rdblue commented on a change in pull request #23266: [SPARK-26313][SQL] move
read related methods from Table to read related mix-in traits
URL: https://github.com/apache/spark/pull/23266#discussion_r240407918
##########
File path:
sql/core/src/main/java/org/apache/spark/sql/sources/v2/SupportsBatchRead.java
##########
@@ -20,14 +20,27 @@
import org.apache.spark.annotation.Evolving;
import org.apache.spark.sql.sources.v2.reader.Scan;
import org.apache.spark.sql.sources.v2.reader.ScanBuilder;
+import org.apache.spark.sql.types.StructType;
/**
- * An empty mix-in interface for {@link Table}, to indicate this table
supports batch scan.
- * <p>
- * If a {@link Table} implements this interface, its {@link
Table#newScanBuilder(DataSourceOptions)}
- * must return a {@link ScanBuilder} that builds {@link Scan} with {@link
Scan#toBatch()}
- * implemented.
- * </p>
+ * A mix-in interface for {@link Table} to provide data reading ability of
batch processing.
*/
@Evolving
-public interface SupportsBatchRead extends Table { }
+public interface SupportsBatchRead extends Table {
+
+ /**
+ * Returns the schema of this table.
+ */
+ StructType schema();
Review comment:
I think `Table` should also have a `schema`. I agree that it is named, but
it is also a data source or sink of some kind, with a schema. I don't see the
value in moving the schema to a different interface, and I think that moving
the schema to an interface specific to the read path is worse because it causes
the problem that a table must be readable to be writable.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
