klion26 commented on code in PR #10157:
URL: https://github.com/apache/arrow-rs/pull/10157#discussion_r3656477810
##########
parquet-variant-compute/src/shred_variant.rs:
##########
@@ -2829,4 +3093,237 @@ mod tests {
let shredding_type = ShreddedSchemaBuilder::default().build();
assert_eq!(shredding_type, DataType::Null);
}
+
+ // This test wants to cover that the variant can/can't be shredded to the
given data type.
+ #[test]
+ fn test_variant_type_shredded_correctly() {
+ // array contains all variant types
+ let mut array_builder = VariantArrayBuilder::new(30);
+ array_builder.append_value(Variant::Null);
+ array_builder.append_value(Variant::Int8(1));
+ array_builder.append_value(Variant::Int16(2));
+ array_builder.append_value(Variant::Int32(3));
+ array_builder.append_value(Variant::Int64(4));
+
array_builder.append_value(Variant::Date(NaiveDate::from_epoch_days(12345).unwrap()));
+ array_builder.append_value(Variant::TimestampMicros(
+ DateTime::from_timestamp_micros(123456789).unwrap(),
+ ));
+ array_builder.append_value(Variant::TimestampNtzMicros(
+ DateTime::from_timestamp_micros(123456789)
+ .unwrap()
+ .naive_utc(),
+ ));
+
array_builder.append_value(Variant::TimestampNanos(DateTime::from_timestamp_nanos(
+ 1234567890000,
+ )));
+ array_builder.append_value(Variant::TimestampNtzNanos(
+ DateTime::from_timestamp_nanos(1234567890000).naive_utc(),
+ ));
+ array_builder.append_value(VariantDecimal4::try_new(123, 0).unwrap());
+ array_builder.append_value(VariantDecimal8::try_new(123, 0).unwrap());
+ array_builder.append_value(VariantDecimal16::try_new(123, 0).unwrap());
+ array_builder.append_value(Variant::Float(5.0));
+ array_builder.append_value(Variant::Double(6f64));
+ array_builder.append_value(Variant::BooleanTrue);
+ array_builder.append_value(Variant::BooleanFalse);
+ array_builder.append_value(Variant::Binary("helow".as_bytes()));
+ array_builder.append_value(Variant::String("hello"));
+ array_builder.append_value(Variant::ShortString(
+ ShortString::try_from("world").unwrap(),
+ ));
+ array_builder.append_value(Variant::Time(
+ NaiveTime::from_num_seconds_from_midnight_opt(12345, 123).unwrap(),
+ ));
+
+ let array = array_builder.build();
+
+ fn can_shred_to(v: &Variant, dt: &DataType) -> bool {
+ matches!(
+ (v, dt),
+ (Variant::Int8(_), DataType::Int8)
+ | (Variant::Int8(_), DataType::Int16)
+ | (Variant::Int8(_), DataType::Int32)
+ | (Variant::Int8(_), DataType::Int64)
+ | (Variant::Int8(_), DataType::Decimal32(_, _))
+ | (Variant::Int8(_), DataType::Decimal64(_, _))
+ | (Variant::Int8(_), DataType::Decimal128(_, _))
+ | (Variant::Int16(_), DataType::Int8)
+ | (Variant::Int16(_), DataType::Int16)
+ | (Variant::Int16(_), DataType::Int32)
+ | (Variant::Int16(_), DataType::Int64)
+ | (Variant::Int16(_), DataType::Decimal32(_, _))
+ | (Variant::Int16(_), DataType::Decimal64(_, _))
+ | (Variant::Int16(_), DataType::Decimal128(_, _))
+ | (Variant::Int32(_), DataType::Int8)
+ | (Variant::Int32(_), DataType::Int16)
+ | (Variant::Int32(_), DataType::Int32)
+ | (Variant::Int32(_), DataType::Int64)
+ | (Variant::Int32(_), DataType::Decimal32(_, _))
+ | (Variant::Int32(_), DataType::Decimal64(_, _))
+ | (Variant::Int32(_), DataType::Decimal128(_, _))
+ | (Variant::Int64(_), DataType::Int8)
+ | (Variant::Int64(_), DataType::Int16)
+ | (Variant::Int64(_), DataType::Int32)
+ | (Variant::Int64(_), DataType::Int64)
+ | (Variant::Int64(_), DataType::Decimal32(_, _))
+ | (Variant::Int64(_), DataType::Decimal64(_, _))
+ | (Variant::Int64(_), DataType::Decimal128(_, _))
+ | (Variant::Date(_), DataType::Date32)
+ | (
+ Variant::TimestampMicros(_),
+ DataType::Timestamp(TimeUnit::Microsecond, Some(_)),
+ )
+ | (
+ Variant::TimestampMicros(_),
+ DataType::Timestamp(TimeUnit::Nanosecond, Some(_))
+ )
+ | (
+ Variant::TimestampNtzMicros(_),
+ DataType::Timestamp(TimeUnit::Microsecond, None),
+ )
+ | (
+ Variant::TimestampNtzMicros(_),
+ DataType::Timestamp(TimeUnit::Nanosecond, None)
+ )
+ | (
+ Variant::TimestampNanos(_),
+ DataType::Timestamp(TimeUnit::Microsecond, Some(_))
Review Comment:
This test wants to cover the happy path(include the other types), will
enhance the test in follow-up for time related pr
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]