kosiew commented on code in PR #23914:
URL: https://github.com/apache/datafusion/pull/23914#discussion_r3819533457
##########
datafusion/common/src/nested_struct.rs:
##########
@@ -1292,6 +1538,566 @@ mod tests {
assert!(b_col.is_null(1));
}
+ fn map_type(key_type: DataType, value_type: DataType) -> DataType {
+ map_type_with_entry_names(key_type, value_type, "keys", "values",
false)
+ }
+
+ fn map_type_with_entry_names(
+ key_type: DataType,
+ value_type: DataType,
+ key_name: &str,
+ value_name: &str,
+ sorted: bool,
+ ) -> DataType {
+ DataType::Map(
+ Arc::new(non_null_field(
+ "entries",
+ struct_type(vec![
+ non_null_field(key_name, key_type),
+ field(value_name, value_type),
+ ]),
+ )),
+ sorted,
+ )
+ }
+
+ fn struct_map_array() -> ArrayRef {
+ struct_map_array_with_sorted(false)
+ }
+
+ fn struct_map_array_with_sorted(sorted: bool) -> ArrayRef {
+ struct_map_array_with_key_fields(sorted, false)
+ }
+
+ fn struct_map_array_with_key_fields(sorted: bool, include_tenant: bool) ->
ArrayRef {
+ let mut key_fields = vec![(
+ arc_field("id", DataType::Int32),
+ Arc::new(Int32Array::from(vec![1])) as ArrayRef,
+ )];
+ if include_tenant {
+ key_fields.push((
+ arc_field("tenant", DataType::Utf8),
+ Arc::new(StringArray::from(vec!["a"])) as ArrayRef,
+ ));
+ }
+ let keys = StructArray::from(key_fields);
+ let values = StructArray::from(vec![
+ (
+ arc_field("amount", DataType::Int32),
+ Arc::new(Int32Array::from(vec![10])) as ArrayRef,
+ ),
+ (
+ arc_field("ignored", DataType::Utf8),
+ Arc::new(StringArray::from(vec!["x"])) as ArrayRef,
+ ),
+ ]);
+ let entries = StructArray::new(
+ vec![
+ Arc::new(non_null_field("keys", keys.data_type().clone())),
+ arc_field("values", values.data_type().clone()),
+ ]
+ .into(),
+ vec![Arc::new(keys), Arc::new(values)],
+ None,
+ );
+ Arc::new(MapArray::new(
+ Arc::new(non_null_field("entries", entries.data_type().clone())),
+ OffsetBuffer::new(vec![0, 1, 1].into()),
+ entries,
+ Some(NullBuffer::from(vec![true, false])),
+ sorted,
+ ))
+ }
+
+ fn nested_struct_map_array(
+ key_name: &str,
+ value_name: &str,
+ sorted: bool,
+ ) -> ArrayRef {
+ let key_nested = StructArray::from(vec![(
+ arc_field("id", DataType::Int32),
+ Arc::new(Int32Array::from(vec![1])) as ArrayRef,
+ )]);
+ let keys = StructArray::from(vec![(
+ arc_field("nested", key_nested.data_type().clone()),
+ Arc::new(key_nested) as ArrayRef,
+ )]);
+ let value_nested = StructArray::from(vec![(
+ arc_field("amount", DataType::Int32),
+ Arc::new(Int32Array::from(vec![10])) as ArrayRef,
+ )]);
+ let values = StructArray::from(vec![(
+ arc_field("nested", value_nested.data_type().clone()),
+ Arc::new(value_nested) as ArrayRef,
+ )]);
+ let entries = StructArray::new(
+ vec![
+ Arc::new(non_null_field(key_name, keys.data_type().clone())),
+ arc_field(value_name, values.data_type().clone()),
+ ]
+ .into(),
+ vec![Arc::new(keys), Arc::new(values)],
+ None,
+ );
+ Arc::new(MapArray::new(
+ Arc::new(non_null_field("entries", entries.data_type().clone())),
+ OffsetBuffer::new(vec![0, 1].into()),
+ entries,
+ None,
+ sorted,
+ ))
+ }
+
+ #[test]
+ fn test_map_entry_names_match_positionally_and_adapt_nested_structs() {
+ let source_col = nested_struct_map_array("source_keys",
"source_values", false);
+ let target_type = map_type_with_entry_names(
+ struct_type(vec![field(
+ "nested",
+ struct_type(vec![
+ field("id", DataType::Int64),
+ field("label", DataType::Utf8),
+ ]),
+ )]),
+ struct_type(vec![field(
+ "nested",
+ struct_type(vec![
+ field("amount", DataType::Int64),
+ field("currency", DataType::Utf8),
+ ]),
+ )]),
+ "key",
+ "value",
+ false,
+ );
+
+ assert!(
+ validate_data_type_compatibility(
+ "map_col",
+ source_col.data_type(),
+ &target_type
+ )
+ .is_ok()
+ );
+ let result =
+ cast_column(&source_col, &target_type,
&DEFAULT_CAST_OPTIONS).unwrap();
+ let map = result.as_any().downcast_ref::<MapArray>().unwrap();
+ let (key_field, value_field) = map.entries_fields();
+ assert_eq!(key_field.name(), "key");
+ assert_eq!(value_field.name(), "value");
+ let keys = map.keys().as_struct();
+ let key_nested = keys.column_by_name("nested").unwrap().as_struct();
+ assert_eq!(get_column_as!(key_nested, "id", Int64Array).value(0), 1);
+ assert!(get_column_as!(key_nested, "label", StringArray).is_null(0));
+ let values = map.values().as_struct();
+ let value_nested =
values.column_by_name("nested").unwrap().as_struct();
+ assert_eq!(
+ get_column_as!(value_nested, "amount", Int64Array).value(0),
+ 10
+ );
+ assert!(get_column_as!(value_nested, "currency",
StringArray).is_null(0));
+ }
+
+ #[test]
+ fn
test_unsorted_map_key_struct_field_removal_rejected_by_planner_and_runtime() {
+ let source_col = struct_map_array_with_key_fields(false, true);
+ let target_type = map_type(
+ struct_type(vec![field("id", DataType::Int32)]),
+ struct_type(vec![
+ field("amount", DataType::Int32),
+ field("ignored", DataType::Utf8),
+ ]),
+ );
+ assert_map_planning_runtime_error(
+ &source_col,
+ &target_type,
+ "Cannot remove field 'tenant' from a Map key Struct",
+ );
+ }
+
+ #[test]
+ fn test_unsorted_map_non_injective_key_cast_rejected() {
Review Comment:
[added planner+runtime Map(Int32, Utf8) -> Map(Int64, Utf8)
test](https://github.com/apache/datafusion/pull/23914/commits/f73a2ee1f9cec6f3754f8d7072ca623fcedbdfaa)
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
