Search in sources :

Example 11 with Sequence

use of org.apache.druid.java.util.common.guava.Sequence in project druid by druid-io.

the class GroupByQueryRunnerTest method testMergeResultsWithLimitPushDownSortByAgg.

@Test
public void testMergeResultsWithLimitPushDownSortByAgg() {
    if (!config.getDefaultStrategy().equals(GroupByStrategySelector.STRATEGY_V2)) {
        return;
    }
    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")).setLimitSpec(new DefaultLimitSpec(Collections.singletonList(new OrderByColumnSpec("idx", OrderByColumnSpec.Direction.DESCENDING)), 5)).overrideContext(ImmutableMap.of(GroupByQueryConfig.CTX_KEY_FORCE_LIMIT_PUSH_DOWN, true)).setGranularity(Granularities.ALL);
    final GroupByQuery allGranQuery = builder.build();
    QueryRunner mergedRunner = factory.getToolchest().mergeResults(new QueryRunner<ResultRow>() {

        @Override
        public Sequence<ResultRow> run(QueryPlus<ResultRow> queryPlus, ResponseContext responseContext) {
            // simulate two daily segments
            final QueryPlus<ResultRow> queryPlus1 = queryPlus.withQuery(queryPlus.getQuery().withQuerySegmentSpec(new MultipleIntervalSegmentSpec(Collections.singletonList(Intervals.of("2011-04-02/2011-04-03")))));
            final QueryPlus<ResultRow> queryPlus2 = queryPlus.withQuery(queryPlus.getQuery().withQuerySegmentSpec(new MultipleIntervalSegmentSpec(Collections.singletonList(Intervals.of("2011-04-03/2011-04-04")))));
            return factory.getToolchest().mergeResults((queryPlus3, responseContext1) -> new MergeSequence<>(queryPlus3.getQuery().getResultOrdering(), Sequences.simple(Arrays.asList(runner.run(queryPlus1, responseContext1), runner.run(queryPlus2, responseContext1))))).run(queryPlus, responseContext);
        }
    });
    List<ResultRow> allGranExpectedResults = Arrays.asList(makeRow(allGranQuery, "2011-04-02", "alias", "mezzanine", "rows", 6L, "idx", 4420L), makeRow(allGranQuery, "2011-04-02", "alias", "premium", "rows", 6L, "idx", 4416L), makeRow(allGranQuery, "2011-04-02", "alias", "entertainment", "rows", 2L, "idx", 319L), makeRow(allGranQuery, "2011-04-02", "alias", "automotive", "rows", 2L, "idx", 269L), makeRow(allGranQuery, "2011-04-02", "alias", "travel", "rows", 2L, "idx", 243L));
    Iterable<ResultRow> results = mergedRunner.run(QueryPlus.wrap(allGranQuery)).toList();
    TestHelper.assertExpectedObjects(allGranExpectedResults, results, "merged");
}
Also used : DefaultLimitSpec(org.apache.druid.query.groupby.orderby.DefaultLimitSpec) LongSumAggregatorFactory(org.apache.druid.query.aggregation.LongSumAggregatorFactory) MultipleIntervalSegmentSpec(org.apache.druid.query.spec.MultipleIntervalSegmentSpec) Sequence(org.apache.druid.java.util.common.guava.Sequence) MergeSequence(org.apache.druid.java.util.common.guava.MergeSequence) DefaultDimensionSpec(org.apache.druid.query.dimension.DefaultDimensionSpec) FinalizeResultsQueryRunner(org.apache.druid.query.FinalizeResultsQueryRunner) ChainedExecutionQueryRunner(org.apache.druid.query.ChainedExecutionQueryRunner) QueryRunner(org.apache.druid.query.QueryRunner) OrderByColumnSpec(org.apache.druid.query.groupby.orderby.OrderByColumnSpec) ResponseContext(org.apache.druid.query.context.ResponseContext) QueryPlus(org.apache.druid.query.QueryPlus) InitializedNullHandlingTest(org.apache.druid.testing.InitializedNullHandlingTest) Test(org.junit.Test)

Example 12 with Sequence

use of org.apache.druid.java.util.common.guava.Sequence in project druid by druid-io.

the class GroupByQueryRunnerTest method testMergedPostAggHavingSpec.

@Test
public void testMergedPostAggHavingSpec() {
    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")).setPostAggregatorSpecs(Collections.singletonList(new ArithmeticPostAggregator("rows_times_10", "*", Arrays.asList(new FieldAccessPostAggregator("rows", "rows"), new ConstantPostAggregator("const", 10L))))).setGranularity(new PeriodGranularity(new Period("P1M"), null, null)).setHavingSpec(new OrHavingSpec(ImmutableList.of(new GreaterThanHavingSpec("rows_times_10", 20L), new EqualToHavingSpec("idx", 217L))));
    GroupByQuery query = builder.build();
    // Same query, but with expressions instead of arithmetic.
    final GroupByQuery expressionQuery = query.withPostAggregatorSpecs(Collections.singletonList(new ExpressionPostAggregator("rows_times_10", "rows * 10.0", null, TestExprMacroTable.INSTANCE)));
    List<ResultRow> expectedResults = Arrays.asList(makeRow(query, "2011-04-01", "alias", "business", "rows", 2L, "idx", 217L, "rows_times_10", 20.0), makeRow(query, "2011-04-01", "alias", "mezzanine", "rows", 6L, "idx", 4420L, "rows_times_10", 60.0), makeRow(query, "2011-04-01", "alias", "premium", "rows", 6L, "idx", 4416L, "rows_times_10", 60.0));
    QueryRunner mergedRunner = factory.getToolchest().mergeResults(new QueryRunner<ResultRow>() {

        @Override
        public Sequence<ResultRow> run(QueryPlus<ResultRow> queryPlus, ResponseContext responseContext) {
            // simulate two daily segments
            final QueryPlus queryPlus1 = queryPlus.withQuery(queryPlus.getQuery().withQuerySegmentSpec(new MultipleIntervalSegmentSpec(Collections.singletonList(Intervals.of("2011-04-02/2011-04-03")))));
            final QueryPlus queryPlus2 = queryPlus.withQuery(queryPlus.getQuery().withQuerySegmentSpec(new MultipleIntervalSegmentSpec(Collections.singletonList(Intervals.of("2011-04-03/2011-04-04")))));
            return new MergeSequence(queryPlus.getQuery().getResultOrdering(), Sequences.simple(Arrays.asList(runner.run(queryPlus1, responseContext), runner.run(queryPlus2, responseContext))));
        }
    });
    ResponseContext context = ResponseContext.createEmpty();
    // add an extra layer of merging, simulate broker forwarding query to historical
    TestHelper.assertExpectedObjects(expectedResults, factory.getToolchest().postMergeQueryDecoration(factory.getToolchest().mergeResults(factory.getToolchest().preMergeQueryDecoration(mergedRunner))).run(QueryPlus.wrap(query)), "merged");
    TestHelper.assertExpectedObjects(expectedResults, factory.getToolchest().postMergeQueryDecoration(factory.getToolchest().mergeResults(factory.getToolchest().preMergeQueryDecoration(mergedRunner))).run(QueryPlus.wrap(expressionQuery)), "merged");
}
Also used : EqualToHavingSpec(org.apache.druid.query.groupby.having.EqualToHavingSpec) ArithmeticPostAggregator(org.apache.druid.query.aggregation.post.ArithmeticPostAggregator) FieldAccessPostAggregator(org.apache.druid.query.aggregation.post.FieldAccessPostAggregator) OrHavingSpec(org.apache.druid.query.groupby.having.OrHavingSpec) LongSumAggregatorFactory(org.apache.druid.query.aggregation.LongSumAggregatorFactory) ConstantPostAggregator(org.apache.druid.query.aggregation.post.ConstantPostAggregator) PeriodGranularity(org.apache.druid.java.util.common.granularity.PeriodGranularity) Period(org.joda.time.Period) MultipleIntervalSegmentSpec(org.apache.druid.query.spec.MultipleIntervalSegmentSpec) Sequence(org.apache.druid.java.util.common.guava.Sequence) MergeSequence(org.apache.druid.java.util.common.guava.MergeSequence) DefaultDimensionSpec(org.apache.druid.query.dimension.DefaultDimensionSpec) FinalizeResultsQueryRunner(org.apache.druid.query.FinalizeResultsQueryRunner) ChainedExecutionQueryRunner(org.apache.druid.query.ChainedExecutionQueryRunner) QueryRunner(org.apache.druid.query.QueryRunner) MergeSequence(org.apache.druid.java.util.common.guava.MergeSequence) ExpressionPostAggregator(org.apache.druid.query.aggregation.post.ExpressionPostAggregator) ResponseContext(org.apache.druid.query.context.ResponseContext) GreaterThanHavingSpec(org.apache.druid.query.groupby.having.GreaterThanHavingSpec) QueryPlus(org.apache.druid.query.QueryPlus) InitializedNullHandlingTest(org.apache.druid.testing.InitializedNullHandlingTest) Test(org.junit.Test)

Example 13 with Sequence

use of org.apache.druid.java.util.common.guava.Sequence in project druid by druid-io.

the class GroupByQueryRunnerTest method testMergeResultsWithLimitPushDownSortByDimAggDim.

@Test
public void testMergeResultsWithLimitPushDownSortByDimAggDim() {
    if (!config.getDefaultStrategy().equals(GroupByStrategySelector.STRATEGY_V2)) {
        return;
    }
    GroupByQuery.Builder builder = makeQueryBuilder().setDataSource(QueryRunnerTestHelper.DATA_SOURCE).setInterval("2011-04-02/2011-04-04").setDimensions(new DefaultDimensionSpec("quality", "alias"), new DefaultDimensionSpec("market", "market")).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT, new LongSumAggregatorFactory("idx", "index")).setLimitSpec(new DefaultLimitSpec(Lists.newArrayList(new OrderByColumnSpec("alias", OrderByColumnSpec.Direction.DESCENDING), new OrderByColumnSpec("idx", OrderByColumnSpec.Direction.DESCENDING), new OrderByColumnSpec("market", OrderByColumnSpec.Direction.DESCENDING)), 5)).overrideContext(ImmutableMap.of(GroupByQueryConfig.CTX_KEY_FORCE_LIMIT_PUSH_DOWN, true)).setGranularity(Granularities.ALL);
    final GroupByQuery allGranQuery = builder.build();
    QueryRunner mergedRunner = factory.getToolchest().mergeResults(new QueryRunner<ResultRow>() {

        @Override
        public Sequence<ResultRow> run(QueryPlus<ResultRow> queryPlus, ResponseContext responseContext) {
            // simulate two daily segments
            final QueryPlus<ResultRow> queryPlus1 = queryPlus.withQuery(queryPlus.getQuery().withQuerySegmentSpec(new MultipleIntervalSegmentSpec(Collections.singletonList(Intervals.of("2011-04-02/2011-04-03")))));
            final QueryPlus<ResultRow> queryPlus2 = queryPlus.withQuery(queryPlus.getQuery().withQuerySegmentSpec(new MultipleIntervalSegmentSpec(Collections.singletonList(Intervals.of("2011-04-03/2011-04-04")))));
            return factory.getToolchest().mergeResults((queryPlus3, responseContext1) -> new MergeSequence<>(queryPlus3.getQuery().getResultOrdering(), Sequences.simple(Arrays.asList(runner.run(queryPlus1, responseContext1), runner.run(queryPlus2, responseContext1))))).run(queryPlus, responseContext);
        }
    });
    List<ResultRow> allGranExpectedResults = Arrays.asList(makeRow(allGranQuery, "2011-04-02", "alias", "travel", "market", "spot", "rows", 2L, "idx", 243L), makeRow(allGranQuery, "2011-04-02", "alias", "technology", "market", "spot", "rows", 2L, "idx", 177L), makeRow(allGranQuery, "2011-04-02", "alias", "premium", "market", "total_market", "rows", 2L, "idx", 2342L), makeRow(allGranQuery, "2011-04-02", "alias", "premium", "market", "upfront", "rows", 2L, "idx", 1817L), makeRow(allGranQuery, "2011-04-02", "alias", "premium", "market", "spot", "rows", 2L, "idx", 257L));
    Iterable<ResultRow> results = mergedRunner.run(QueryPlus.wrap(allGranQuery)).toList();
    TestHelper.assertExpectedObjects(allGranExpectedResults, results, "merged");
}
Also used : DefaultLimitSpec(org.apache.druid.query.groupby.orderby.DefaultLimitSpec) LongSumAggregatorFactory(org.apache.druid.query.aggregation.LongSumAggregatorFactory) MultipleIntervalSegmentSpec(org.apache.druid.query.spec.MultipleIntervalSegmentSpec) Sequence(org.apache.druid.java.util.common.guava.Sequence) MergeSequence(org.apache.druid.java.util.common.guava.MergeSequence) DefaultDimensionSpec(org.apache.druid.query.dimension.DefaultDimensionSpec) FinalizeResultsQueryRunner(org.apache.druid.query.FinalizeResultsQueryRunner) ChainedExecutionQueryRunner(org.apache.druid.query.ChainedExecutionQueryRunner) QueryRunner(org.apache.druid.query.QueryRunner) OrderByColumnSpec(org.apache.druid.query.groupby.orderby.OrderByColumnSpec) ResponseContext(org.apache.druid.query.context.ResponseContext) QueryPlus(org.apache.druid.query.QueryPlus) InitializedNullHandlingTest(org.apache.druid.testing.InitializedNullHandlingTest) Test(org.junit.Test)

Example 14 with Sequence

use of org.apache.druid.java.util.common.guava.Sequence in project druid by druid-io.

the class GroupByQueryRunnerTest method doTestMergeResultsWithOrderBy.

private void doTestMergeResultsWithOrderBy(GroupByQuery baseQuery, LimitSpec limitSpec, List<ResultRow> expectedResults) {
    QueryRunner mergedRunner = factory.getToolchest().mergeResults(new QueryRunner<ResultRow>() {

        @Override
        public Sequence<ResultRow> run(QueryPlus<ResultRow> queryPlus, ResponseContext responseContext) {
            // simulate two daily segments
            final QueryPlus queryPlus1 = queryPlus.withQuery(queryPlus.getQuery().withQuerySegmentSpec(new MultipleIntervalSegmentSpec(Collections.singletonList(Intervals.of("2011-04-02/2011-04-03")))));
            final QueryPlus queryPlus2 = queryPlus.withQuery(queryPlus.getQuery().withQuerySegmentSpec(new MultipleIntervalSegmentSpec(Collections.singletonList(Intervals.of("2011-04-03/2011-04-04")))));
            return new MergeSequence(queryPlus.getQuery().getResultOrdering(), Sequences.simple(Arrays.asList(runner.run(queryPlus1, responseContext), runner.run(queryPlus2, responseContext))));
        }
    });
    final GroupByQuery query = baseQuery.withLimitSpec(limitSpec);
    TestHelper.assertExpectedObjects(expectedResults, mergedRunner.run(QueryPlus.wrap(query)), "merged");
}
Also used : MergeSequence(org.apache.druid.java.util.common.guava.MergeSequence) ResponseContext(org.apache.druid.query.context.ResponseContext) MultipleIntervalSegmentSpec(org.apache.druid.query.spec.MultipleIntervalSegmentSpec) Sequence(org.apache.druid.java.util.common.guava.Sequence) MergeSequence(org.apache.druid.java.util.common.guava.MergeSequence) FinalizeResultsQueryRunner(org.apache.druid.query.FinalizeResultsQueryRunner) ChainedExecutionQueryRunner(org.apache.druid.query.ChainedExecutionQueryRunner) QueryRunner(org.apache.druid.query.QueryRunner) QueryPlus(org.apache.druid.query.QueryPlus)

Example 15 with Sequence

use of org.apache.druid.java.util.common.guava.Sequence in project druid by druid-io.

the class CachingClusteredClientTest method testOutOfOrderSequenceMerging.

@Test
public void testOutOfOrderSequenceMerging() {
    List<Sequence<Result<TopNResultValue>>> sequences = ImmutableList.of(Sequences.simple(makeTopNResultsWithoutRename(DateTimes.of("2011-01-07"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, DateTimes.of("2011-01-08"), "a", 50, 4988, "b", 50, 4987, "c", 50, 4986, DateTimes.of("2011-01-09"), "a", 50, 4985, "b", 50, 4984, "c", 50, 4983)), Sequences.simple(makeTopNResultsWithoutRename(DateTimes.of("2011-01-06T01"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, DateTimes.of("2011-01-07T01"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, DateTimes.of("2011-01-08T01"), "a", 50, 4988, "b", 50, 4987, "c", 50, 4986, DateTimes.of("2011-01-09T01"), "a", 50, 4985, "b", 50, 4984, "c", 50, 4983)));
    TestHelper.assertExpectedResults(makeTopNResultsWithoutRename(DateTimes.of("2011-01-06T01"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, DateTimes.of("2011-01-07"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, DateTimes.of("2011-01-07T01"), "a", 50, 4991, "b", 50, 4990, "c", 50, 4989, DateTimes.of("2011-01-08"), "a", 50, 4988, "b", 50, 4987, "c", 50, 4986, DateTimes.of("2011-01-08T01"), "a", 50, 4988, "b", 50, 4987, "c", 50, 4986, DateTimes.of("2011-01-09"), "a", 50, 4985, "b", 50, 4984, "c", 50, 4983, DateTimes.of("2011-01-09T01"), "a", 50, 4985, "b", 50, 4984, "c", 50, 4983), mergeSequences(new TopNQueryBuilder().dataSource("test").intervals("2011-01-06/2011-01-10").dimension("a").metric("b").threshold(3).aggregators(new CountAggregatorFactory("b")).randomQueryId().build(), sequences));
}
Also used : TopNResultValue(org.apache.druid.query.topn.TopNResultValue) TopNQueryBuilder(org.apache.druid.query.topn.TopNQueryBuilder) CountAggregatorFactory(org.apache.druid.query.aggregation.CountAggregatorFactory) Sequence(org.apache.druid.java.util.common.guava.Sequence) Test(org.junit.Test)

Aggregations

Sequence (org.apache.druid.java.util.common.guava.Sequence)102 Test (org.junit.Test)53 List (java.util.List)44 DefaultDimensionSpec (org.apache.druid.query.dimension.DefaultDimensionSpec)37 ResponseContext (org.apache.druid.query.context.ResponseContext)32 ImmutableList (com.google.common.collect.ImmutableList)29 Intervals (org.apache.druid.java.util.common.Intervals)28 Granularities (org.apache.druid.java.util.common.granularity.Granularities)28 QueryRunner (org.apache.druid.query.QueryRunner)28 ArrayList (java.util.ArrayList)27 VirtualColumns (org.apache.druid.segment.VirtualColumns)26 Cursor (org.apache.druid.segment.Cursor)25 QueryPlus (org.apache.druid.query.QueryPlus)24 Result (org.apache.druid.query.Result)24 NullHandling (org.apache.druid.common.config.NullHandling)22 InitializedNullHandlingTest (org.apache.druid.testing.InitializedNullHandlingTest)22 MultipleIntervalSegmentSpec (org.apache.druid.query.spec.MultipleIntervalSegmentSpec)21 QueryableIndexStorageAdapter (org.apache.druid.segment.QueryableIndexStorageAdapter)20 DataSegment (org.apache.druid.timeline.DataSegment)20 ImmutableMap (com.google.common.collect.ImmutableMap)18