Examples with ExpressionVirtualColumn org.apache.druid.segment.virtual.ExpressionVirtualColumn used on opensource projects

Example 11 with ExpressionVirtualColumn

use of org.apache.druid.segment.virtual.ExpressionVirtualColumn in project druid by druid-io.

the class GroupByQueryRunnerTest method testNestedGroupByWithLongArrays.

@Test
public void testNestedGroupByWithLongArrays() {
    if (config.getDefaultStrategy().equals(GroupByStrategySelector.STRATEGY_V1)) {
        expectedException.expect(UnsupportedOperationException.class);
        expectedException.expectMessage("GroupBy v1 only supports dimensions with an outputType of STRING");
    }
    cannotVectorize();
    GroupByQuery inner = makeQueryBuilder().setDataSource(QueryRunnerTestHelper.DATA_SOURCE).setQuerySegmentSpec(QueryRunnerTestHelper.FIRST_TO_THIRD).setVirtualColumns(new ExpressionVirtualColumn("v0", "array(1,2)", ColumnType.LONG_ARRAY, ExprMacroTable.nil())).setDimensions(new DefaultDimensionSpec("v0", "alias", ColumnType.LONG_ARRAY)).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT).setGranularity(QueryRunnerTestHelper.ALL_GRAN).build();
    GroupByQuery outer = makeQueryBuilder().setDataSource(new QueryDataSource(inner)).setQuerySegmentSpec(QueryRunnerTestHelper.FIRST_TO_THIRD).setDimensions(new DefaultDimensionSpec("alias", "alias_outer", ColumnType.LONG_ARRAY)).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT).setGranularity(QueryRunnerTestHelper.ALL_GRAN).build();
    List<ResultRow> expectedResults = ImmutableList.of(makeRow(outer, "2011-04-01", "alias_outer", new ComparableList(ImmutableList.of(1L, 2L)), "rows", 1L));
    Iterable<ResultRow> results = GroupByQueryRunnerTestHelper.runQuery(factory, runner, outer);
    TestHelper.assertExpectedObjects(expectedResults, results, "long-nested-groupby-arrays");
}

Example 12 with ExpressionVirtualColumn

use of org.apache.druid.segment.virtual.ExpressionVirtualColumn in project druid by druid-io.

the class GroupByQueryRunnerTest method testGroupByWithSubtotalsSpecOfDimensionsPrefixes.

@Test
public void testGroupByWithSubtotalsSpecOfDimensionsPrefixes() {
    if (!config.getDefaultStrategy().equals(GroupByStrategySelector.STRATEGY_V2)) {
        return;
    }
    GroupByQuery query = makeQueryBuilder().setDataSource(QueryRunnerTestHelper.DATA_SOURCE).setQuerySegmentSpec(QueryRunnerTestHelper.FIRST_TO_THIRD).setVirtualColumns(new ExpressionVirtualColumn("alias", "quality", ColumnType.STRING, TestExprMacroTable.INSTANCE)).setDimensions(Lists.newArrayList(new DefaultDimensionSpec("market", "market2"), new DefaultDimensionSpec("alias", "alias2"))).setAggregatorSpecs(Arrays.asList(QueryRunnerTestHelper.ROWS_COUNT, new LongSumAggregatorFactory("idx", "index"))).setGranularity(QueryRunnerTestHelper.DAY_GRAN).setSubtotalsSpec(ImmutableList.of(ImmutableList.of("market2"), ImmutableList.of())).build();
    List<ResultRow> expectedResults = Arrays.asList(makeRow(query, "2011-04-01T00:00:00.000Z", "market2", "spot", "rows", 9L, "idx", 1102L), makeRow(query, "2011-04-01T00:00:00.000Z", "market2", "total_market", "rows", 2L, "idx", 2836L), makeRow(query, "2011-04-01T00:00:00.000Z", "market2", "upfront", "rows", 2L, "idx", 2681L), makeRow(query, "2011-04-02T00:00:00.000Z", "market2", "spot", "rows", 9L, "idx", 1120L), makeRow(query, "2011-04-02T00:00:00.000Z", "market2", "total_market", "rows", 2L, "idx", 2514L), makeRow(query, "2011-04-02T00:00:00.000Z", "market2", "upfront", "rows", 2L, "idx", 2193L), makeRow(query, "2011-04-01T00:00:00.000Z", "rows", 13L, "idx", 6619L), makeRow(query, "2011-04-02T00:00:00.000Z", "rows", 13L, "idx", 5827L));
    Iterable<ResultRow> results = GroupByQueryRunnerTestHelper.runQuery(factory, runner, query);
    TestHelper.assertExpectedObjects(expectedResults, results, "subtotal");
}

Example 13 with ExpressionVirtualColumn

use of org.apache.druid.segment.virtual.ExpressionVirtualColumn in project druid by druid-io.

the class GroupByQueryRunnerTest method testDifferentGroupingSubquery.

@Test
public void testDifferentGroupingSubquery() {
    GroupByQuery subquery = makeQueryBuilder().setDataSource(QueryRunnerTestHelper.DATA_SOURCE).setQuerySegmentSpec(QueryRunnerTestHelper.FIRST_TO_THIRD).setDimensions(new DefaultDimensionSpec("quality", "alias")).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT, new LongSumAggregatorFactory("idx", "index"), new LongSumAggregatorFactory("indexMaxPlusTen", "indexMaxPlusTen")).setGranularity(QueryRunnerTestHelper.DAY_GRAN).build();
    GroupByQuery query = makeQueryBuilder().setDataSource(subquery).setQuerySegmentSpec(QueryRunnerTestHelper.FIRST_TO_THIRD).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT, new DoubleMaxAggregatorFactory("idx", "idx"), new DoubleMaxAggregatorFactory("indexMaxPlusTen", "indexMaxPlusTen")).setGranularity(QueryRunnerTestHelper.DAY_GRAN).build();
    List<ResultRow> expectedResults = makeRows(query, new String[] { "__time", "rows", "idx", "indexMaxPlusTen" }, new Object[] { "2011-04-01", 9L, 2900.0, 2930.0 }, new Object[] { "2011-04-02", 9L, 2505.0, 2535.0 });
    TestHelper.assertExpectedObjects(expectedResults, GroupByQueryRunnerTestHelper.runQuery(factory, runner, query), "subquery");
    subquery = makeQueryBuilder(subquery).setVirtualColumns(new ExpressionVirtualColumn("expr", "-index + 100", ColumnType.FLOAT, TestExprMacroTable.INSTANCE)).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT, new LongSumAggregatorFactory("idx", "expr"), new LongSumAggregatorFactory("indexMaxPlusTen", "indexMaxPlusTen")).build();
    query = (GroupByQuery) query.withDataSource(new QueryDataSource(subquery));
    expectedResults = makeRows(query, new String[] { "__time", "rows", "idx", "indexMaxPlusTen" }, new Object[] { "2011-04-01", 9L, 21.0, 2930.0 }, new Object[] { "2011-04-02", 9L, 2.0, 2535.0 });
    TestHelper.assertExpectedObjects(expectedResults, GroupByQueryRunnerTestHelper.runQuery(factory, runner, query), "subquery");
}

Example 14 with ExpressionVirtualColumn

use of org.apache.druid.segment.virtual.ExpressionVirtualColumn in project druid by druid-io.

the class GroupByQueryRunnerTest method testExtractionStringArraySpecWithMultiValueVirtualDimAsInput.

@Test
public void testExtractionStringArraySpecWithMultiValueVirtualDimAsInput() {
    if (config.getDefaultStrategy().equals(GroupByStrategySelector.STRATEGY_V1)) {
        expectedException.expect(UnsupportedOperationException.class);
        expectedException.expectMessage("GroupBy v1 only supports dimensions with an outputType of STRING");
    } else if (!vectorize) {
        expectedException.expect(RuntimeException.class);
        expectedException.expectMessage("Not supported for multi-value dimensions");
    }
    cannotVectorize();
    GroupByQuery query = makeQueryBuilder().setDataSource(QueryRunnerTestHelper.DATA_SOURCE).setQuerySegmentSpec(QueryRunnerTestHelper.FIRST_TO_THIRD).setVirtualColumns(new ExpressionVirtualColumn("v0", "mv_to_array(placementish)", ColumnType.STRING_ARRAY, ExprMacroTable.nil())).setDimensions(new ExtractionDimensionSpec("v0", "alias", ColumnType.STRING_ARRAY, new SubstringDimExtractionFn(1, 1))).setAggregatorSpecs(QueryRunnerTestHelper.ROWS_COUNT, new LongSumAggregatorFactory("idx", "index")).setGranularity(QueryRunnerTestHelper.ALL_GRAN).build();
    GroupByQueryRunnerTestHelper.runQuery(factory, runner, query);
}

Example 15 with ExpressionVirtualColumn

use of org.apache.druid.segment.virtual.ExpressionVirtualColumn in project druid by druid-io.

the class GroupByTimeseriesQueryRunnerTest method constructorFeeder.

@SuppressWarnings("unchecked")
@Parameterized.Parameters(name = "{0}, vectorize = {1}")
public static Iterable<Object[]> constructorFeeder() {
    GroupByQueryConfig config = new GroupByQueryConfig();
    config.setMaxIntermediateRows(10000);
    final Pair<GroupByQueryRunnerFactory, Closer> factoryAndCloser = GroupByQueryRunnerTest.makeQueryRunnerFactory(config);
    final GroupByQueryRunnerFactory factory = factoryAndCloser.lhs;
    RESOURCE_CLOSER.register(factoryAndCloser.rhs);
    final List<Object[]> constructors = new ArrayList<>();
    for (QueryRunner<ResultRow> runner : QueryRunnerTestHelper.makeQueryRunners(factory)) {
        final QueryRunner modifiedRunner = new QueryRunner() {

            @Override
            public Sequence run(QueryPlus queryPlus, ResponseContext responseContext) {
                TimeseriesQuery tsQuery = (TimeseriesQuery) queryPlus.getQuery();
                QueryRunner<ResultRow> newRunner = factory.mergeRunners(Execs.directExecutor(), ImmutableList.of(runner));
                QueryToolChest toolChest = factory.getToolchest();
                newRunner = new FinalizeResultsQueryRunner<>(toolChest.mergeResults(toolChest.preMergeQueryDecoration(newRunner)), toolChest);
                final String timeDimension = tsQuery.getTimestampResultField();
                final List<VirtualColumn> virtualColumns = new ArrayList<>(Arrays.asList(tsQuery.getVirtualColumns().getVirtualColumns()));
                Map<String, Object> theContext = tsQuery.getContext();
                if (timeDimension != null) {
                    theContext = new HashMap<>(tsQuery.getContext());
                    final PeriodGranularity granularity = (PeriodGranularity) tsQuery.getGranularity();
                    virtualColumns.add(new ExpressionVirtualColumn("v0", StringUtils.format("timestamp_floor(__time, '%s')", granularity.getPeriod()), ColumnType.LONG, TestExprMacroTable.INSTANCE));
                    theContext.put(GroupByQuery.CTX_TIMESTAMP_RESULT_FIELD, timeDimension);
                    theContext.put(GroupByQuery.CTX_TIMESTAMP_RESULT_FIELD_GRANULARITY, granularity);
                    theContext.put(GroupByQuery.CTX_TIMESTAMP_RESULT_FIELD_INDEX, 0);
                }
                GroupByQuery newQuery = GroupByQuery.builder().setDataSource(tsQuery.getDataSource()).setQuerySegmentSpec(tsQuery.getQuerySegmentSpec()).setGranularity(tsQuery.getGranularity()).setDimFilter(tsQuery.getDimensionsFilter()).setDimensions(timeDimension == null ? ImmutableList.of() : ImmutableList.of(new DefaultDimensionSpec("v0", timeDimension, ColumnType.LONG))).setAggregatorSpecs(tsQuery.getAggregatorSpecs()).setPostAggregatorSpecs(tsQuery.getPostAggregatorSpecs()).setVirtualColumns(VirtualColumns.create(virtualColumns)).setContext(theContext).build();
                return Sequences.map(newRunner.run(queryPlus.withQuery(newQuery), responseContext), new Function<ResultRow, Result<TimeseriesResultValue>>() {

                    @Override
                    public Result<TimeseriesResultValue> apply(final ResultRow input) {
                        final MapBasedRow mapBasedRow = input.toMapBasedRow(newQuery);
                        return new Result<>(mapBasedRow.getTimestamp(), new TimeseriesResultValue(mapBasedRow.getEvent()));
                    }
                });
            }

            @Override
            public String toString() {
                return runner.toString();
            }
        };
        for (boolean vectorize : ImmutableList.of(false, true)) {
            // Add vectorization tests for any indexes that support it.
            if (!vectorize || QueryRunnerTestHelper.isTestRunnerVectorizable(runner)) {
                constructors.add(new Object[] { modifiedRunner, vectorize });
            }
        }
    }
    return constructors;
}