notfilippo commented on issue #11513:
URL: https://github.com/apache/datafusion/issues/11513#issuecomment-2247283135
@jayzhan211 & @doki23 -- I'll try to answer both with a single comment: the
proposal includes (you can see it where LogicalPhysicalType is defined) an
implementation of the `From<DataType>` trait for `LogicalType`. This implies
that if we introduce a LogicalPhysicalSchema type, containing
LogicalPhysicalFields a possible way to convert between types would be:
```rust
let schema = Schema::new(/* ... */);
let logical_physical_schema: LogicalPhysicalSchema = schema.into();
// Suppose that DFSchema is edited to support LogicalPhysicalSchemaRef
instead of SchemaRef
let df_schema = DFSchema::try_from_qualified_schema("t1",
&logical_physical_schema).unwrap();
```
The `into()` call will convert all known arrow::DataTypes in the schema to a
`NativeType` which will keep a reference to the original DataType and its
logical representation.
```rust
let string_view = DataType::StringView;
let logical_string_view: LogicalPhysicalType = string_view.into();
// logical_string_view will be:
// LogicalPhysicalType(Arc(NativeType { logical: Utf8, physical: StringView
}))
```
Instead for user defined types instead of converting an arrow::Schema into a
LogicalPhysicalSchema you would directly define a LogicalPhysicalSchema where
you can use your own arrow-compatible LogicalPhysicalTypes.
--
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]
