter or something is interrupted
                    // By this stage, there are at maximum download.internal.queue.size items remaining (default 1000)
                    while (writerThread.isAlive() && !writerThread.isInterrupted() && !interruptFound.get()
                            && !Thread.currentThread().isInterrupted() && !rw.finalised()
                            && !((System.currentTimeMillis()
                                    - completionStartTime) > downloadMaxCompletionTime)) {
                        Thread.sleep(downloadCheckBusyWaitSleep);
                    }
                } finally {
                    try {
                        // Attempt all actions that could trigger the writer thread to finalise, as by this stage we are in hard shutdown mode

                        // Signal that we are in hard shutdown mode
                        interruptFound.set(true);

                        // Add the sentinel or clear the queue and try again until it gets onto the queue
                        // We are in hard shutdown mode, so only priority is that the queue either
                        // gets the sentinel or the thread is interrupted to clean up resources
                        while (!queue.offer(sentinel)) {
                            queue.clear();
                        }

                        // Interrupt the single writer thread
                        writerThread.interrupt();

                        // Explicitly call finalise on the RecordWriter as a backup
                        // In normal circumstances it is called via the sentinel or the interrupt
                        // This will not block if finalise has been called previously in the current three implementations
                        rw.finalise();
                    } finally {
                        if (rw != null && rw.hasError()) {
                            throw RecordWriterException.newRecordWriterException(dd, downloadParams, true, rw);
                        } else {
                            // Flush whatever output was still pending for more deterministic debugging
                            out.flush();
                        }
                    }
                }
            }
        }
    } catch (SolrServerException ex) {
        logger.error("Problem communicating with SOLR server while processing download. " + ex.getMessage(),
                ex);
    }
    return uidStats;
}

From source file:au.org.ala.biocache.dao.SearchDAOImpl.java

License:Open Source License

/**
 * Split a list of queries by a fq./*  w w w .java 2s . co m*/
 */
private List<SolrQuery> splitQueries(List<SolrQuery> queries, String fq, String[] fqFields,
        String[] notFqFields) {
    List<SolrQuery> notFQ = new ArrayList<SolrQuery>();
    List<SolrQuery> fQ = new ArrayList<SolrQuery>();

    for (SolrQuery query : queries) {
        SolrQuery nsq = query.getCopy().addFilterQuery("-(" + fq + ")");
        if (notFqFields != null) {
            for (String field : notFqFields)
                nsq.addField(field);
        }
        notFQ.add(nsq);

        SolrQuery sq = query.getCopy().addFilterQuery(fq);
        if (fqFields != null) {
            for (String field : fqFields)
                sq.addField(field);
        }
        fQ.add(sq);
    }

    queries.clear();
    queries.addAll(notFQ);
    queries.addAll(fQ);

    return fQ;
}

From source file:de.qaware.chronix.solr.client.stream.SolrStreamingService.java

License:Apache License

private void initStreamingService(SolrQuery query, SolrClient connection) {
    SolrQuery solrQuery = query.getCopy();
    solrQuery.setRows(0);/*w w  w. ja  v a2s  . co m*/
    //we do not need any data from the server expect the total number of found documents
    solrQuery.setFields("");

    try {
        QueryResponse response = connection.query(solrQuery);
        nrOfAvailableTimeSeries = response.getResults().getNumFound();
        queryStart = (long) response.getResponseHeader().get(ChronixSolrStorageConstants.QUERY_START_LONG);
        queryEnd = (long) response.getResponseHeader().get(ChronixSolrStorageConstants.QUERY_END_LONG);

    } catch (SolrServerException | IOException e) {
        LOGGER.error("SolrServerException occurred while querying server.", e);
    }
}

From source file:de.qaware.chronix.storage.solr.stream.SolrStreamingService.java

License:Apache License

private void initialStream(SolrQuery query, SolrClient connection) {
    try {/*from   w w w .j  a  v  a  2s  . co  m*/
        SolrQuery solrQuery = query.getCopy();
        solrQuery.setRows(nrOfTimeSeriesPerBatch);
        solrQuery.setStart(currentDocumentCount);

        solrStreamingHandler.init(nrOfTimeSeriesPerBatch, currentDocumentCount);
        QueryResponse response = connection.queryAndStreamResponse(solrQuery, solrStreamingHandler);

        nrOfAvailableTimeSeries = response.getResults().getNumFound();
        queryStart = 0;//(long) response.getResponseHeader().get(ChronixSolrStorageConstants.QUERY_START_LONG);
        queryEnd = Long.MAX_VALUE;//(long) response.getResponseHeader().get(ChronixSolrStorageConstants.QUERY_END_LONG);

        needStream = false;
    } catch (SolrServerException | IOException e) {
        LOGGER.error("SolrServerException occurred while querying server.", e);
    }
}

From source file:edu.ku.brc.sgr.MatchableIndexedId.java

License:Open Source License

@Override
public MatchResults doMatch(SolrServer server, SolrQuery baseQuery) {
    final SolrQuery query = baseQuery.getCopy();
    query.setQuery("id:" + id);

    final QueryResponse resp;
    try {/*from  w  w  w  . j a v a  2  s  .c o  m*/
        resp = server.query(query);
    } catch (SolrServerException e) {
        // TODO Auto-generated catch block
        throw new RuntimeException(e);
    }

    final SolrDocumentList docs = resp.getResults();
    final float maxScore = (docs != null) ? docs.getMaxScore() : 0.0f;
    final ImmutableList.Builder<Match> msBuilder = ImmutableList.builder();
    if (docs != null) {
        for (SolrDocument doc : docs) {
            float score = (Float) doc.getFieldValue("score");
            SGRRecord match = SGRRecord.fromSolrDocument(doc);
            msBuilder.add(new Match(match, score));
        }
    }
    return new MatchResults(id, resp.getQTime(), maxScore, msBuilder.build());
}

From source file:edu.ku.brc.sgr.MatchableRecord.java

License:Open Source License

@Override
public MatchResults doMatch(SolrServer server, SolrQuery baseQuery) {
    final ModifiableSolrParams query = baseQuery.getCopy().set(MoreLikeThisParams.DOC_SUPPLIED, true);

    final MoreLikeThisDocumentRequest req = new MoreLikeThisDocumentRequest(query)
            .addDocument(record.asSolrDocument());

    final QueryResponse resp;
    try {/*  w  w w .  j a  v  a2 s.  c o m*/
        resp = req.process(server);
    } catch (Exception e) {
        // TODO Auto-generated catch block
        throw new RuntimeException(e);
    }

    final SolrDocumentList docs = resp.getResults();
    Map<String, String> explain = resp.getExplainMap();
    if (explain == null)
        explain = Maps.newHashMap();

    final ImmutableList.Builder<Match> msBuilder = ImmutableList.builder();
    final float maxScore = (docs != null) ? docs.getMaxScore() : 0.0f;
    if (docs != null) {
        for (SolrDocument doc : docs) {
            float score = (Float) doc.getFieldValue("score");
            SGRRecord match = SGRRecord.fromSolrDocument(doc);
            String explained = explain.get(match.id);
            msBuilder.add(new Match(match, score, explained));
        }
    }
    return new MatchResults(record.id, resp.getQTime(), maxScore, msBuilder.build());
}

From source file:edu.unc.lib.dl.ui.service.SolrQueryLayerService.java

License:Apache License

public void getChildrenCounts(List<BriefObjectMetadata> resultList, AccessGroupSet accessGroups,
        String countName, String queryAddendum, SolrQuery baseQuery) {
    long startTime = System.currentTimeMillis();
    if (resultList == null || resultList.size() == 0)
        return;//  w ww .java2  s .  c o m

    String ancestorPathField = solrSettings.getFieldName(SearchFieldKeys.ANCESTOR_PATH.name());
    SolrQuery solrQuery;
    if (baseQuery == null) {
        // Create a base query since we didn't receive one
        solrQuery = new SolrQuery();
        StringBuilder query = new StringBuilder("*:*");
        try {
            // Add access restrictions to query
            addAccessRestrictions(query, accessGroups);
        } catch (AccessRestrictionException e) {
            // If the user doesn't have any access groups, they don't have access to anything, return null.
            LOG.error(e.getMessage());
            return;
        }

        solrQuery.setStart(0);
        solrQuery.setRows(0);

        solrQuery.setQuery(query.toString());
    } else {
        // Starting from a base query
        solrQuery = baseQuery.getCopy();
        // Make sure we aren't returning any normal results
        solrQuery.setRows(0);
        // Remove all facet fields so we are only getting ancestor path
        if (solrQuery.getFacetFields() != null) {
            for (String facetField : solrQuery.getFacetFields()) {
                solrQuery.removeFacetField(facetField);
            }
        }
    }

    if (queryAddendum != null) {
        solrQuery.setQuery(solrQuery.getQuery() + " AND " + queryAddendum);
    }

    solrQuery.setFacet(true);
    solrQuery.setFacetMinCount(1);
    solrQuery.addFacetField(ancestorPathField);

    Integer countPageSize;
    try {
        countPageSize = new Integer(searchSettings.getProperty("search.facet.countPageSize"));
    } catch (NumberFormatException e) {
        countPageSize = 20;
    }

    solrQuery.add("f." + ancestorPathField + ".facet.limit", Integer.toString(Integer.MAX_VALUE));
    // Sort by value rather than count so that earlier tiers will come first in case the result gets cut off
    solrQuery.setFacetSort("index");

    java.util.Map<Integer, StringBuilder> tierQueryMap = new java.util.HashMap<Integer, StringBuilder>();
    java.util.Map<Integer, List<BriefObjectMetadata>> containerMap = new java.util.HashMap<Integer, List<BriefObjectMetadata>>();

    // Pare the list of ids we are searching for and assigning counts to down to just containers
    for (BriefObjectMetadata metadataObject : resultList) {
        if (metadataObject.getPath() != null && metadataObject.getContentModel() != null
                && metadataObject.getContentModel().contains(ContentModelHelper.Model.CONTAINER.toString())) {
            CutoffFacetNode highestTier = metadataObject.getPath().getHighestTierNode();
            StringBuilder tierQuery = tierQueryMap.get(highestTier.getTier());
            List<BriefObjectMetadata> containerObjects = containerMap.get(highestTier.getTier());
            if (tierQuery == null) {
                tierQuery = new StringBuilder();
                tierQueryMap.put(highestTier.getTier(), tierQuery);

                containerObjects = new ArrayList<BriefObjectMetadata>();
                containerMap.put(highestTier.getTier(), containerObjects);
            }

            if (tierQuery.length() == 0) {
                tierQuery.append(ancestorPathField).append(":(");
            } else {
                tierQuery.append(" OR ");
            }

            tierQuery.append(SolrSettings.sanitize(highestTier.getSearchValue())).append(",*");
            containerObjects.add(metadataObject);

            // If there are a lot of results, then do a partial lookup
            if (containerObjects.size() >= countPageSize) {
                tierQuery.append(")");
                this.executeChildrenCounts(tierQuery, containerObjects, solrQuery, countName,
                        highestTier.getTier());
                LOG.info("Partial query done at " + System.currentTimeMillis() + " ("
                        + (System.currentTimeMillis() - startTime) + ")");
                containerMap.remove(highestTier.getTier());
                tierQueryMap.remove(highestTier.getTier());
            }
        }
    }

    Iterator<java.util.Map.Entry<Integer, StringBuilder>> queryIt = tierQueryMap.entrySet().iterator();
    while (queryIt.hasNext()) {
        java.util.Map.Entry<Integer, StringBuilder> tierQueryEntry = queryIt.next();
        tierQueryEntry.getValue().append(')');
        this.executeChildrenCounts(tierQueryEntry.getValue(), containerMap.get(tierQueryEntry.getKey()),
                solrQuery, countName, tierQueryEntry.getKey());
    }
    LOG.info("Child count query done at " + System.currentTimeMillis() + " ("
            + (System.currentTimeMillis() - startTime) + ")");
}

From source file:edu.unc.lib.dl.ui.service.SolrQueryLayerService.java

License:Apache License

/**
 * Retrieves results for populating a hierarchical browse view. Supports all the regular navigation available to
 * searches. Results contain child counts for each item (all items returned are containers), and a map containing the
 * number of nested subcontainers per container. Children counts are retrieved based on facet counts.
 * /*from  ww w. j  ava2s.  c  o  m*/
 * @param browseRequest
 * @return
 */
public HierarchicalBrowseResultResponse getHierarchicalBrowseResults(HierarchicalBrowseRequest browseRequest) {
    AccessGroupSet accessGroups = browseRequest.getAccessGroups();
    SearchState browseState = (SearchState) browseRequest.getSearchState().clone();
    HierarchicalBrowseResultResponse browseResults = new HierarchicalBrowseResultResponse();

    CutoffFacet rootPath = null;
    BriefObjectMetadataBean rootNode = null;
    if (browseRequest.getRootPid() != null) {
        rootNode = getObjectById(
                new SimpleIdRequest(browseRequest.getRootPid(), browseRequest.getAccessGroups()));
        if (rootNode != null) {
            rootPath = rootNode.getPath();
            browseState.getFacets().put(SearchFieldKeys.ANCESTOR_PATH.name(), rootPath);
            browseResults.setSelectedContainer(rootNode);
        }
    }
    // Default the ancestor path to the collections object so we always have a root
    if (rootNode == null) {
        rootPath = (CutoffFacet) browseState.getFacets().get(SearchFieldKeys.ANCESTOR_PATH.name());
        if (rootPath == null) {
            rootPath = new CutoffFacet(SearchFieldKeys.ANCESTOR_PATH.name(),
                    "1," + this.collectionsPid.getPid());
            browseState.getFacets().put(SearchFieldKeys.ANCESTOR_PATH.name(), rootPath);
        }

        rootNode = getObjectById(new SimpleIdRequest(rootPath.getSearchKey(), browseRequest.getAccessGroups()));
    }
    boolean rootIsAStub = rootNode == null;
    if (rootIsAStub) {
        // Parent is not found, but children are, so make a stub for the parent.
        rootNode = new BriefObjectMetadataBean();
        rootNode.setId(rootPath.getSearchKey());
        rootNode.setAncestorPathFacet(rootPath);
    }

    SearchState hierarchyState = searchStateFactory.createHierarchyListSearchState();
    // Use the ancestor path facet from the state where we will have set a default value
    hierarchyState.getFacets().put(SearchFieldKeys.ANCESTOR_PATH.name(), rootPath);
    hierarchyState.setRowsPerPage(0);

    SearchRequest hierarchyRequest = new SearchRequest(hierarchyState, accessGroups, false);

    SolrQuery baseQuery = this.generateSearch(hierarchyRequest);
    // Get the set of all applicable containers
    SolrQuery hierarchyQuery = baseQuery.getCopy();
    hierarchyQuery.setRows(new Integer(searchSettings.getProperty("search.results.maxBrowsePerPage")));

    // Reusable query segment for limiting the results to just the depth asked for
    StringBuilder cutoffQuery = new StringBuilder();
    cutoffQuery.append('!').append(solrSettings.getFieldName(SearchFieldKeys.ANCESTOR_PATH.name())).append(":");
    cutoffQuery.append(rootPath.getHighestTier() + browseRequest.getRetrievalDepth());
    cutoffQuery.append(searchSettings.facetSubfieldDelimiter).append('*');
    hierarchyQuery.addFilterQuery(cutoffQuery.toString());

    SearchResultResponse results;
    try {
        results = this.executeSearch(hierarchyQuery, hierarchyState, false, false);
        browseResults.setSearchResultResponse(results);
    } catch (SolrServerException e) {
        LOG.error("Error while getting container results for hierarchical browse results", e);
        return null;
    }
    // Add the root node into the result set
    browseResults.getResultList().add(0, rootNode);

    if (browseRequest.isRetrieveFacets() && browseRequest.getSearchState().getFacetsToRetrieve() != null) {
        SearchState facetState = (SearchState) browseState.clone();
        facetState.setRowsPerPage(0);
        SearchRequest facetRequest = new SearchRequest(facetState, browseRequest.getAccessGroups(), true);
        SolrQuery facetQuery = this.generateSearch(facetRequest);
        try {
            SearchResultResponse facetResponse = this.executeSearch(facetQuery, facetState, true, false);
            browseResults.setFacetFields(facetResponse.getFacetFields());
        } catch (SolrServerException e) {
            LOG.warn("Failed to retrieve facet results for " + facetQuery.toString(), e);
        }
    }

    // Don't need to manipulate the container list any further unless either the root is a real record or there are
    // subcontainers
    if (!rootIsAStub || results.getResultCount() > 0) {
        // Get the children counts per container
        SearchRequest filteredChildrenRequest = new SearchRequest(browseState, browseRequest.getAccessGroups(),
                true);
        this.getChildrenCounts(results.getResultList(), accessGroups, "child", null,
                this.generateSearch(filteredChildrenRequest));

        this.getChildrenCounts(results.getResultList(), accessGroups, "containers",
                "contentModel:" + SolrSettings.sanitize(ContentModelHelper.Model.CONTAINER.toString()),
                this.generateSearch(filteredChildrenRequest));

        try {
            // If anything that constituted a search is in the request then trim out possible empty folders
            if (browseState.getFacets().size() > 1 || browseState.getRangeFields().size() > 0
                    || browseState.getSearchFields().size() > 0 || browseState.getAccessTypeFilter() != null) {
                // Get the list of any direct matches for the current query
                browseResults
                        .setMatchingContainerPids(this.getDirectContainerMatches(browseState, accessGroups));
                // Remove all containers that are not direct matches for the user's query and have 0 children
                browseResults.removeContainersWithoutContents();
            }
        } catch (SolrServerException e) {
            LOG.error("Error while getting children counts for hierarchical browse", e);
            return null;
        }
    }

    // Retrieve normal item search results, which are restricted to a max number per page
    if (browseRequest.isIncludeFiles() && browseState.getRowsPerPage() > 0) {
        browseState.getResourceTypes().add(searchSettings.resourceTypeFile);
        SearchState fileSearchState = new SearchState(browseState);
        List<String> resourceTypes = new ArrayList<String>();
        resourceTypes.add(searchSettings.resourceTypeFile);
        fileSearchState.setResourceTypes(resourceTypes);
        CutoffFacet ancestorPath = (CutoffFacet) fileSearchState.getFacets()
                .get(SearchFieldKeys.ANCESTOR_PATH.name());
        ancestorPath.setCutoff(rootPath.getHighestTier() + 1);
        fileSearchState.setFacetsToRetrieve(null);
        SearchRequest fileSearchRequest = new SearchRequest(fileSearchState, browseRequest.getAccessGroups());
        SearchResultResponse fileResults = this.getSearchResults(fileSearchRequest);
        browseResults.populateItemResults(fileResults.getResultList());
    }

    browseResults.generateResultTree();

    return browseResults;
}

From source file:eu.europeana.core.BeanQueryModelFactory.java

License:EUPL

SolrQuery addHiddenQueryFilters(SolrQuery solrQuery, Map<String, String[]> params) {
    SolrQuery dCopy = solrQuery.getCopy();
    if (params != null && params.containsKey("hqf")) {
        final String[] hqf_fields = SolrQueryUtil.getFilterQueriesAsPhrases(params.get("hqf"));
        for (String hqf_field : hqf_fields) {
            dCopy.addFilterQuery(hqf_field);
        }/*from  w w w. ja  va  2  s.com*/
    }
    final PortalTheme theme = ThemeFilter.getTheme();
    if (theme != null && !theme.getHiddenQueryFilters().isEmpty()) {
        final String[] themeHqf = SolrQueryUtil
                .getFilterQueriesAsPhrases(theme.getHiddenQueryFilters().split(","));
        for (String qf : themeHqf) {
            dCopy.addFilterQuery(qf);
        }
    }
    dCopy.setFilterQueries(SolrQueryUtil.getFilterQueriesAsOrQueries(dCopy,
            ThemeFilter.getTheme().getRecordDefinition().getFacetMap()));
    return dCopy;
}

From source file:org.apache.nutch.indexwriter.solr.TestSolrJ.java

License:Apache License

/**
 * query the example/*w  w w .  j a v  a 2 s  .c  om*/
 */
@Test
public void testQuery() throws Exception {
    SolrClient client = new HttpSolrClient.Builder(serverUrl).build();

    // Empty the database...
    client.deleteByQuery("*:*");// delete everything!

    // Now add something...
    SolrInputDocument doc = new SolrInputDocument();
    String docID = "1112211111";
    doc.addField("id", docID, 1.0f);
    doc.addField("name", "my name!", 1.0f);

    assertEquals(null, doc.getField("foo"));
    assertTrue(doc.getField("name").getValue() != null);

    UpdateResponse upres = client.add(doc);
    // System.out.println( "ADD:"+upres.getResponse() );
    assertEquals(0, upres.getStatus());

    upres = client.commit(true, true);
    // System.out.println( "COMMIT:"+upres.getResponse() );
    assertEquals(0, upres.getStatus());

    upres = client.optimize(true, true);
    // System.out.println( "OPTIMIZE:"+upres.getResponse() );
    assertEquals(0, upres.getStatus());

    SolrQuery query = new SolrQuery();
    query.setQuery("id:" + docID);
    QueryResponse response = client.query(query);

    assertEquals(docID, response.getResults().get(0).getFieldValue("id"));

    // Now add a few docs for facet testing...
    List<SolrInputDocument> docs = new ArrayList<>();
    SolrInputDocument doc2 = new SolrInputDocument();
    doc2.addField("id", "2", 1.0f);
    doc2.addField("inStock", true, 1.0f);
    doc2.addField("price", 2, 1.0f);
    doc2.addField("timestamp_dt", new java.util.Date(), 1.0f);
    docs.add(doc2);
    SolrInputDocument doc3 = new SolrInputDocument();
    doc3.addField("id", "3", 1.0f);
    doc3.addField("inStock", false, 1.0f);
    doc3.addField("price", 3, 1.0f);
    doc3.addField("timestamp_dt", new java.util.Date(), 1.0f);
    docs.add(doc3);
    SolrInputDocument doc4 = new SolrInputDocument();
    doc4.addField("id", "4", 1.0f);
    doc4.addField("inStock", true, 1.0f);
    doc4.addField("price", 4, 1.0f);
    doc4.addField("timestamp_dt", new java.util.Date(), 1.0f);
    docs.add(doc4);
    SolrInputDocument doc5 = new SolrInputDocument();
    doc5.addField("id", "5", 1.0f);
    doc5.addField("inStock", false, 1.0f);
    doc5.addField("price", 5, 1.0f);
    doc5.addField("timestamp_dt", new java.util.Date(), 1.0f);
    docs.add(doc5);

    upres = client.add(docs);
    // System.out.println( "ADD:"+upres.getResponse() );
    assertEquals(0, upres.getStatus());

    upres = client.commit(true, true);
    // System.out.println( "COMMIT:"+upres.getResponse() );
    assertEquals(0, upres.getStatus());

    upres = client.optimize(true, true);
    // System.out.println( "OPTIMIZE:"+upres.getResponse() );
    assertEquals(0, upres.getStatus());

    query = new SolrQuery("*:*");
    query.addFacetQuery("price:[* TO 2]");
    query.addFacetQuery("price:[2 TO 4]");
    query.addFacetQuery("price:[5 TO *]");
    query.addFacetField("inStock");
    query.addFacetField("price");
    query.addFacetField("timestamp_dt");
    query.removeFilterQuery("inStock:true");

    response = client.query(query);
    assertEquals(0, response.getStatus());
    assertEquals(5, response.getResults().getNumFound());
    assertEquals(3, response.getFacetQuery().size());
    assertEquals(2, response.getFacetField("inStock").getValueCount());
    assertEquals(4, response.getFacetField("price").getValueCount());

    // test a second query, test making a copy of the main query
    SolrQuery query2 = query.getCopy();
    query2.addFilterQuery("inStock:true");
    response = client.query(query2);
    assertEquals(1, query2.getFilterQueries().length);
    assertEquals(0, response.getStatus());
    assertEquals(2, response.getResults().getNumFound());
    assertFalse(query.getFilterQueries() == query2.getFilterQueries());

    // sanity check round tripping of params...
    query = new SolrQuery("foo");
    query.addFilterQuery("{!field f=inStock}true");
    query.addFilterQuery("{!term f=name}hoss");
    query.addFacetQuery("price:[* TO 2]");
    query.addFacetQuery("price:[2 TO 4]");

    response = client.query(query);
    assertTrue("echoed params are not a NamedList: " + response.getResponseHeader().get("params").getClass(),
            response.getResponseHeader().get("params") instanceof NamedList);
    NamedList echo = (NamedList) response.getResponseHeader().get("params");
    List values = null;
    assertEquals("foo", echo.get("q"));
    assertTrue("echoed fq is not a List: " + echo.get("fq").getClass(), echo.get("fq") instanceof List);
    values = (List) echo.get("fq");
    assertEquals(2, values.size());
    assertEquals("{!field f=inStock}true", values.get(0));
    assertEquals("{!term f=name}hoss", values.get(1));
    assertTrue("echoed facet.query is not a List: " + echo.get("facet.query").getClass(),
            echo.get("facet.query") instanceof List);
    values = (List) echo.get("facet.query");
    assertEquals(2, values.size());
    assertEquals("price:[* TO 2]", values.get(0));
    assertEquals("price:[2 TO 4]", values.get(1));
}