Examples with DimensionsSpec io.druid.data.input.impl.DimensionsSpec used on opensource projects

Example 1 with DimensionsSpec

use of io.druid.data.input.impl.DimensionsSpec in project druid by druid-io.

the class GroupByQueryHelper method createIndexAccumulatorPair.

public static <T> Pair<IncrementalIndex, Accumulator<IncrementalIndex, T>> createIndexAccumulatorPair(final GroupByQuery query, final GroupByQueryConfig config, StupidPool<ByteBuffer> bufferPool, final boolean combine) {
    final GroupByQueryConfig querySpecificConfig = config.withOverrides(query);
    final Granularity gran = query.getGranularity();
    final long timeStart = query.getIntervals().get(0).getStartMillis();
    long granTimeStart = timeStart;
    if (!(Granularities.ALL.equals(gran))) {
        granTimeStart = gran.bucketStart(new DateTime(timeStart)).getMillis();
    }
    final List<AggregatorFactory> aggs;
    if (combine) {
        aggs = Lists.transform(query.getAggregatorSpecs(), new Function<AggregatorFactory, AggregatorFactory>() {

            @Override
            public AggregatorFactory apply(AggregatorFactory input) {
                return input.getCombiningFactory();
            }
        });
    } else {
        aggs = query.getAggregatorSpecs();
    }
    final List<String> dimensions = Lists.transform(query.getDimensions(), new Function<DimensionSpec, String>() {

        @Override
        public String apply(DimensionSpec input) {
            return input.getOutputName();
        }
    });
    final IncrementalIndex index;
    final boolean sortResults = query.getContextValue(CTX_KEY_SORT_RESULTS, true);
    // All groupBy dimensions are strings, for now.
    final List<DimensionSchema> dimensionSchemas = Lists.newArrayList();
    for (DimensionSpec dimension : query.getDimensions()) {
        dimensionSchemas.add(new StringDimensionSchema(dimension.getOutputName()));
    }
    final IncrementalIndexSchema indexSchema = new IncrementalIndexSchema.Builder().withDimensionsSpec(new DimensionsSpec(dimensionSchemas, null, null)).withMetrics(aggs.toArray(new AggregatorFactory[aggs.size()])).withQueryGranularity(gran).withMinTimestamp(granTimeStart).build();
    if (query.getContextValue("useOffheap", false)) {
        index = new OffheapIncrementalIndex(indexSchema, false, true, sortResults, querySpecificConfig.getMaxResults(), bufferPool);
    } else {
        index = new OnheapIncrementalIndex(indexSchema, false, true, sortResults, querySpecificConfig.getMaxResults());
    }
    Accumulator<IncrementalIndex, T> accumulator = new Accumulator<IncrementalIndex, T>() {

        @Override
        public IncrementalIndex accumulate(IncrementalIndex accumulated, T in) {
            if (in instanceof MapBasedRow) {
                try {
                    MapBasedRow row = (MapBasedRow) in;
                    accumulated.add(new MapBasedInputRow(row.getTimestamp(), dimensions, row.getEvent()));
                } catch (IndexSizeExceededException e) {
                    throw new ResourceLimitExceededException(e.getMessage());
                }
            } else {
                throw new ISE("Unable to accumulate something of type [%s]", in.getClass());
            }
            return accumulated;
        }
    };
    return new Pair<>(index, accumulator);
}

Example 2 with DimensionsSpec

use of io.druid.data.input.impl.DimensionsSpec in project druid by druid-io.

the class RealtimePlumberSchoolTest method setUp.

@Before
public void setUp() throws Exception {
    tmpDir = Files.createTempDir();
    ObjectMapper jsonMapper = new DefaultObjectMapper();
    schema = new DataSchema("test", jsonMapper.convertValue(new StringInputRowParser(new JSONParseSpec(new TimestampSpec("timestamp", "auto", null), new DimensionsSpec(null, null, null), null, null), null), Map.class), new AggregatorFactory[] { new CountAggregatorFactory("rows") }, new UniformGranularitySpec(Granularities.HOUR, Granularities.NONE, null), jsonMapper);
    schema2 = new DataSchema("test", jsonMapper.convertValue(new StringInputRowParser(new JSONParseSpec(new TimestampSpec("timestamp", "auto", null), new DimensionsSpec(null, null, null), null, null), null), Map.class), new AggregatorFactory[] { new CountAggregatorFactory("rows") }, new UniformGranularitySpec(Granularities.YEAR, Granularities.NONE, null), jsonMapper);
    announcer = EasyMock.createMock(DataSegmentAnnouncer.class);
    announcer.announceSegment(EasyMock.<DataSegment>anyObject());
    EasyMock.expectLastCall().anyTimes();
    segmentPublisher = EasyMock.createNiceMock(SegmentPublisher.class);
    dataSegmentPusher = EasyMock.createNiceMock(DataSegmentPusher.class);
    handoffNotifierFactory = EasyMock.createNiceMock(SegmentHandoffNotifierFactory.class);
    handoffNotifier = EasyMock.createNiceMock(SegmentHandoffNotifier.class);
    EasyMock.expect(handoffNotifierFactory.createSegmentHandoffNotifier(EasyMock.anyString())).andReturn(handoffNotifier).anyTimes();
    EasyMock.expect(handoffNotifier.registerSegmentHandoffCallback(EasyMock.<SegmentDescriptor>anyObject(), EasyMock.<Executor>anyObject(), EasyMock.<Runnable>anyObject())).andReturn(true).anyTimes();
    emitter = EasyMock.createMock(ServiceEmitter.class);
    EasyMock.replay(announcer, segmentPublisher, dataSegmentPusher, handoffNotifierFactory, handoffNotifier, emitter);
    tuningConfig = new RealtimeTuningConfig(1, null, null, null, new IntervalStartVersioningPolicy(), rejectionPolicy, null, null, null, buildV9Directly, 0, 0, false, null);
    realtimePlumberSchool = new RealtimePlumberSchool(emitter, new DefaultQueryRunnerFactoryConglomerate(Maps.<Class<? extends Query>, QueryRunnerFactory>newHashMap()), dataSegmentPusher, announcer, segmentPublisher, handoffNotifierFactory, MoreExecutors.sameThreadExecutor(), TestHelper.getTestIndexMerger(), TestHelper.getTestIndexMergerV9(), TestHelper.getTestIndexIO(), MapCache.create(0), FireDepartmentTest.NO_CACHE_CONFIG, TestHelper.getObjectMapper());
    metrics = new FireDepartmentMetrics();
    plumber = (RealtimePlumber) realtimePlumberSchool.findPlumber(schema, tuningConfig, metrics);
}

Example 3 with DimensionsSpec

use of io.druid.data.input.impl.DimensionsSpec in project druid by druid-io.

the class DataSchemaTest method testExplicitInclude.

@Test
public void testExplicitInclude() throws Exception {
    Map<String, Object> parser = jsonMapper.convertValue(new StringInputRowParser(new JSONParseSpec(new TimestampSpec("time", "auto", null), new DimensionsSpec(DimensionsSpec.getDefaultSchemas(ImmutableList.of("time", "dimA", "dimB", "col2")), ImmutableList.of("dimC"), null), null, null), null), new TypeReference<Map<String, Object>>() {
    });
    DataSchema schema = new DataSchema("test", parser, new AggregatorFactory[] { new DoubleSumAggregatorFactory("metric1", "col1"), new DoubleSumAggregatorFactory("metric2", "col2") }, new ArbitraryGranularitySpec(Granularities.DAY, ImmutableList.of(Interval.parse("2014/2015"))), jsonMapper);
    Assert.assertEquals(ImmutableSet.of("dimC", "col1", "metric1", "metric2"), schema.getParser().getParseSpec().getDimensionsSpec().getDimensionExclusions());
}

Example 4 with DimensionsSpec

use of io.druid.data.input.impl.DimensionsSpec in project druid by druid-io.

the class DataSchemaTest method testSerde.

@Test
public void testSerde() throws Exception {
    String jsonStr = "{" + "\"dataSource\":\"test\"," + "\"parser\":{" + "\"type\":\"string\"," + "\"parseSpec\":{" + "\"format\":\"json\"," + "\"timestampSpec\":{\"column\":\"xXx\", \"format\": \"auto\", \"missingValue\": null}," + "\"dimensionsSpec\":{\"dimensions\":[], \"dimensionExclusions\":[]}," + "\"flattenSpec\":{\"useFieldDiscovery\":true, \"fields\":[]}," + "\"featureSpec\":{}}," + "\"encoding\":\"UTF-8\"" + "}," + "\"metricsSpec\":[{\"type\":\"doubleSum\",\"name\":\"metric1\",\"fieldName\":\"col1\"}]," + "\"granularitySpec\":{" + "\"type\":\"arbitrary\"," + "\"queryGranularity\":{\"type\":\"duration\",\"duration\":86400000,\"origin\":\"1970-01-01T00:00:00.000Z\"}," + "\"intervals\":[\"2014-01-01T00:00:00.000Z/2015-01-01T00:00:00.000Z\"]}}";
    DataSchema actual = jsonMapper.readValue(jsonMapper.writeValueAsString(jsonMapper.readValue(jsonStr, DataSchema.class)), DataSchema.class);
    Assert.assertEquals(actual.getDataSource(), "test");
    Assert.assertEquals(actual.getParser().getParseSpec(), new JSONParseSpec(new TimestampSpec("xXx", null, null), new DimensionsSpec(null, Arrays.asList("metric1", "xXx", "col1"), null), null, null));
    Assert.assertEquals(actual.getAggregators(), new AggregatorFactory[] { new DoubleSumAggregatorFactory("metric1", "col1") });
    Assert.assertEquals(actual.getGranularitySpec(), new ArbitraryGranularitySpec(new DurationGranularity(86400000, null), ImmutableList.of(Interval.parse("2014/2015"))));
}

Example 5 with DimensionsSpec

use of io.druid.data.input.impl.DimensionsSpec in project druid by druid-io.

the class DataSchemaTest method testOverlapMetricNameAndDim.

@Test(expected = IAE.class)
public void testOverlapMetricNameAndDim() throws Exception {
    Map<String, Object> parser = jsonMapper.convertValue(new StringInputRowParser(new JSONParseSpec(new TimestampSpec("time", "auto", null), new DimensionsSpec(DimensionsSpec.getDefaultSchemas(ImmutableList.of("time", "dimA", "dimB", "metric1")), ImmutableList.of("dimC"), null), null, null), null), new TypeReference<Map<String, Object>>() {
    });
    DataSchema schema = new DataSchema("test", parser, new AggregatorFactory[] { new DoubleSumAggregatorFactory("metric1", "col1"), new DoubleSumAggregatorFactory("metric2", "col2") }, new ArbitraryGranularitySpec(Granularities.DAY, ImmutableList.of(Interval.parse("2014/2015"))), jsonMapper);
    schema.getParser();
}