Github user rdblue commented on a diff in the pull request:
https://github.com/apache/spark/pull/21308#discussion_r216382704
--- Diff:
sql/core/src/main/java/org/apache/spark/sql/sources/v2/DeleteSupport.java ---
@@ -0,0 +1,46 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.sql.sources.v2;
+
+import org.apache.spark.sql.sources.Filter;
+
+/**
+ * A mix-in interface for {@link DataSourceV2} delete support. Data
sources can implement this
+ * interface to provide the ability to delete data from tables that
matches filter expressions.
+ * <p>
+ * Data sources must implement this interface to support logical
operations that combine writing
+ * data with deleting data, like overwriting partitions.
+ */
+public interface DeleteSupport extends DataSourceV2 {
+ /**
+ * Delete data from a data source table that matches filter expressions.
+ * <p>
+ * Rows are deleted from the data source iff all of the filter
expressions match. That is, the
+ * expressions must be interpreted as a set of filters that are ANDed
together.
+ * <p>
+ * Implementations may reject a delete operation if the delete isn't
possible without significant
+ * effort. For example, partitioned data sources may reject deletes that
do not filter by
+ * partition columns because the filter may require rewriting files
without deleted records.
+ * To reject a delete implementations should throw {@link
IllegalArgumentException} with a clear
+ * error message that identifies which expression was rejected.
+ *
+ * @param filters filter expressions, used to select rows to delete when
all expressions match
+ * @throws IllegalArgumentException If the delete is rejected due to
required effort
+ */
+ void deleteWhere(Filter[] filters);
--- End diff --
Maybe it's a little unclear: this delete is not a write. It is a
driver-side operation using table metadata, like dropping matching partitions
in a Hive table or dropping matching files in an Iceberg table. That way, there
are no tasks and we don't need to use the commit protocol.
If we want to filter data files, the overwrite API I've proposed is the
right way to do it. Spark could read, filter the rows, and replace all of the
files that were read.
If there are files that have both rows that should be removed and rows that
should be kept, the source should throw IllegalArgumentException to reject the
delete.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
