Github user rdblue commented on a diff in the pull request:
https://github.com/apache/spark/pull/20387#discussion_r166386661
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/v2/DataSourceV2Relation.scala
---
@@ -17,17 +17,151 @@
package org.apache.spark.sql.execution.datasources.v2
+import java.util.UUID
+
+import scala.collection.JavaConverters._
+import scala.collection.mutable
+
+import org.apache.spark.sql.{AnalysisException, SaveMode}
+import org.apache.spark.sql.catalyst.TableIdentifier
import org.apache.spark.sql.catalyst.analysis.MultiInstanceRelation
-import org.apache.spark.sql.catalyst.expressions.AttributeReference
-import org.apache.spark.sql.catalyst.plans.logical.{LeafNode, Statistics}
-import org.apache.spark.sql.sources.v2.reader._
+import org.apache.spark.sql.catalyst.expressions.{AttributeReference,
Expression}
+import org.apache.spark.sql.catalyst.plans.logical.{LeafNode, LogicalPlan,
Statistics}
+import org.apache.spark.sql.execution.datasources.DataSourceStrategy
+import org.apache.spark.sql.sources.{DataSourceRegister, Filter}
+import org.apache.spark.sql.sources.v2.{DataSourceOptions, DataSourceV2,
ReadSupport, ReadSupportWithSchema, WriteSupport}
+import org.apache.spark.sql.sources.v2.reader.{DataSourceReader,
SupportsPushDownCatalystFilters, SupportsPushDownFilters,
SupportsPushDownRequiredColumns, SupportsReportStatistics}
+import org.apache.spark.sql.sources.v2.writer.DataSourceWriter
+import org.apache.spark.sql.types.StructType
case class DataSourceV2Relation(
- fullOutput: Seq[AttributeReference],
- reader: DataSourceReader)
- extends LeafNode with MultiInstanceRelation with DataSourceReaderHolder {
+ source: DataSourceV2,
+ options: Map[String, String],
+ path: Option[String] = None,
+ table: Option[TableIdentifier] = None,
--- End diff --
I think that `TableIdentifier` and a string-to-string `Map` should be
passed to `DataSourceV2Relation` and that either the relation or
`DataSourceOptions` should be responsible for creating `DataSourceOptions` with
well-defined properties to to pass the table information to implementations.
This minimizes the number of places that need to handle `DataSourceOptions`
(which is specific to v2) and uses `TableIdentifier` and `Map` to match the
rest of the planner nodes. For example, other read paths that can create
`DataSourceV2Relation`, like resolution rules, use `TableIdentifier`.
I'm not currently advocating a position on how to configure
`DataFrameReader` or `DataFrameWriter` or how to handle schemas.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
