Examples with QueryDataSource org.apache.druid.query.QueryDataSource used on opensource projects

Example 1 with QueryDataSource

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

the class CalciteQueryTest method testRequireTimeConditionPositive.

@Test
public void testRequireTimeConditionPositive() throws Exception {
    // simple timeseries
    testQuery(PLANNER_CONFIG_REQUIRE_TIME_CONDITION, "SELECT SUM(cnt), gran FROM (\n" + "  SELECT __time as t, floor(__time TO month) AS gran,\n" + "  cnt FROM druid.foo\n" + ") AS x\n" + "WHERE t >= '2000-01-01' and t < '2002-01-01'" + "GROUP BY gran\n" + "ORDER BY gran", CalciteTests.REGULAR_USER_AUTH_RESULT, ImmutableList.of(Druids.newTimeseriesQueryBuilder().dataSource(CalciteTests.DATASOURCE1).intervals(querySegmentSpec(Intervals.of("2000-01-01/2002-01-01"))).granularity(Granularities.MONTH).aggregators(aggregators(new LongSumAggregatorFactory("a0", "cnt"))).context(getTimeseriesContextWithFloorTime(TIMESERIES_CONTEXT_BY_GRAN, "d0")).build()), ImmutableList.of(new Object[] { 3L, timestamp("2000-01-01") }, new Object[] { 3L, timestamp("2001-01-01") }));
    // nested groupby only requires time condition for inner most query
    testQuery(PLANNER_CONFIG_REQUIRE_TIME_CONDITION, "SELECT\n" + "  SUM(cnt),\n" + "  COUNT(*)\n" + "FROM (SELECT dim2, SUM(cnt) AS cnt FROM druid.foo WHERE __time >= '2000-01-01' GROUP BY dim2)", CalciteTests.REGULAR_USER_AUTH_RESULT, ImmutableList.of(GroupByQuery.builder().setDataSource(new QueryDataSource(GroupByQuery.builder().setDataSource(CalciteTests.DATASOURCE1).setInterval(querySegmentSpec(Intervals.utc(DateTimes.of("2000-01-01").getMillis(), JodaUtils.MAX_INSTANT))).setGranularity(Granularities.ALL).setDimensions(dimensions(new DefaultDimensionSpec("dim2", "d0"))).setAggregatorSpecs(aggregators(new LongSumAggregatorFactory("a0", "cnt"))).setContext(QUERY_CONTEXT_DEFAULT).build())).setInterval(querySegmentSpec(Filtration.eternity())).setGranularity(Granularities.ALL).setAggregatorSpecs(aggregators(new LongSumAggregatorFactory("_a0", "a0"), new CountAggregatorFactory("_a1"))).setContext(QUERY_CONTEXT_DEFAULT).build()), NullHandling.replaceWithDefault() ? ImmutableList.of(new Object[] { 6L, 3L }) : ImmutableList.of(new Object[] { 6L, 4L }));
    // Cannot vectorize next test due to extraction dimension spec.
    cannotVectorize();
    // semi-join requires time condition on both left and right query
    testQuery(PLANNER_CONFIG_REQUIRE_TIME_CONDITION, "SELECT COUNT(*) FROM druid.foo\n" + "WHERE __time >= '2000-01-01' AND SUBSTRING(dim2, 1, 1) IN (\n" + "  SELECT SUBSTRING(dim1, 1, 1) FROM druid.foo\n" + "  WHERE dim1 <> '' AND __time >= '2000-01-01'\n" + ")", CalciteTests.REGULAR_USER_AUTH_RESULT, ImmutableList.of(Druids.newTimeseriesQueryBuilder().dataSource(join(new TableDataSource(CalciteTests.DATASOURCE1), new QueryDataSource(GroupByQuery.builder().setDataSource(CalciteTests.DATASOURCE1).setInterval(querySegmentSpec(Intervals.utc(DateTimes.of("2000-01-01").getMillis(), JodaUtils.MAX_INSTANT))).setDimFilter(not(selector("dim1", NullHandling.sqlCompatible() ? "" : null, null))).setGranularity(Granularities.ALL).setDimensions(new ExtractionDimensionSpec("dim1", "d0", ColumnType.STRING, new SubstringDimExtractionFn(0, 1))).setContext(QUERY_CONTEXT_DEFAULT).build()), "j0.", equalsCondition(makeExpression("substring(\"dim2\", 0, 1)"), DruidExpression.ofColumn(ColumnType.STRING, "j0.d0")), JoinType.INNER)).intervals(querySegmentSpec(Intervals.utc(DateTimes.of("2000-01-01").getMillis(), JodaUtils.MAX_INSTANT))).granularity(Granularities.ALL).aggregators(aggregators(new CountAggregatorFactory("a0"))).context(QUERY_CONTEXT_DEFAULT).build()), ImmutableList.of(new Object[] { 3L }));
}

Example 2 with QueryDataSource

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

the class CalciteQueryTest method testQueryWithSelectProjectAndIdentityProjectDoesNotRename.

@Test
public void testQueryWithSelectProjectAndIdentityProjectDoesNotRename() throws Exception {
    cannotVectorize();
    requireMergeBuffers(3);
    testQuery(PLANNER_CONFIG_NO_HLL.withOverrides(ImmutableMap.of(PlannerConfig.CTX_KEY_USE_GROUPING_SET_FOR_EXACT_DISTINCT, "true")), "SELECT\n" + "(SUM(CASE WHEN (TIMESTAMP '2000-01-04 17:00:00'<=__time AND __time<TIMESTAMP '2022-01-05 17:00:00') THEN 1 ELSE 0 END)*1.0/COUNT(DISTINCT CASE WHEN (TIMESTAMP '2000-01-04 17:00:00'<=__time AND __time<TIMESTAMP '2022-01-05 17:00:00') THEN dim1 END))\n" + "FROM druid.foo\n" + "GROUP BY ()", CalciteTests.REGULAR_USER_AUTH_RESULT, ImmutableList.of(GroupByQuery.builder().setDataSource(new QueryDataSource(GroupByQuery.builder().setDataSource(CalciteTests.DATASOURCE1).setInterval(querySegmentSpec(Filtration.eternity())).setGranularity(Granularities.ALL).setVirtualColumns(expressionVirtualColumn("v0", "case_searched(((947005200000 <= \"__time\") && (\"__time\" < 1641402000000)),\"dim1\",null)", ColumnType.STRING)).setDimensions(dimensions(new DefaultDimensionSpec("v0", "d0", ColumnType.STRING))).setAggregatorSpecs(aggregators(new LongSumAggregatorFactory("a0", null, "case_searched(((947005200000 <= \"__time\") && (\"__time\" < 1641402000000)),1,0)", ExprMacroTable.nil()), new GroupingAggregatorFactory("a1", ImmutableList.of("v0")))).setSubtotalsSpec(ImmutableList.of(ImmutableList.of("d0"), ImmutableList.of())).setContext(QUERY_CONTEXT_DEFAULT).build())).setInterval(querySegmentSpec(Filtration.eternity())).setGranularity(Granularities.ALL).setAggregatorSpecs(aggregators(new FilteredAggregatorFactory(new LongMinAggregatorFactory("_a0", "a0"), selector("a1", "1", null)), new FilteredAggregatorFactory(new CountAggregatorFactory("_a1"), and(not(selector("d0", null, null)), selector("a1", "0", null))))).setPostAggregatorSpecs(Collections.singletonList(new ExpressionPostAggregator("p0", "((\"_a0\" * 1.0) / \"_a1\")", null, ExprMacroTable.nil()))).setContext(QUERY_CONTEXT_DEFAULT).build()), ImmutableList.of(new Object[] { 1.0d }));
}

Example 3 with QueryDataSource

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

the class CalciteQueryTest method testSelfJoin.

@Test
public void testSelfJoin() throws Exception {
    // Cannot vectorize due to virtual columns.
    cannotVectorize();
    testQuery("SELECT COUNT(*) FROM druid.foo x, druid.foo y\n", ImmutableList.of(Druids.newTimeseriesQueryBuilder().dataSource(join(new TableDataSource(CalciteTests.DATASOURCE1), new QueryDataSource(newScanQueryBuilder().dataSource(CalciteTests.DATASOURCE1).intervals(querySegmentSpec(Filtration.eternity())).columns(ImmutableList.of("__time", "cnt", "dim1", "dim2", "dim3", "m1", "m2", "unique_dim1")).resultFormat(ScanQuery.ResultFormat.RESULT_FORMAT_COMPACTED_LIST).context(QUERY_CONTEXT_DEFAULT).build()), "j0.", "1", JoinType.INNER)).intervals(querySegmentSpec(Filtration.eternity())).granularity(Granularities.ALL).aggregators(aggregators(new CountAggregatorFactory("a0"))).context(QUERY_CONTEXT_DEFAULT).build()), ImmutableList.of(new Object[] { 36L }));
}

Example 4 with QueryDataSource

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

the class CalciteQueryTest method testViewAndJoin.

@Test
public void testViewAndJoin() throws Exception {
    cannotVectorize();
    Map<String, Object> queryContext = withLeftDirectAccessEnabled(QUERY_CONTEXT_DEFAULT);
    testQuery("SELECT COUNT(*) FROM view.cview as a INNER JOIN druid.foo d on d.dim2 = a.dim2 WHERE a.dim1_firstchar <> 'z' ", queryContext, ImmutableList.of(Druids.newTimeseriesQueryBuilder().dataSource(join(join(new TableDataSource(CalciteTests.DATASOURCE1), new QueryDataSource(newScanQueryBuilder().dataSource(CalciteTests.DATASOURCE3).intervals(querySegmentSpec(Filtration.eternity())).columns("dim2").context(queryContext).build()), "j0.", "(\"dim2\" == \"j0.dim2\")", JoinType.INNER, bound("dim2", "a", "a", false, false, null, null)), new QueryDataSource(newScanQueryBuilder().dataSource(CalciteTests.DATASOURCE1).intervals(querySegmentSpec(Filtration.eternity())).columns("dim2").context(queryContext).build()), "_j0.", "('a' == \"_j0.dim2\")", JoinType.INNER)).intervals(querySegmentSpec(Filtration.eternity())).filters(not(selector("dim1", "z", new SubstringDimExtractionFn(0, 1)))).granularity(Granularities.ALL).aggregators(aggregators(new CountAggregatorFactory("a0"))).context(withLeftDirectAccessEnabled(QUERY_CONTEXT_DEFAULT)).build()), ImmutableList.of(new Object[] { 8L }));
}

Example 5 with QueryDataSource

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

the class CalciteQueryTest method testExactCountDistinctUsingSubqueryWithWherePushDown.

@Test
public void testExactCountDistinctUsingSubqueryWithWherePushDown() throws Exception {
    testQuery("SELECT\n" + "  SUM(cnt),\n" + "  COUNT(*)\n" + "FROM (SELECT dim2, SUM(cnt) AS cnt FROM druid.foo GROUP BY dim2)\n" + "WHERE dim2 <> ''", ImmutableList.of(GroupByQuery.builder().setDataSource(new QueryDataSource(GroupByQuery.builder().setDataSource(CalciteTests.DATASOURCE1).setInterval(querySegmentSpec(Filtration.eternity())).setDimFilter(not(selector("dim2", "", null))).setGranularity(Granularities.ALL).setDimensions(dimensions(new DefaultDimensionSpec("dim2", "d0"))).setAggregatorSpecs(aggregators(new LongSumAggregatorFactory("a0", "cnt"))).setContext(QUERY_CONTEXT_DEFAULT).build())).setInterval(querySegmentSpec(Filtration.eternity())).setGranularity(Granularities.ALL).setAggregatorSpecs(aggregators(new LongSumAggregatorFactory("_a0", "a0"), new CountAggregatorFactory("_a1"))).setContext(QUERY_CONTEXT_DEFAULT).build()), NullHandling.replaceWithDefault() ? ImmutableList.of(new Object[] { 3L, 2L }) : ImmutableList.of(new Object[] { 5L, 3L }));
    testQuery("SELECT\n" + "  SUM(cnt),\n" + "  COUNT(*)\n" + "FROM (SELECT dim2, SUM(cnt) AS cnt FROM druid.foo GROUP BY dim2)\n" + "WHERE dim2 IS NOT NULL", ImmutableList.of(GroupByQuery.builder().setDataSource(new QueryDataSource(GroupByQuery.builder().setDataSource(CalciteTests.DATASOURCE1).setInterval(querySegmentSpec(Filtration.eternity())).setDimFilter(not(selector("dim2", null, null))).setGranularity(Granularities.ALL).setDimensions(dimensions(new DefaultDimensionSpec("dim2", "d0"))).setAggregatorSpecs(aggregators(new LongSumAggregatorFactory("a0", "cnt"))).setContext(QUERY_CONTEXT_DEFAULT).build())).setInterval(querySegmentSpec(Filtration.eternity())).setGranularity(Granularities.ALL).setAggregatorSpecs(aggregators(new LongSumAggregatorFactory("_a0", "a0"), new CountAggregatorFactory("_a1"))).setContext(QUERY_CONTEXT_DEFAULT).build()), NullHandling.replaceWithDefault() ? ImmutableList.of(new Object[] { 3L, 2L }) : ImmutableList.of(new Object[] { 4L, 3L }));
}