Examples with Document datawave.query.attributes.Document used on opensource projects

Example 11 with Document

use of datawave.query.attributes.Document in project datawave by NationalSecurityAgency.

the class QueryLogicTestHarness method assertLogicResults.

// =============================================
// assert methods
/**
 * Determines if the correct results were obtained for a query.
 *
 * @param logic
 *            key/value response data
 * @param expected
 *            list of key values expected within response data
 * @param checkers
 *            list of additional validation methods
 */
public void assertLogicResults(BaseQueryLogic<Map.Entry<Key, Value>> logic, Collection<String> expected, List<DocumentChecker> checkers) {
    Set<String> actualResults = new HashSet<>();
    if (log.isDebugEnabled()) {
        log.debug("    ======  expected id(s)  ======");
        for (String e : expected) {
            log.debug("id(" + e + ")");
        }
    }
    for (Map.Entry<Key, Value> entry : logic) {
        if (FinalDocumentTrackingIterator.isFinalDocumentKey(entry.getKey())) {
            continue;
        }
        final Document document = this.deserializer.apply(entry).getValue();
        // check all of the types to ensure that all are keepers as defined in the
        // AttributeFactory class
        int count = 0;
        for (Attribute<? extends Comparable<?>> attribute : document.getAttributes()) {
            if (attribute instanceof TimingMetadata) {
            // ignore
            } else if (attribute instanceof Attributes) {
                Attributes attrs = (Attributes) attribute;
                Collection<Class<?>> types = new HashSet<>();
                for (Attribute<? extends Comparable<?>> attr : attrs.getAttributes()) {
                    count++;
                    if (attr instanceof TypeAttribute) {
                        Type<? extends Comparable<?>> type = ((TypeAttribute<?>) attr).getType();
                        if (Objects.nonNull(type)) {
                            types.add(type.getClass());
                        }
                    }
                }
                Assert.assertEquals(AttributeFactory.getKeepers(types), types);
            } else {
                count++;
            }
        }
        // ignore empty documents (possible when only passing FinalDocument back)
        if (count == 0) {
            continue;
        }
        // parse the document
        String extractedResult = this.parser.parse(entry.getKey(), document);
        log.debug("result(" + extractedResult + ") key(" + entry.getKey() + ") document(" + document + ")");
        // verify expected results
        Assert.assertNotNull("extracted result", extractedResult);
        Assert.assertFalse("duplicate result(" + extractedResult + ") key(" + entry.getKey() + ")", actualResults.contains(extractedResult));
        actualResults.add(extractedResult);
        // perform any custom assert checks on document
        for (final DocumentChecker check : checkers) {
            check.assertValid(document);
        }
    }
    log.info("total records found(" + actualResults.size() + ") expected(" + expected.size() + ")");
    // ensure that the complete expected result set exists
    if (expected.size() > actualResults.size()) {
        final Set<String> notFound = new HashSet<>(expected);
        notFound.removeAll(actualResults);
        for (final String m : notFound) {
            log.error("missing result(" + m + ")");
        }
    } else if (expected.size() < actualResults.size()) {
        final Set<String> extra = new HashSet<>(actualResults);
        extra.removeAll(expected);
        for (final String r : extra) {
            log.error("unexpected result(" + r + ")");
        }
    }
    Assert.assertEquals("results do not match expected", expected.size(), actualResults.size());
    Assert.assertTrue("expected and actual values do not match", expected.containsAll(actualResults));
    Assert.assertTrue("expected and actual values do not match", actualResults.containsAll(expected));
}

Example 12 with Document

use of datawave.query.attributes.Document in project datawave by NationalSecurityAgency.

the class CompositeFunctionsTest method runTestQuery.

protected void runTestQuery(List<String> expected, String querystr, Date startDate, Date endDate, Map<String, String> extraParms, Connector connector, ShardQueryLogic logic) throws Exception {
    log.debug("runTestQuery");
    log.trace("Creating QueryImpl");
    QueryImpl settings = new QueryImpl();
    settings.setBeginDate(startDate);
    settings.setEndDate(endDate);
    settings.setPagesize(Integer.MAX_VALUE);
    settings.setQueryAuthorizations(auths.serialize());
    settings.setQuery(querystr);
    settings.setParameters(extraParms);
    settings.setId(UUID.randomUUID());
    log.debug("query: " + settings.getQuery());
    log.debug("logic: " + settings.getQueryLogicName());
    logic.setMaxEvaluationPipelines(1);
    GenericQueryConfiguration config = logic.initialize(connector, settings, authSet);
    logic.setupQuery(config);
    HashSet<String> expectedSet = new HashSet<>(expected);
    HashSet<String> resultSet;
    resultSet = new HashSet<>();
    Set<Document> docs = new HashSet<>();
    for (Entry<Key, Value> entry : logic) {
        Document d = deserializer.apply(entry).getValue();
        log.debug(entry.getKey() + " => " + d);
        Attribute<?> attr = d.get("UUID");
        if (attr == null) {
            attr = d.get("UUID.0");
        }
        Assert.assertNotNull("Result Document did not contain a 'UUID'", attr);
        Assert.assertTrue("Expected result to be an instance of DatwawaveTypeAttribute, was: " + attr.getClass().getName(), attr instanceof TypeAttribute || attr instanceof PreNormalizedAttribute);
        TypeAttribute<?> UUIDAttr = (TypeAttribute<?>) attr;
        String UUID = UUIDAttr.getType().getDelegate().toString();
        Assert.assertTrue("Received unexpected UUID: " + UUID, expected.contains(UUID));
        resultSet.add(UUID);
        docs.add(d);
    }
    if (expected.size() > resultSet.size()) {
        expectedSet.addAll(expected);
        expectedSet.removeAll(resultSet);
        for (String s : expectedSet) {
            log.warn("Missing: " + s);
        }
    }
    if (!expected.containsAll(resultSet)) {
        log.error("Expected results " + expected + " differ form actual results " + resultSet);
    }
    Assert.assertTrue("Expected results " + expected + " differ form actual results " + resultSet, expected.containsAll(resultSet));
    Assert.assertEquals("Unexpected number of records", expected.size(), resultSet.size());
}

Example 13 with Document

use of datawave.query.attributes.Document in project datawave by NationalSecurityAgency.

the class HitsAreAlwaysIncludedCommonalityTokenTest method runTestQuery.

protected void runTestQuery(Connector connector, String queryString, Date startDate, Date endDate, Map<String, String> extraParms, Collection<String> goodResults) throws Exception {
    QueryImpl settings = new QueryImpl();
    settings.setBeginDate(startDate);
    settings.setEndDate(endDate);
    settings.setPagesize(Integer.MAX_VALUE);
    settings.setQueryAuthorizations(auths.serialize());
    settings.setQuery(queryString);
    settings.setParameters(extraParms);
    settings.setId(UUID.randomUUID());
    log.debug("query: " + settings.getQuery());
    log.debug("logic: " + settings.getQueryLogicName());
    GenericQueryConfiguration config = logic.initialize(connector, settings, authSet);
    logic.setupQuery(config);
    Set<Document> docs = new HashSet<>();
    for (Entry<Key, Value> entry : logic) {
        Document d = deserializer.apply(entry).getValue();
        log.trace(entry.getKey() + " => " + d);
        docs.add(d);
        Attribute hitAttribute = d.get(JexlEvaluation.HIT_TERM_FIELD);
        if (hitAttribute instanceof Attributes) {
            Attributes attributes = (Attributes) hitAttribute;
            for (Attribute attr : attributes.getAttributes()) {
                if (attr instanceof Content) {
                    Content content = (Content) attr;
                    Assert.assertTrue(goodResults.contains(content.getContent()));
                }
            }
        } else if (hitAttribute instanceof Content) {
            Content content = (Content) hitAttribute;
            Assert.assertTrue(goodResults.contains(content.getContent()));
        }
        // remove from goodResults as we find the expected return fields
        log.debug("goodResults: " + goodResults);
        Map<String, Attribute<? extends Comparable<?>>> dictionary = d.getDictionary();
        log.debug("dictionary:" + dictionary);
        for (Entry<String, Attribute<? extends Comparable<?>>> dictionaryEntry : dictionary.entrySet()) {
            Attribute<? extends Comparable<?>> attribute = dictionaryEntry.getValue();
            if (attribute instanceof Attributes) {
                for (Attribute attr : ((Attributes) attribute).getAttributes()) {
                    String toFind = dictionaryEntry.getKey() + ":" + attr;
                    boolean found = goodResults.remove(toFind);
                    if (found)
                        log.debug("removed " + toFind);
                    else
                        log.debug("Did not remove " + toFind);
                }
            } else {
                String toFind = dictionaryEntry.getKey() + ":" + dictionaryEntry.getValue();
                boolean found = goodResults.remove(toFind);
                if (found)
                    log.debug("removed " + toFind);
                else
                    log.debug("Did not remove " + toFind);
            }
        }
        Assert.assertTrue(goodResults + " was not empty", goodResults.isEmpty());
    }
    Assert.assertTrue("No docs were returned!", !docs.isEmpty());
}

Example 14 with Document

use of datawave.query.attributes.Document in project datawave by NationalSecurityAgency.

the class IvaratorInterruptTest method runTestQuery.

protected void runTestQuery(List<String> expected, String querystr, Date startDate, Date endDate, Map<String, String> extraParms) throws Exception {
    log.debug("runTestQuery");
    log.trace("Creating QueryImpl");
    QueryImpl settings = new QueryImpl();
    settings.setBeginDate(startDate);
    settings.setEndDate(endDate);
    settings.setPagesize(Integer.MAX_VALUE);
    settings.setQueryAuthorizations(auths.serialize());
    settings.setQuery(querystr);
    settings.setParameters(extraParms);
    settings.setId(UUID.randomUUID());
    log.debug("query: " + settings.getQuery());
    log.debug("logic: " + settings.getQueryLogicName());
    logic.setMaxEvaluationPipelines(1);
    GenericQueryConfiguration config = logic.initialize(connector, settings, authSet);
    logic.setupQuery(config);
    HashSet<String> expectedSet = new HashSet<>(expected);
    HashSet<String> resultSet;
    resultSet = new HashSet<>();
    Set<Document> docs = new HashSet<>();
    for (Map.Entry<Key, Value> entry : logic) {
        Document d = deserializer.apply(entry).getValue();
        log.debug(entry.getKey() + " => " + d);
        Attribute<?> attr = d.get("UUID");
        if (attr == null)
            attr = d.get("UUID.0");
        Assert.assertNotNull("Result Document did not contain a 'UUID'", attr);
        Assert.assertTrue("Expected result to be an instance of DatwawaveTypeAttribute, was: " + attr.getClass().getName(), attr instanceof TypeAttribute || attr instanceof PreNormalizedAttribute);
        TypeAttribute<?> UUIDAttr = (TypeAttribute<?>) attr;
        String UUID = UUIDAttr.getType().getDelegate().toString();
        Assert.assertTrue("Received unexpected UUID: " + UUID, expected.contains(UUID));
        resultSet.add(UUID);
        docs.add(d);
    }
    if (expected.size() > resultSet.size()) {
        expectedSet.addAll(expected);
        expectedSet.removeAll(resultSet);
        for (String s : expectedSet) {
            log.warn("Missing: " + s);
        }
    }
    if (!expected.containsAll(resultSet)) {
        log.error("Expected results " + expected + " differ form actual results " + resultSet);
    }
    Assert.assertTrue("Expected results " + expected + " differ form actual results " + resultSet, expected.containsAll(resultSet));
    Assert.assertEquals("Unexpected number of records", expected.size(), resultSet.size());
}

Example 15 with Document

use of datawave.query.attributes.Document in project datawave by NationalSecurityAgency.

the class QueryIterator method seek.

@Override
public void seek(Range range, Collection<ByteSequence> columnFamilies, boolean inclusive) throws IOException {
    // preserve the original range for use with the Final Document tracking iterator because it is placed after the ResultCountingIterator
    // so the FinalDocumentTracking iterator needs the start key with the count already appended
    originalRange = range;
    getActiveQueryLog().get(getQueryId()).beginCall(this.originalRange, ActiveQuery.CallType.SEEK);
    Span span = Trace.start("QueryIterator.seek");
    if (!this.isIncludeGroupingContext() && (this.query.contains("grouping:") || this.query.contains("matchesInGroup") || this.query.contains("MatchesInGroup") || this.query.contains("atomValuesMatch"))) {
        this.setIncludeGroupingContext(true);
        this.groupingContextAddedByMe = true;
    } else {
        this.groupingContextAddedByMe = false;
    }
    try {
        if (log.isDebugEnabled()) {
            log.debug("Seek range: " + range + " " + query);
        }
        this.range = range;
        // determine whether this is a teardown/rebuild range
        long resultCount = 0;
        if (!range.isStartKeyInclusive()) {
            // see if we can fail fast. If we were rebuilt with the FinalDocument key, then we are already completely done
            if (collectTimingDetails && FinalDocumentTrackingIterator.isFinalDocumentKey(range.getStartKey())) {
                this.seekKeySource = new EmptyTreeIterable();
                this.serializedDocuments = EmptyIterator.emptyIterator();
                prepareKeyValue(span);
                return;
            }
            // see if we have a count in the cf
            Key startKey = range.getStartKey();
            String[] parts = StringUtils.split(startKey.getColumnFamily().toString(), '\0');
            if (parts.length == 3) {
                resultCount = NumericalEncoder.decode(parts[0]).longValue();
                // remove the count from the range
                startKey = new Key(startKey.getRow(), new Text(parts[1] + '\0' + parts[2]), startKey.getColumnQualifier(), startKey.getColumnVisibility(), startKey.getTimestamp());
                this.range = range = new Range(startKey, range.isStartKeyInclusive(), range.getEndKey(), range.isEndKeyInclusive());
            }
        }
        // determine whether this is a document specific range
        Range documentRange = isDocumentSpecificRange(range) ? range : null;
        // is done
        if (documentRange != null && !documentRange.isStartKeyInclusive()) {
            if (log.isTraceEnabled()) {
                log.trace("Received non-inclusive event specific range: " + documentRange);
            }
            if (gatherTimingDetails()) {
                this.seekKeySource = new EvaluationTrackingNestedIterator(QuerySpan.Stage.EmptyTree, trackingSpan, new EmptyTreeIterable(), myEnvironment);
            } else {
                this.seekKeySource = new EmptyTreeIterable();
            }
        } else // if the Range is for a single document and the query doesn't reference any index-only or tokenized fields
        if (documentRange != null && (!this.isContainsIndexOnlyTerms() && this.getTermFrequencyFields().isEmpty() && !super.mustUseFieldIndex)) {
            if (log.isTraceEnabled()) {
                log.trace("Received event specific range: " + documentRange);
            }
            // We can take a shortcut to the directly to the event
            Map.Entry<Key, Document> documentKey = Maps.immutableEntry(super.getDocumentKey.apply(documentRange), new Document());
            if (log.isTraceEnabled()) {
                log.trace("Transformed document key: " + documentKey);
            }
            if (gatherTimingDetails()) {
                this.seekKeySource = new EvaluationTrackingNestedIterator(QuerySpan.Stage.DocumentSpecificTree, trackingSpan, new DocumentSpecificNestedIterator(documentKey), myEnvironment);
            } else {
                this.seekKeySource = new DocumentSpecificNestedIterator(documentKey);
            }
        } else {
            this.seekKeySource = buildDocumentIterator(documentRange, range, columnFamilies, inclusive);
        }
        // Create the pipeline iterator for document aggregation and
        // evaluation within a thread pool
        PipelineIterator pipelineIter = PipelineFactory.createIterator(this.seekKeySource, getMaxEvaluationPipelines(), getMaxPipelineCachedResults(), getSerialPipelineRequest(), querySpanCollector, trackingSpan, this, sourceForDeepCopies.deepCopy(myEnvironment), myEnvironment, yield, yieldThresholdMs, columnFamilies, inclusive);
        pipelineIter.setCollectTimingDetails(collectTimingDetails);
        // TODO pipelineIter.setStatsdHostAndPort(statsdHostAndPort);
        pipelineIter.startPipeline();
        // gather Key,Document Entries from the pipelines
        Iterator<Entry<Key, Document>> pipelineDocuments = pipelineIter;
        if (log.isTraceEnabled()) {
            pipelineDocuments = Iterators.filter(pipelineDocuments, keyDocumentEntry -> {
                log.trace("after pipeline, keyDocumentEntry:" + keyDocumentEntry);
                return true;
            });
        }
        // now apply the unique transform if requested
        UniqueTransform uniquify = getUniqueTransform();
        if (uniquify != null) {
            pipelineDocuments = Iterators.filter(pipelineDocuments, uniquify.getUniquePredicate());
        }
        // apply the grouping transform if requested and if the batch size is greater than zero
        // if the batch size is 0, then grouping is computed only on the web server
        GroupingTransform groupify = getGroupingTransform();
        if (groupify != null && this.groupFieldsBatchSize > 0) {
            pipelineDocuments = groupingTransform.getGroupingIterator(pipelineDocuments, this.groupFieldsBatchSize, this.yield);
            if (log.isTraceEnabled()) {
                pipelineDocuments = Iterators.filter(pipelineDocuments, keyDocumentEntry -> {
                    log.trace("after grouping, keyDocumentEntry:" + keyDocumentEntry);
                    return true;
                });
            }
        }
        pipelineDocuments = Iterators.filter(pipelineDocuments, keyDocumentEntry -> {
            // last chance before the documents are serialized
            getActiveQueryLog().get(getQueryId()).recordStats(keyDocumentEntry.getValue(), querySpanCollector.getCombinedQuerySpan(null));
            // Always return true since we just want to record data in the ActiveQueryLog
            return true;
        });
        if (this.getReturnType() == ReturnType.kryo) {
            // Serialize the Document using Kryo
            this.serializedDocuments = Iterators.transform(pipelineDocuments, new KryoDocumentSerializer(isReducedResponse(), isCompressResults()));
        } else if (this.getReturnType() == ReturnType.writable) {
            // Use the Writable interface to serialize the Document
            this.serializedDocuments = Iterators.transform(pipelineDocuments, new WritableDocumentSerializer(isReducedResponse()));
        } else if (this.getReturnType() == ReturnType.tostring) {
            // Just return a toString() representation of the document
            this.serializedDocuments = Iterators.transform(pipelineDocuments, new ToStringDocumentSerializer(isReducedResponse()));
        } else {
            throw new IllegalArgumentException("Unknown return type of: " + this.getReturnType());
        }
        if (log.isTraceEnabled()) {
            KryoDocumentDeserializer dser = new KryoDocumentDeserializer();
            this.serializedDocuments = Iterators.filter(this.serializedDocuments, keyValueEntry -> {
                log.trace("after serializing, keyValueEntry:" + dser.apply(keyValueEntry));
                return true;
            });
        }
        // Cannot do this on document specific ranges as the count would place the keys outside the initial range
        if (!sortedUIDs && documentRange == null) {
            this.serializedDocuments = new ResultCountingIterator(serializedDocuments, resultCount, yield);
        } else if (this.sortedUIDs) {
            // we have sorted UIDs, so we can mask out the cq
            this.serializedDocuments = new KeyAdjudicator<>(serializedDocuments, yield);
        }
        // only add the final document tracking iterator which sends stats back to the client if collectTimingDetails is true
        if (collectTimingDetails) {
            // if there is no document to return, then add an empty document
            // to store the timing metadata
            this.serializedDocuments = new FinalDocumentTrackingIterator(querySpanCollector, trackingSpan, originalRange, this.serializedDocuments, this.getReturnType(), this.isReducedResponse(), this.isCompressResults(), this.yield);
        }
        if (log.isTraceEnabled()) {
            KryoDocumentDeserializer dser = new KryoDocumentDeserializer();
            this.serializedDocuments = Iterators.filter(this.serializedDocuments, keyValueEntry -> {
                log.debug("finally, considering:" + dser.apply(keyValueEntry));
                return true;
            });
        }
        // Determine if we have items to return
        prepareKeyValue(span);
    } catch (Exception e) {
        handleException(e);
    } finally {
        if (gatherTimingDetails() && trackingSpan != null && querySpanCollector != null) {
            querySpanCollector.addQuerySpan(trackingSpan);
        }
        if (null != span) {
            span.stop();
        }
        QueryStatsDClient client = getStatsdClient();
        if (client != null) {
            client.flush();
        }
        getActiveQueryLog().get(getQueryId()).endCall(this.originalRange, ActiveQuery.CallType.SEEK);
        if (this.key == null && this.value == null) {
            // no entries to return
            getActiveQueryLog().remove(getQueryId(), this.originalRange);
        }
    }
}