use of io.trino.spi.predicate.Domain in project trino by trinodb.
the class KafkaFilterManager method filterRangeByDomain.
@VisibleForTesting
public static Optional<Range> filterRangeByDomain(Domain domain) {
Long low = INVALID_KAFKA_RANGE_INDEX;
Long high = INVALID_KAFKA_RANGE_INDEX;
if (domain.isSingleValue()) {
// still return range for single value case like (_partition_offset=XXX or _timestamp=XXX)
low = (long) domain.getSingleValue();
high = (long) domain.getSingleValue();
} else {
ValueSet valueSet = domain.getValues();
if (valueSet instanceof SortedRangeSet) {
// still return range for single value case like (_partition_offset in (XXX1,XXX2) or _timestamp in XXX1, XXX2)
Ranges ranges = ((SortedRangeSet) valueSet).getRanges();
List<io.trino.spi.predicate.Range> rangeList = ranges.getOrderedRanges();
if (rangeList.stream().allMatch(io.trino.spi.predicate.Range::isSingleValue)) {
List<Long> values = rangeList.stream().map(range -> (Long) range.getSingleValue()).collect(toImmutableList());
low = Collections.min(values);
high = Collections.max(values);
} else {
io.trino.spi.predicate.Range span = ranges.getSpan();
low = getLowIncludedValue(span).orElse(low);
high = getHighIncludedValue(span).orElse(high);
}
}
}
if (high != INVALID_KAFKA_RANGE_INDEX) {
high = high + 1;
}
return Optional.of(new Range(low, high));
}
use of io.trino.spi.predicate.Domain in project trino by trinodb.
the class TestPrometheusSplit method testPredicatePushDownSetsUpperBoundOnly.
@Test
public void testPredicatePushDownSetsUpperBoundOnly() {
long predicateHighValue = 1568638171999L;
Range highRange = Range.lessThanOrEqual(TIMESTAMP_COLUMN_TYPE, packDateTimeWithZone(predicateHighValue, UTC_KEY));
ValueSet valueSet = ValueSet.ofRanges(highRange);
Domain testDomain = Domain.create(valueSet, false);
TupleDomain<ColumnHandle> testTupleDomain = TupleDomain.withColumnDomains(ImmutableMap.of(new PrometheusColumnHandle("timestamp", TIMESTAMP_COLUMN_TYPE, 2), testDomain));
PrometheusTableHandle prometheusTableHandle = new PrometheusTableHandle("schemaName", "tableName").withPredicate(testTupleDomain);
io.airlift.units.Duration maxQueryRangeDuration = new io.airlift.units.Duration(120, TimeUnit.SECONDS);
io.airlift.units.Duration queryChunkSizeDuration = new io.airlift.units.Duration(30, TimeUnit.SECONDS);
Instant now = ofEpochMilli(1568638171999L + 600000L);
List<String> splitTimes = PrometheusSplitManager.generateTimesForSplits(now, maxQueryRangeDuration, queryChunkSizeDuration, prometheusTableHandle);
TemporalAmount expectedMaxQueryAsTime = java.time.Duration.ofMillis(maxQueryRangeDuration.toMillis() + ((splitTimes.size() - 1) * OFFSET_MILLIS));
String lastSplit = splitTimes.get(splitTimes.size() - 1);
Instant lastSplitAsTime = ofEpochMilli(longFromDecimalSecondString(lastSplit));
String earliestSplit = splitTimes.get(0);
Instant earliestSplitAsTime = ofEpochMilli(longFromDecimalSecondString(earliestSplit));
TemporalAmount queryChunkAsTime = java.time.Duration.ofMillis(queryChunkSizeDuration.toMillis());
java.time.Duration actualMaxDuration = Duration.between(earliestSplitAsTime.minus(queryChunkAsTime), lastSplitAsTime);
assertEquals(lastSplitAsTime.toEpochMilli(), 1568638171999L);
assertEquals(actualMaxDuration, expectedMaxQueryAsTime);
}
use of io.trino.spi.predicate.Domain in project trino by trinodb.
the class ShardMetadataRecordCursor method getTableIds.
@VisibleForTesting
static Iterator<Long> getTableIds(Jdbi dbi, TupleDomain<Integer> tupleDomain) {
Map<Integer, Domain> domains = tupleDomain.getDomains().get();
Domain schemaNameDomain = domains.get(getColumnIndex(SHARD_METADATA, SCHEMA_NAME));
Domain tableNameDomain = domains.get(getColumnIndex(SHARD_METADATA, TABLE_NAME));
List<String> values = new ArrayList<>();
StringBuilder sql = new StringBuilder("SELECT table_id FROM tables ");
if (schemaNameDomain != null || tableNameDomain != null) {
sql.append("WHERE ");
List<String> predicates = new ArrayList<>();
if (tableNameDomain != null && tableNameDomain.isSingleValue()) {
predicates.add("table_name = ?");
values.add(getStringValue(tableNameDomain.getSingleValue()));
}
if (schemaNameDomain != null && schemaNameDomain.isSingleValue()) {
predicates.add("schema_name = ?");
values.add(getStringValue(schemaNameDomain.getSingleValue()));
}
sql.append(Joiner.on(" AND ").join(predicates));
}
ImmutableList.Builder<Long> tableIds = ImmutableList.builder();
try (Connection connection = dbi.open().getConnection();
PreparedStatement statement = connection.prepareStatement(sql.toString())) {
for (int i = 0; i < values.size(); i++) {
statement.setString(i + 1, values.get(i));
}
try (ResultSet resultSet = statement.executeQuery()) {
while (resultSet.next()) {
tableIds.add(resultSet.getLong("table_id"));
}
}
} catch (SQLException | JdbiException e) {
throw metadataError(e);
}
return tableIds.build().iterator();
}
use of io.trino.spi.predicate.Domain in project trino by trinodb.
the class BaseSqlServerConnectorTest method testPredicatePushdown.
@Test
public void testPredicatePushdown() {
// varchar equality
assertThat(query("SELECT regionkey, nationkey, name FROM nation WHERE name = 'ROMANIA'")).matches("VALUES (BIGINT '3', BIGINT '19', CAST('ROMANIA' AS varchar(25)))").isNotFullyPushedDown(FilterNode.class);
// varchar range
assertThat(query("SELECT regionkey, nationkey, name FROM nation WHERE name BETWEEN 'POLAND' AND 'RPA'")).matches("VALUES (BIGINT '3', BIGINT '19', CAST('ROMANIA' AS varchar(25)))").isNotFullyPushedDown(FilterNode.class);
// varchar IN without domain compaction
assertThat(query("SELECT regionkey, nationkey, name FROM nation WHERE name IN ('POLAND', 'ROMANIA', 'VIETNAM')")).matches("VALUES " + "(BIGINT '3', BIGINT '19', CAST('ROMANIA' AS varchar(25))), " + "(BIGINT '2', BIGINT '21', CAST('VIETNAM' AS varchar(25)))").isNotFullyPushedDown(node(FilterNode.class, // verify that pushed down constraint is applied by the connector
tableScan(tableHandle -> {
TupleDomain<ColumnHandle> constraint = ((JdbcTableHandle) tableHandle).getConstraint();
ColumnHandle nameColumn = constraint.getDomains().orElseThrow().keySet().stream().map(JdbcColumnHandle.class::cast).filter(column -> column.getColumnName().equals("name")).collect(onlyElement());
return constraint.getDomains().get().get(nameColumn).equals(Domain.multipleValues(createVarcharType(25), ImmutableList.of(utf8Slice("POLAND"), utf8Slice("ROMANIA"), utf8Slice("VIETNAM"))));
}, TupleDomain.all(), ImmutableMap.of())));
// varchar IN with small compaction threshold
assertThat(query(Session.builder(getSession()).setCatalogSessionProperty("sqlserver", "domain_compaction_threshold", "1").build(), "SELECT regionkey, nationkey, name FROM nation WHERE name IN ('POLAND', 'ROMANIA', 'VIETNAM')")).matches("VALUES " + "(BIGINT '3', BIGINT '19', CAST('ROMANIA' AS varchar(25))), " + "(BIGINT '2', BIGINT '21', CAST('VIETNAM' AS varchar(25)))").isNotFullyPushedDown(node(FilterNode.class, // verify that no constraint is applied by the connector
tableScan(tableHandle -> ((JdbcTableHandle) tableHandle).getConstraint().isAll(), TupleDomain.all(), ImmutableMap.of())));
// varchar different case
assertThat(query("SELECT regionkey, nationkey, name FROM nation WHERE name = 'romania'")).returnsEmptyResult().isNotFullyPushedDown(FilterNode.class);
// bigint equality
assertThat(query("SELECT regionkey, nationkey, name FROM nation WHERE nationkey = 19")).matches("VALUES (BIGINT '3', BIGINT '19', CAST('ROMANIA' AS varchar(25)))").isFullyPushedDown();
// bigint equality with small compaction threshold
assertThat(query(Session.builder(getSession()).setCatalogSessionProperty("sqlserver", "domain_compaction_threshold", "1").build(), "SELECT regionkey, nationkey, name FROM nation WHERE nationkey IN (19, 21)")).matches("VALUES " + "(BIGINT '3', BIGINT '19', CAST('ROMANIA' AS varchar(25))), " + "(BIGINT '2', BIGINT '21', CAST('VIETNAM' AS varchar(25)))").isNotFullyPushedDown(FilterNode.class);
// bigint range, with decimal to bigint simplification
assertThat(query("SELECT regionkey, nationkey, name FROM nation WHERE nationkey BETWEEN 18.5 AND 19.5")).matches("VALUES (BIGINT '3', BIGINT '19', CAST('ROMANIA' AS varchar(25)))").isFullyPushedDown();
// date equality
assertThat(query("SELECT orderkey FROM orders WHERE orderdate = DATE '1992-09-29'")).matches("VALUES BIGINT '1250', 34406, 38436, 57570").isFullyPushedDown();
// predicate over aggregation key (likely to be optimized before being pushed down into the connector)
assertThat(query("SELECT * FROM (SELECT regionkey, sum(nationkey) FROM nation GROUP BY regionkey) WHERE regionkey = 3")).matches("VALUES (BIGINT '3', BIGINT '77')").isFullyPushedDown();
// predicate over aggregation result
assertThat(query("SELECT regionkey, sum(nationkey) FROM nation GROUP BY regionkey HAVING sum(nationkey) = 77")).matches("VALUES (BIGINT '3', BIGINT '77')").isFullyPushedDown();
// decimals
try (TestTable testTable = new TestTable(onRemoteDatabase(), "test_decimal_pushdown", "(short_decimal decimal(9, 3), long_decimal decimal(30, 10))", List.of("123.321, 123456789.987654321"))) {
assertThat(query("SELECT * FROM " + testTable.getName() + " WHERE short_decimal <= 124")).matches("VALUES (CAST(123.321 AS decimal(9,3)), CAST(123456789.987654321 AS decimal(30, 10)))").isFullyPushedDown();
assertThat(query("SELECT * FROM " + testTable.getName() + " WHERE short_decimal <= 124")).matches("VALUES (CAST(123.321 AS decimal(9,3)), CAST(123456789.987654321 AS decimal(30, 10)))").isFullyPushedDown();
assertThat(query("SELECT * FROM " + testTable.getName() + " WHERE long_decimal <= 123456790")).matches("VALUES (CAST(123.321 AS decimal(9,3)), CAST(123456789.987654321 AS decimal(30, 10)))").isFullyPushedDown();
assertThat(query("SELECT * FROM " + testTable.getName() + " WHERE short_decimal <= 123.321")).matches("VALUES (CAST(123.321 AS decimal(9,3)), CAST(123456789.987654321 AS decimal(30, 10)))").isFullyPushedDown();
assertThat(query("SELECT * FROM " + testTable.getName() + " WHERE long_decimal <= 123456789.987654321")).matches("VALUES (CAST(123.321 AS decimal(9,3)), CAST(123456789.987654321 AS decimal(30, 10)))").isFullyPushedDown();
assertThat(query("SELECT * FROM " + testTable.getName() + " WHERE short_decimal = 123.321")).matches("VALUES (CAST(123.321 AS decimal(9,3)), CAST(123456789.987654321 AS decimal(30, 10)))").isFullyPushedDown();
assertThat(query("SELECT * FROM " + testTable.getName() + " WHERE long_decimal = 123456789.987654321")).matches("VALUES (CAST(123.321 AS decimal(9,3)), CAST(123456789.987654321 AS decimal(30, 10)))").isFullyPushedDown();
}
}
use of io.trino.spi.predicate.Domain in project trino by trinodb.
the class IcebergPageSourceProvider method getParquetTupleDomain.
private static TupleDomain<ColumnDescriptor> getParquetTupleDomain(Map<List<String>, RichColumnDescriptor> descriptorsByPath, TupleDomain<IcebergColumnHandle> effectivePredicate) {
if (effectivePredicate.isNone()) {
return TupleDomain.none();
}
ImmutableMap.Builder<ColumnDescriptor, Domain> predicate = ImmutableMap.builder();
effectivePredicate.getDomains().get().forEach((columnHandle, domain) -> {
String baseType = columnHandle.getType().getTypeSignature().getBase();
// skip looking up predicates for complex types as Parquet only stores stats for primitives
if (!baseType.equals(StandardTypes.MAP) && !baseType.equals(StandardTypes.ARRAY) && !baseType.equals(StandardTypes.ROW)) {
RichColumnDescriptor descriptor = descriptorsByPath.get(ImmutableList.of(columnHandle.getName()));
if (descriptor != null) {
predicate.put(descriptor, domain);
}
}
});
return TupleDomain.withColumnDomains(predicate.buildOrThrow());
}
Aggregations