use of io.trino.spi.type.TimestampType in project trino by trinodb.
the class TupleDomainParquetPredicate method getDomain.
/**
* Get a domain for the ranges defined by each pair of elements from {@code minimums} and {@code maximums}.
* Both arrays must have the same length.
*/
private static Domain getDomain(ColumnDescriptor column, Type type, List<Object> minimums, List<Object> maximums, boolean hasNullValue, DateTimeZone timeZone) {
checkArgument(minimums.size() == maximums.size(), "Expected minimums and maximums to have the same size");
if (type.equals(BOOLEAN)) {
boolean hasTrueValues = minimums.stream().anyMatch(value -> (boolean) value) || maximums.stream().anyMatch(value -> (boolean) value);
boolean hasFalseValues = minimums.stream().anyMatch(value -> !(boolean) value) || maximums.stream().anyMatch(value -> !(boolean) value);
if (hasTrueValues && hasFalseValues) {
return Domain.all(type);
}
if (hasTrueValues) {
return Domain.create(ValueSet.of(type, true), hasNullValue);
}
if (hasFalseValues) {
return Domain.create(ValueSet.of(type, false), hasNullValue);
}
// All nulls case is handled earlier
throw new VerifyException("Impossible boolean statistics");
}
if (type.equals(BIGINT) || type.equals(INTEGER) || type.equals(DATE) || type.equals(SMALLINT) || type.equals(TINYINT)) {
List<Range> ranges = new ArrayList<>();
for (int i = 0; i < minimums.size(); i++) {
long min = asLong(minimums.get(i));
long max = asLong(maximums.get(i));
if (isStatisticsOverflow(type, min, max)) {
return Domain.create(ValueSet.all(type), hasNullValue);
}
ranges.add(Range.range(type, min, true, max, true));
}
return Domain.create(ValueSet.ofRanges(ranges), hasNullValue);
}
if (type instanceof DecimalType) {
DecimalType decimalType = (DecimalType) type;
List<Range> ranges = new ArrayList<>();
if (decimalType.isShort()) {
for (int i = 0; i < minimums.size(); i++) {
Object min = minimums.get(i);
Object max = maximums.get(i);
long minValue = min instanceof Binary ? getShortDecimalValue(((Binary) min).getBytes()) : asLong(min);
long maxValue = min instanceof Binary ? getShortDecimalValue(((Binary) max).getBytes()) : asLong(max);
if (isStatisticsOverflow(type, minValue, maxValue)) {
return Domain.create(ValueSet.all(type), hasNullValue);
}
ranges.add(Range.range(type, minValue, true, maxValue, true));
}
} else {
for (int i = 0; i < minimums.size(); i++) {
Int128 min = Int128.fromBigEndian(((Binary) minimums.get(i)).getBytes());
Int128 max = Int128.fromBigEndian(((Binary) maximums.get(i)).getBytes());
ranges.add(Range.range(type, min, true, max, true));
}
}
return Domain.create(ValueSet.ofRanges(ranges), hasNullValue);
}
if (type.equals(REAL)) {
List<Range> ranges = new ArrayList<>();
for (int i = 0; i < minimums.size(); i++) {
Float min = (Float) minimums.get(i);
Float max = (Float) maximums.get(i);
if (min.isNaN() || max.isNaN()) {
return Domain.create(ValueSet.all(type), hasNullValue);
}
ranges.add(Range.range(type, (long) floatToRawIntBits(min), true, (long) floatToRawIntBits(max), true));
}
return Domain.create(ValueSet.ofRanges(ranges), hasNullValue);
}
if (type.equals(DOUBLE)) {
List<Range> ranges = new ArrayList<>();
for (int i = 0; i < minimums.size(); i++) {
Double min = (Double) minimums.get(i);
Double max = (Double) maximums.get(i);
if (min.isNaN() || max.isNaN()) {
return Domain.create(ValueSet.all(type), hasNullValue);
}
ranges.add(Range.range(type, min, true, max, true));
}
return Domain.create(ValueSet.ofRanges(ranges), hasNullValue);
}
if (type instanceof VarcharType) {
List<Range> ranges = new ArrayList<>();
for (int i = 0; i < minimums.size(); i++) {
Slice min = Slices.wrappedBuffer(((Binary) minimums.get(i)).toByteBuffer());
Slice max = Slices.wrappedBuffer(((Binary) maximums.get(i)).toByteBuffer());
ranges.add(Range.range(type, min, true, max, true));
}
return Domain.create(ValueSet.ofRanges(ranges), hasNullValue);
}
if (type instanceof TimestampType) {
if (column.getPrimitiveType().getPrimitiveTypeName().equals(INT96)) {
TrinoTimestampEncoder<?> timestampEncoder = createTimestampEncoder((TimestampType) type, timeZone);
List<Object> values = new ArrayList<>();
for (int i = 0; i < minimums.size(); i++) {
Object min = minimums.get(i);
Object max = maximums.get(i);
// available and valid in that special case
if (!(min instanceof Binary) || !(max instanceof Binary) || !min.equals(max)) {
return Domain.create(ValueSet.all(type), hasNullValue);
}
values.add(timestampEncoder.getTimestamp(decodeInt96Timestamp((Binary) min)));
}
return Domain.multipleValues(type, values, hasNullValue);
}
if (column.getPrimitiveType().getPrimitiveTypeName().equals(INT64)) {
LogicalTypeAnnotation logicalTypeAnnotation = column.getPrimitiveType().getLogicalTypeAnnotation();
if (!(logicalTypeAnnotation instanceof TimestampLogicalTypeAnnotation)) {
// Invalid statistics. Unit and UTC adjustment are not known
return Domain.create(ValueSet.all(type), hasNullValue);
}
TimestampLogicalTypeAnnotation timestampTypeAnnotation = (TimestampLogicalTypeAnnotation) logicalTypeAnnotation;
// Bail out if the precision is not known
if (timestampTypeAnnotation.getUnit() == null) {
return Domain.create(ValueSet.all(type), hasNullValue);
}
TrinoTimestampEncoder<?> timestampEncoder = createTimestampEncoder((TimestampType) type, DateTimeZone.UTC);
List<Range> ranges = new ArrayList<>();
for (int i = 0; i < minimums.size(); i++) {
long min = (long) minimums.get(i);
long max = (long) maximums.get(i);
ranges.add(Range.range(type, timestampEncoder.getTimestamp(decodeInt64Timestamp(min, timestampTypeAnnotation.getUnit())), true, timestampEncoder.getTimestamp(decodeInt64Timestamp(max, timestampTypeAnnotation.getUnit())), true));
}
return Domain.create(ValueSet.ofRanges(ranges), hasNullValue);
}
}
return Domain.create(ValueSet.all(type), hasNullValue);
}
use of io.trino.spi.type.TimestampType in project trino by trinodb.
the class ParquetSchemaConverter method getPrimitiveType.
private org.apache.parquet.schema.Type getPrimitiveType(Type type, String name, List<String> parent) {
List<String> fullName = ImmutableList.<String>builder().addAll(parent).add(name).build();
primitiveTypes.put(fullName, type);
if (BOOLEAN.equals(type)) {
return Types.primitive(PrimitiveType.PrimitiveTypeName.BOOLEAN, OPTIONAL).named(name);
}
if (INTEGER.equals(type) || SMALLINT.equals(type) || TINYINT.equals(type)) {
return Types.primitive(PrimitiveType.PrimitiveTypeName.INT32, OPTIONAL).named(name);
}
if (type instanceof DecimalType) {
DecimalType decimalType = (DecimalType) type;
if (decimalType.getPrecision() <= 9) {
return Types.optional(PrimitiveType.PrimitiveTypeName.INT32).as(OriginalType.DECIMAL).precision(decimalType.getPrecision()).scale(decimalType.getScale()).named(name);
} else if (decimalType.isShort()) {
return Types.optional(PrimitiveType.PrimitiveTypeName.INT64).as(OriginalType.DECIMAL).precision(decimalType.getPrecision()).scale(decimalType.getScale()).named(name);
} else {
return Types.optional(PrimitiveType.PrimitiveTypeName.FIXED_LEN_BYTE_ARRAY).length(16).as(OriginalType.DECIMAL).precision(decimalType.getPrecision()).scale(decimalType.getScale()).named(name);
}
}
if (DATE.equals(type)) {
return Types.optional(PrimitiveType.PrimitiveTypeName.INT32).as(OriginalType.DATE).named(name);
}
if (BIGINT.equals(type)) {
return Types.primitive(PrimitiveType.PrimitiveTypeName.INT64, OPTIONAL).named(name);
}
if (type instanceof TimestampType) {
TimestampType timestampType = (TimestampType) type;
if (timestampType.getPrecision() <= 3) {
return Types.primitive(PrimitiveType.PrimitiveTypeName.INT64, OPTIONAL).as(LogicalTypeAnnotation.timestampType(false, LogicalTypeAnnotation.TimeUnit.MILLIS)).named(name);
}
if (timestampType.getPrecision() <= 6) {
return Types.primitive(PrimitiveType.PrimitiveTypeName.INT64, OPTIONAL).as(LogicalTypeAnnotation.timestampType(false, LogicalTypeAnnotation.TimeUnit.MICROS)).named(name);
}
if (timestampType.getPrecision() <= 9) {
// even though it can only hold values within 1677-09-21 00:12:43 and 2262-04-11 23:47:16 range.
return Types.primitive(PrimitiveType.PrimitiveTypeName.INT64, OPTIONAL).as(LogicalTypeAnnotation.timestampType(false, LogicalTypeAnnotation.TimeUnit.NANOS)).named(name);
}
}
if (DOUBLE.equals(type)) {
return Types.primitive(PrimitiveType.PrimitiveTypeName.DOUBLE, OPTIONAL).named(name);
}
if (RealType.REAL.equals(type)) {
return Types.primitive(PrimitiveType.PrimitiveTypeName.FLOAT, OPTIONAL).named(name);
}
if (type instanceof VarcharType || type instanceof CharType) {
return Types.primitive(PrimitiveType.PrimitiveTypeName.BINARY, OPTIONAL).as(LogicalTypeAnnotation.stringType()).named(name);
}
if (type instanceof VarbinaryType) {
return Types.primitive(PrimitiveType.PrimitiveTypeName.BINARY, OPTIONAL).named(name);
}
throw new TrinoException(NOT_SUPPORTED, format("Unsupported primitive type: %s", type));
}
use of io.trino.spi.type.TimestampType in project trino by trinodb.
the class MaterializedResult method writeValue.
private static void writeValue(Type type, BlockBuilder blockBuilder, Object value) {
if (value == null) {
blockBuilder.appendNull();
} else if (BIGINT.equals(type)) {
type.writeLong(blockBuilder, (Long) value);
} else if (INTEGER.equals(type)) {
type.writeLong(blockBuilder, (Integer) value);
} else if (SMALLINT.equals(type)) {
type.writeLong(blockBuilder, (Short) value);
} else if (TINYINT.equals(type)) {
type.writeLong(blockBuilder, (Byte) value);
} else if (REAL.equals(type)) {
type.writeLong(blockBuilder, floatToRawIntBits(((Float) value)));
} else if (DOUBLE.equals(type)) {
type.writeDouble(blockBuilder, (Double) value);
} else if (BOOLEAN.equals(type)) {
type.writeBoolean(blockBuilder, (Boolean) value);
} else if (JSON.equals(type)) {
type.writeSlice(blockBuilder, Slices.utf8Slice((String) value));
} else if (type instanceof VarcharType) {
type.writeSlice(blockBuilder, Slices.utf8Slice((String) value));
} else if (type instanceof CharType) {
type.writeSlice(blockBuilder, Slices.utf8Slice((String) value));
} else if (VARBINARY.equals(type)) {
type.writeSlice(blockBuilder, Slices.wrappedBuffer((byte[]) value));
} else if (DATE.equals(type)) {
int days = ((SqlDate) value).getDays();
type.writeLong(blockBuilder, days);
} else if (type instanceof TimeType) {
SqlTime time = (SqlTime) value;
type.writeLong(blockBuilder, time.getPicos());
} else if (type instanceof TimeWithTimeZoneType) {
long nanos = roundDiv(((SqlTimeWithTimeZone) value).getPicos(), PICOSECONDS_PER_NANOSECOND);
int offsetMinutes = ((SqlTimeWithTimeZone) value).getOffsetMinutes();
type.writeLong(blockBuilder, packTimeWithTimeZone(nanos, offsetMinutes));
} else if (type instanceof TimestampType) {
long micros = ((SqlTimestamp) value).getEpochMicros();
if (((TimestampType) type).getPrecision() <= TimestampType.MAX_SHORT_PRECISION) {
type.writeLong(blockBuilder, micros);
} else {
type.writeObject(blockBuilder, new LongTimestamp(micros, ((SqlTimestamp) value).getPicosOfMicros()));
}
} else if (TIMESTAMP_WITH_TIME_ZONE.equals(type)) {
long millisUtc = ((SqlTimestampWithTimeZone) value).getMillisUtc();
TimeZoneKey timeZoneKey = ((SqlTimestampWithTimeZone) value).getTimeZoneKey();
type.writeLong(blockBuilder, packDateTimeWithZone(millisUtc, timeZoneKey));
} else if (type instanceof ArrayType) {
List<?> list = (List<?>) value;
Type elementType = ((ArrayType) type).getElementType();
BlockBuilder arrayBlockBuilder = blockBuilder.beginBlockEntry();
for (Object element : list) {
writeValue(elementType, arrayBlockBuilder, element);
}
blockBuilder.closeEntry();
} else if (type instanceof MapType) {
Map<?, ?> map = (Map<?, ?>) value;
Type keyType = ((MapType) type).getKeyType();
Type valueType = ((MapType) type).getValueType();
BlockBuilder mapBlockBuilder = blockBuilder.beginBlockEntry();
for (Entry<?, ?> entry : map.entrySet()) {
writeValue(keyType, mapBlockBuilder, entry.getKey());
writeValue(valueType, mapBlockBuilder, entry.getValue());
}
blockBuilder.closeEntry();
} else if (type instanceof RowType) {
List<?> row = (List<?>) value;
List<Type> fieldTypes = type.getTypeParameters();
BlockBuilder rowBlockBuilder = blockBuilder.beginBlockEntry();
for (int field = 0; field < row.size(); field++) {
writeValue(fieldTypes.get(field), rowBlockBuilder, row.get(field));
}
blockBuilder.closeEntry();
} else {
throw new IllegalArgumentException("Unsupported type " + type);
}
}
use of io.trino.spi.type.TimestampType in project trino by trinodb.
the class TestTupleDomainParquetPredicate method testTimestampInt96.
@Test(dataProvider = "timestampPrecision")
public void testTimestampInt96(int precision, LocalDateTime baseTime, Object baseDomainValue) throws ParquetCorruptionException {
ColumnDescriptor columnDescriptor = createColumnDescriptor(INT96, "TimestampColumn");
TimestampType timestampType = createTimestampType(precision);
assertEquals(getDomain(columnDescriptor, timestampType, 0, null, ID, UTC), all(timestampType));
assertEquals(getDomain(columnDescriptor, timestampType, 10, timestampColumnStats(baseTime, baseTime), ID, UTC), singleValue(timestampType, baseDomainValue));
// INT96 binary ranges ignored when min <> max
assertEquals(getDomain(columnDescriptor, timestampType, 10, timestampColumnStats(baseTime.minusSeconds(10), baseTime), ID, UTC), create(ValueSet.all(timestampType), false));
}
use of io.trino.spi.type.TimestampType in project trino by trinodb.
the class GenericHiveRecordCursor method parseObjectColumn.
private void parseObjectColumn(int column) {
loaded[column] = true;
Object fieldData = rowInspector.getStructFieldData(rowData, structFields[column]);
if (fieldData == null) {
nulls[column] = true;
} else {
Type type = types[column];
if (type.getJavaType() == Block.class) {
objects[column] = getBlockObject(type, fieldData, fieldInspectors[column]);
} else if (type instanceof TimestampType) {
Timestamp timestamp = (Timestamp) ((PrimitiveObjectInspector) fieldInspectors[column]).getPrimitiveJavaObject(fieldData);
objects[column] = longTimestamp(timestamp, column);
} else {
throw new IllegalStateException("Unsupported type: " + type);
}
nulls[column] = false;
}
}
Aggregations