Examples with BoundDimFilter org.apache.druid.query.filter.BoundDimFilter used on opensource projects

Example 16 with BoundDimFilter

use of org.apache.druid.query.filter.BoundDimFilter in project druid by druid-io.

the class GroupByQueryRunnerTest method testBySegmentResultsWithAllFiltersWithExtractionFns.

@Test
public void testBySegmentResultsWithAllFiltersWithExtractionFns() {
    String extractionJsFn = "function(str) { return 'super-' + str; }";
    String jsFn = "function(x) { return(x === 'super-mezzanine') }";
    ExtractionFn extractionFn = new JavaScriptExtractionFn(extractionJsFn, false, JavaScriptConfig.getEnabledInstance());
    List<DimFilter> superFilterList = new ArrayList<>();
    superFilterList.add(new SelectorDimFilter("quality", "super-mezzanine", extractionFn));
    superFilterList.add(new InDimFilter("quality", Arrays.asList("not-super-mezzanine", "FOOBAR", "super-mezzanine"), extractionFn));
    superFilterList.add(new BoundDimFilter("quality", "super-mezzanine", "super-mezzanine", false, false, true, extractionFn, StringComparators.ALPHANUMERIC));
    superFilterList.add(new RegexDimFilter("quality", "super-mezzanine", extractionFn));
    superFilterList.add(new SearchQueryDimFilter("quality", new ContainsSearchQuerySpec("super-mezzanine", true), extractionFn));
    superFilterList.add(new JavaScriptDimFilter("quality", jsFn, extractionFn, JavaScriptConfig.getEnabledInstance()));
    DimFilter superFilter = new AndDimFilter(superFilterList);
    GroupByQuery.Builder builder = makeQueryBuilder().setDataSource(QueryRunnerTestHelper.DATA_SOURCE).setInterval("2011-04-02/2011-04-04").setDimensions(new DefaultDimensionSpec("quality", "alias")).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT, new LongSumAggregatorFactory("idx", "index")).setGranularity(new PeriodGranularity(new Period("P1M"), null, null)).setDimFilter(superFilter).overrideContext(ImmutableMap.of(QueryContexts.BY_SEGMENT_KEY, true));
    final GroupByQuery fullQuery = builder.build();
    int segmentCount = 32;
    Result<BySegmentResultValue> singleSegmentResult = new Result<>(DateTimes.of("2011-01-12T00:00:00.000Z"), new BySegmentResultValueClass<>(Collections.singletonList(makeRow(fullQuery, "2011-04-01", "alias", "mezzanine", "rows", 6L, "idx", 4420L)), QueryRunnerTestHelper.SEGMENT_ID.toString(), Intervals.of("2011-04-02T00:00:00.000Z/2011-04-04T00:00:00.000Z")));
    List<Result> bySegmentResults = new ArrayList<>();
    for (int i = 0; i < segmentCount; i++) {
        bySegmentResults.add(singleSegmentResult);
    }
    QueryToolChest toolChest = factory.getToolchest();
    List<QueryRunner<ResultRow>> singleSegmentRunners = new ArrayList<>();
    for (int i = 0; i < segmentCount; i++) {
        singleSegmentRunners.add(toolChest.preMergeQueryDecoration(runner));
    }
    ExecutorService exec = Executors.newCachedThreadPool();
    QueryRunner theRunner = toolChest.postMergeQueryDecoration(new FinalizeResultsQueryRunner<>(toolChest.mergeResults(factory.mergeRunners(Executors.newCachedThreadPool(), singleSegmentRunners)), toolChest));
    TestHelper.assertExpectedObjects(bySegmentResults, theRunner.run(QueryPlus.wrap(fullQuery)), "bySegment-filter");
    exec.shutdownNow();
}

Example 17 with BoundDimFilter

use of org.apache.druid.query.filter.BoundDimFilter in project druid by druid-io.

the class GroupByQueryRunnerTest method testDimFilterHavingSpecWithExtractionFns.

@Test
public void testDimFilterHavingSpecWithExtractionFns() {
    String extractionJsFn = "function(str) { return 'super-' + str; }";
    ExtractionFn extractionFn = new JavaScriptExtractionFn(extractionJsFn, false, JavaScriptConfig.getEnabledInstance());
    String extractionJsFn2 = "function(num) { return num + 10; }";
    ExtractionFn extractionFn2 = new JavaScriptExtractionFn(extractionJsFn2, false, JavaScriptConfig.getEnabledInstance());
    final DimFilterHavingSpec havingSpec = new DimFilterHavingSpec(new OrDimFilter(ImmutableList.of(new BoundDimFilter("rows", "12", null, true, false, null, extractionFn2, StringComparators.NUMERIC), new SelectorDimFilter("idx", "super-217", extractionFn))), null);
    GroupByQuery.Builder builder = makeQueryBuilder().setDataSource(QueryRunnerTestHelper.DATA_SOURCE).setInterval("2011-04-02/2011-04-04").setDimensions(new DefaultDimensionSpec("quality", "alias")).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT, new LongSumAggregatorFactory("idx", "index")).setGranularity(new PeriodGranularity(new Period("P1M"), null, null)).setHavingSpec(havingSpec);
    final GroupByQuery fullQuery = builder.build();
    List<ResultRow> expectedResults = Arrays.asList(makeRow(fullQuery, "2011-04-01", "alias", "business", "rows", 2L, "idx", 217L), makeRow(fullQuery, "2011-04-01", "alias", "mezzanine", "rows", 6L, "idx", 4420L), makeRow(fullQuery, "2011-04-01", "alias", "premium", "rows", 6L, "idx", 4416L));
    TestHelper.assertExpectedObjects(expectedResults, GroupByQueryRunnerTestHelper.runQuery(factory, runner, fullQuery), "extractionfn-havingspec");
}

Example 18 with BoundDimFilter

use of org.apache.druid.query.filter.BoundDimFilter in project druid by druid-io.

the class GroupByQueryRunnerTest method testDimFilterHavingSpec.

@Test
public void testDimFilterHavingSpec() {
    final DimFilterHavingSpec havingSpec = new DimFilterHavingSpec(new AndDimFilter(ImmutableList.of(new OrDimFilter(ImmutableList.of(new BoundDimFilter("rows", "2", null, true, false, null, null, StringComparators.NUMERIC), new SelectorDimFilter("idx", "217", null))), new SelectorDimFilter("__time", String.valueOf(DateTimes.of("2011-04-01").getMillis()), null))), null);
    GroupByQuery.Builder builder = makeQueryBuilder().setDataSource(QueryRunnerTestHelper.DATA_SOURCE).setInterval("2011-04-02/2011-04-04").setDimensions(new DefaultDimensionSpec("quality", "alias")).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT, new LongSumAggregatorFactory("idx", "index"), QueryRunnerTestHelper.INDEX_LONG_MIN, QueryRunnerTestHelper.INDEX_LONG_MAX, QueryRunnerTestHelper.INDEX_DOUBLE_MIN, QueryRunnerTestHelper.INDEX_DOUBLE_MAX, QueryRunnerTestHelper.INDEX_FLOAT_MIN, QueryRunnerTestHelper.INDEX_FLOAT_MAX).setGranularity(new PeriodGranularity(new Period("P1M"), null, null)).setHavingSpec(havingSpec);
    final GroupByQuery fullQuery = builder.build();
    List<ResultRow> expectedResults = Arrays.asList(makeRow(fullQuery, "2011-04-01", "alias", "business", "rows", 2L, "idx", 217L, QueryRunnerTestHelper.LONG_MIN_INDEX_METRIC, 105L, QueryRunnerTestHelper.LONG_MAX_INDEX_METRIC, 112L, QueryRunnerTestHelper.DOUBLE_MIN_INDEX_METRIC, 105.735462D, QueryRunnerTestHelper.DOUBLE_MAX_INDEX_METRIC, 112.987027D, QueryRunnerTestHelper.FLOAT_MIN_INDEX_METRIC, 105.73546F, QueryRunnerTestHelper.FLOAT_MAX_INDEX_METRIC, 112.98703F), makeRow(fullQuery, "2011-04-01", "alias", "mezzanine", "rows", 6L, "idx", 4420L, QueryRunnerTestHelper.LONG_MIN_INDEX_METRIC, 107L, QueryRunnerTestHelper.LONG_MAX_INDEX_METRIC, 1193L, QueryRunnerTestHelper.DOUBLE_MIN_INDEX_METRIC, 107.047773D, QueryRunnerTestHelper.DOUBLE_MAX_INDEX_METRIC, 1193.556278D, QueryRunnerTestHelper.FLOAT_MIN_INDEX_METRIC, 107.047775F, QueryRunnerTestHelper.FLOAT_MAX_INDEX_METRIC, 1193.5563F), makeRow(fullQuery, "2011-04-01", "alias", "premium", "rows", 6L, "idx", 4416L, QueryRunnerTestHelper.LONG_MIN_INDEX_METRIC, 122L, QueryRunnerTestHelper.LONG_MAX_INDEX_METRIC, 1321L, QueryRunnerTestHelper.DOUBLE_MIN_INDEX_METRIC, 122.141707D, QueryRunnerTestHelper.DOUBLE_MAX_INDEX_METRIC, 1321.375057D, QueryRunnerTestHelper.FLOAT_MIN_INDEX_METRIC, 122.14171F, QueryRunnerTestHelper.FLOAT_MAX_INDEX_METRIC, 1321.375F));
    TestHelper.assertExpectedObjects(expectedResults, GroupByQueryRunnerTestHelper.runQuery(factory, runner, fullQuery), "dimfilter-havingspec");
}

Example 19 with BoundDimFilter

use of org.apache.druid.query.filter.BoundDimFilter in project druid by druid-io.

the class CachingClusteredClientTest method testSingleDimensionPruning.

@Test
public void testSingleDimensionPruning() {
    DimFilter filter = new AndDimFilter(new OrDimFilter(new SelectorDimFilter("dim1", "a", null), new BoundDimFilter("dim1", "from", "to", false, false, false, null, StringComparators.LEXICOGRAPHIC)), new AndDimFilter(new InDimFilter("dim2", Arrays.asList("a", "c", "e", "g"), null), new BoundDimFilter("dim2", "aaa", "hi", false, false, false, null, StringComparators.LEXICOGRAPHIC), new BoundDimFilter("dim2", "e", "zzz", true, true, false, null, StringComparators.LEXICOGRAPHIC)));
    final Druids.TimeseriesQueryBuilder builder = Druids.newTimeseriesQueryBuilder().dataSource(DATA_SOURCE).filters(filter).granularity(GRANULARITY).intervals(SEG_SPEC).context(CONTEXT).intervals("2011-01-05/2011-01-10").aggregators(RENAMED_AGGS).postAggregators(RENAMED_POST_AGGS);
    TimeseriesQuery query = builder.randomQueryId().build();
    final Interval interval1 = Intervals.of("2011-01-06/2011-01-07");
    final Interval interval2 = Intervals.of("2011-01-07/2011-01-08");
    final Interval interval3 = Intervals.of("2011-01-08/2011-01-09");
    QueryRunner runner = new FinalizeResultsQueryRunner(getDefaultQueryRunner(), new TimeseriesQueryQueryToolChest());
    final DruidServer lastServer = servers[random.nextInt(servers.length)];
    ServerSelector selector1 = makeMockSingleDimensionSelector(lastServer, "dim1", null, "b", 0);
    ServerSelector selector2 = makeMockSingleDimensionSelector(lastServer, "dim1", "e", "f", 1);
    ServerSelector selector3 = makeMockSingleDimensionSelector(lastServer, "dim1", "hi", "zzz", 2);
    ServerSelector selector4 = makeMockSingleDimensionSelector(lastServer, "dim2", "a", "e", 0);
    ServerSelector selector5 = makeMockSingleDimensionSelector(lastServer, "dim2", null, null, 1);
    ServerSelector selector6 = makeMockSingleDimensionSelector(lastServer, "other", "b", null, 0);
    timeline.add(interval1, "v", new NumberedPartitionChunk<>(0, 3, selector1));
    timeline.add(interval1, "v", new NumberedPartitionChunk<>(1, 3, selector2));
    timeline.add(interval1, "v", new NumberedPartitionChunk<>(2, 3, selector3));
    timeline.add(interval2, "v", new NumberedPartitionChunk<>(0, 2, selector4));
    timeline.add(interval2, "v", new NumberedPartitionChunk<>(1, 2, selector5));
    timeline.add(interval3, "v", new NumberedPartitionChunk<>(0, 1, selector6));
    final Capture<QueryPlus> capture = Capture.newInstance();
    final Capture<ResponseContext> contextCap = Capture.newInstance();
    QueryRunner mockRunner = EasyMock.createNiceMock(QueryRunner.class);
    EasyMock.expect(mockRunner.run(EasyMock.capture(capture), EasyMock.capture(contextCap))).andReturn(Sequences.empty()).anyTimes();
    EasyMock.expect(serverView.getQueryRunner(lastServer)).andReturn(mockRunner).anyTimes();
    EasyMock.replay(serverView);
    EasyMock.replay(mockRunner);
    List<SegmentDescriptor> descriptors = new ArrayList<>();
    descriptors.add(new SegmentDescriptor(interval1, "v", 0));
    descriptors.add(new SegmentDescriptor(interval1, "v", 2));
    descriptors.add(new SegmentDescriptor(interval2, "v", 1));
    descriptors.add(new SegmentDescriptor(interval3, "v", 0));
    MultipleSpecificSegmentSpec expected = new MultipleSpecificSegmentSpec(descriptors);
    runner.run(QueryPlus.wrap(query)).toList();
    Assert.assertEquals(expected, ((TimeseriesQuery) capture.getValue().getQuery()).getQuerySegmentSpec());
}

Example 20 with BoundDimFilter

use of org.apache.druid.query.filter.BoundDimFilter in project druid by druid-io.

the class CachingClusteredClientTest method testNoSegmentPruningForHashPartitionedSegments.

private void testNoSegmentPruningForHashPartitionedSegments(boolean enableSegmentPruning, @Nullable HashPartitionFunction partitionFunction, boolean useEmptyPartitionDimensions) {
    DimFilter filter = new AndDimFilter(new SelectorDimFilter("dim1", "a", null), new BoundDimFilter("dim2", "e", "zzz", true, true, false, null, StringComparators.LEXICOGRAPHIC), // Equivalent filter of dim3 below is InDimFilter("dim3", Arrays.asList("c"), null)
    new AndDimFilter(new InDimFilter("dim3", Arrays.asList("a", "c", "e", "g"), null), new BoundDimFilter("dim3", "aaa", "ddd", false, false, false, null, StringComparators.LEXICOGRAPHIC)));
    final Map<String, Object> context = new HashMap<>(CONTEXT);
    context.put(QueryContexts.SECONDARY_PARTITION_PRUNING_KEY, enableSegmentPruning);
    final Druids.TimeseriesQueryBuilder builder = Druids.newTimeseriesQueryBuilder().dataSource(DATA_SOURCE).filters(filter).granularity(GRANULARITY).intervals(SEG_SPEC).intervals("2011-01-05/2011-01-10").aggregators(RENAMED_AGGS).postAggregators(RENAMED_POST_AGGS).context(context).randomQueryId();
    TimeseriesQuery query = builder.build();
    QueryRunner runner = new FinalizeResultsQueryRunner(getDefaultQueryRunner(), new TimeseriesQueryQueryToolChest());
    final Interval interval1 = Intervals.of("2011-01-06/2011-01-07");
    final Interval interval2 = Intervals.of("2011-01-07/2011-01-08");
    final Interval interval3 = Intervals.of("2011-01-08/2011-01-09");
    final DruidServer lastServer = servers[random.nextInt(servers.length)];
    List<String> partitionDimensions = useEmptyPartitionDimensions ? ImmutableList.of() : ImmutableList.of("dim1");
    final int numPartitions1 = 6;
    for (int i = 0; i < numPartitions1; i++) {
        ServerSelector selector = makeMockHashBasedSelector(lastServer, partitionDimensions, partitionFunction, i, numPartitions1);
        timeline.add(interval1, "v", new NumberedPartitionChunk<>(i, numPartitions1, selector));
    }
    partitionDimensions = useEmptyPartitionDimensions ? ImmutableList.of() : ImmutableList.of("dim2");
    final int numPartitions2 = 3;
    for (int i = 0; i < numPartitions2; i++) {
        ServerSelector selector = makeMockHashBasedSelector(lastServer, partitionDimensions, partitionFunction, i, numPartitions2);
        timeline.add(interval2, "v", new NumberedPartitionChunk<>(i, numPartitions2, selector));
    }
    partitionDimensions = useEmptyPartitionDimensions ? ImmutableList.of() : ImmutableList.of("dim1", "dim3");
    final int numPartitions3 = 4;
    for (int i = 0; i < numPartitions3; i++) {
        ServerSelector selector = makeMockHashBasedSelector(lastServer, partitionDimensions, partitionFunction, i, numPartitions3);
        timeline.add(interval3, "v", new NumberedPartitionChunk<>(i, numPartitions3, selector));
    }
    final Capture<QueryPlus> capture = Capture.newInstance();
    final Capture<ResponseContext> contextCap = Capture.newInstance();
    QueryRunner mockRunner = EasyMock.createNiceMock(QueryRunner.class);
    EasyMock.expect(mockRunner.run(EasyMock.capture(capture), EasyMock.capture(contextCap))).andReturn(Sequences.empty()).anyTimes();
    EasyMock.expect(serverView.getQueryRunner(lastServer)).andReturn(mockRunner).anyTimes();
    EasyMock.replay(serverView);
    EasyMock.replay(mockRunner);
    // Expected to read all segments
    Set<SegmentDescriptor> expcetedDescriptors = new HashSet<>();
    IntStream.range(0, numPartitions1).forEach(i -> expcetedDescriptors.add(new SegmentDescriptor(interval1, "v", i)));
    IntStream.range(0, numPartitions2).forEach(i -> expcetedDescriptors.add(new SegmentDescriptor(interval2, "v", i)));
    IntStream.range(0, numPartitions3).forEach(i -> expcetedDescriptors.add(new SegmentDescriptor(interval3, "v", i)));
    runner.run(QueryPlus.wrap(query)).toList();
    QuerySegmentSpec querySegmentSpec = ((TimeseriesQuery) capture.getValue().getQuery()).getQuerySegmentSpec();
    Assert.assertSame(MultipleSpecificSegmentSpec.class, querySegmentSpec.getClass());
    final Set<SegmentDescriptor> actualDescriptors = new HashSet<>(((MultipleSpecificSegmentSpec) querySegmentSpec).getDescriptors());
    Assert.assertEquals(expcetedDescriptors, actualDescriptors);
}