[GitHub] spark pull request #21308: SPARK-24253: Add DeleteSupport mix-in for DataSou...

[rdblue]

Github user rdblue commented on a diff in the pull request:

https://github.com/apache/spark/pull/21308#discussion_r210382412
  
--- Diff: 
sql/core/src/main/java/org/apache/spark/sql/sources/v2/DeleteSupport.java ---
@@ -0,0 +1,51 @@
+/*
+ * 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.catalyst.expressions.Expression;
+
+/**
+ * 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.
+ * 
+ * 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.
+   * 
+   * 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.
+   * 
+   * 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.
+   * 
+   * Implementations may throw {@link UnsupportedOperationException} if 
the delete operation is not
--- End diff --

After updating this to use Filter, the UnsupportedOperationException is no 
longer needed, so I removed it. That should also cut down on the confusion here.


---

-
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

[GitHub] spark pull request #21308: SPARK-24253: Add DeleteSupport mix-in for DataSou...

[rdblue]

Github user rdblue commented on a diff in the pull request:

https://github.com/apache/spark/pull/21308#discussion_r191021410
  
--- Diff: 
sql/core/src/main/java/org/apache/spark/sql/sources/v2/DeleteSupport.java ---
@@ -0,0 +1,51 @@
+/*
+ * 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.catalyst.expressions.Expression;
+
+/**
+ * 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.
+ * 
+ * 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.
+   * 
+   * 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.
+   * 
+   * 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.
+   * 
+   * Implementations may throw {@link UnsupportedOperationException} if 
the delete operation is not
+   * supported because one of the filter expressions is not supported. 
Implementations should throw
+   * this exception with a clear error message that identifies the 
unsupported expression.
+   *
+   * @param filters filter expressions, used to select rows to delete when 
all expressions match
+   * @throws UnsupportedOperationException If one or more filter 
expressions is not supported
+   * @throws IllegalArgumentException If the delete is rejected due to 
required effort
+   */
+  void deleteWhere(Expression[] filters);
--- End diff --

There aren't necessarily partitions in these data sources, so I wouldn't 
add partitions to the method name. I think we can make this more clear with 
better docs though.


---

-
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

[GitHub] spark pull request #21308: SPARK-24253: Add DeleteSupport mix-in for DataSou...

[bersprockets]

Github user bersprockets commented on a diff in the pull request:

https://github.com/apache/spark/pull/21308#discussion_r190963247
  
--- Diff: 
sql/core/src/main/java/org/apache/spark/sql/sources/v2/DeleteSupport.java ---
@@ -0,0 +1,51 @@
+/*
+ * 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.catalyst.expressions.Expression;
+
+/**
+ * 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.
+ * 
+ * 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.
+   * 
+   * 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.
+   * 
+   * 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.
+   * 
+   * Implementations may throw {@link UnsupportedOperationException} if 
the delete operation is not
+   * supported because one of the filter expressions is not supported. 
Implementations should throw
+   * this exception with a clear error message that identifies the 
unsupported expression.
+   *
+   * @param filters filter expressions, used to select rows to delete when 
all expressions match
+   * @throws UnsupportedOperationException If one or more filter 
expressions is not supported
+   * @throws IllegalArgumentException If the delete is rejected due to 
required effort
+   */
+  void deleteWhere(Expression[] filters);
--- End diff --

>Do you think it would be more clear if this were explicitly a driver-side 
operation?

Possibly. Maybe in the big data world this is already obvious. To me, it 
looks like a general purpose delete. Maybe deletePartitions? (I am bad at 
naming things, however).


---

-
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

[GitHub] spark pull request #21308: SPARK-24253: Add DeleteSupport mix-in for DataSou...

[rdblue]

Github user rdblue commented on a diff in the pull request:

https://github.com/apache/spark/pull/21308#discussion_r190723800
  
--- Diff: 
sql/core/src/main/java/org/apache/spark/sql/sources/v2/DeleteSupport.java ---
@@ -0,0 +1,51 @@
+/*
+ * 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.catalyst.expressions.Expression;
+
+/**
+ * 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.
+ * 
+ * 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.
+   * 
+   * 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.
+   * 
+   * 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.
+   * 
+   * Implementations may throw {@link UnsupportedOperationException} if 
the delete operation is not
+   * supported because one of the filter expressions is not supported. 
Implementations should throw
+   * this exception with a clear error message that identifies the 
unsupported expression.
+   *
+   * @param filters filter expressions, used to select rows to delete when 
all expressions match
+   * @throws UnsupportedOperationException If one or more filter 
expressions is not supported
+   * @throws IllegalArgumentException If the delete is rejected due to 
required effort
+   */
+  void deleteWhere(Expression[] filters);
--- End diff --

Do you think it would be more clear if this were explicitly a driver-side 
operation?


---

-
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

[GitHub] spark pull request #21308: SPARK-24253: Add DeleteSupport mix-in for DataSou...

[rdblue]

Github user rdblue commented on a diff in the pull request:

https://github.com/apache/spark/pull/21308#discussion_r190712870
  
--- Diff: 
sql/core/src/main/java/org/apache/spark/sql/sources/v2/DeleteSupport.java ---
@@ -0,0 +1,51 @@
+/*
+ * 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.catalyst.expressions.Expression;
+
+/**
+ * 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.
+ * 
+ * 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.
+   * 
+   * 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.
+   * 
+   * 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.
+   * 
+   * Implementations may throw {@link UnsupportedOperationException} if 
the delete operation is not
--- End diff --

No, these are distinct.

`UnsupportedOperationException` indicates that the source doesn't 
understand a filter. For example, it could be `date(ts) = '2018-05-13'` and the 
source doesn't support the conversion from timestamp to date.

`IllegalArgumentException` is thrown when the expression is understood by 
the source, but the work required to perform the delete is not supported. For 
example, if you have data partitioned by `hour(ts)` and the delete expression 
is `ts > '2018-05-13T00:05:00 and ts < '2018-05-13T00:10:00'`. Deleting a 
5-minute window when data is partitioned by hour probably isn't possible 
without rewriting data files, so the source can reject it.


---

-
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

[GitHub] spark pull request #21308: SPARK-24253: Add DeleteSupport mix-in for DataSou...

[jose-torres]

Github user jose-torres commented on a diff in the pull request:

https://github.com/apache/spark/pull/21308#discussion_r190699855
  
--- Diff: 
sql/core/src/main/java/org/apache/spark/sql/sources/v2/DeleteSupport.java ---
@@ -0,0 +1,51 @@
+/*
+ * 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.catalyst.expressions.Expression;
+
+/**
+ * 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.
+ * 
+ * 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.
+   * 
+   * 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.
+   * 
+   * 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.
+   * 
+   * Implementations may throw {@link UnsupportedOperationException} if 
the delete operation is not
--- End diff --

nit: is this a duplicate of the above paragraph


---

-
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

[GitHub] spark pull request #21308: SPARK-24253: Add DeleteSupport mix-in for DataSou...

[rdblue]

Github user rdblue commented on a diff in the pull request:

https://github.com/apache/spark/pull/21308#discussion_r189730152
  
--- Diff: 
sql/core/src/main/java/org/apache/spark/sql/sources/v2/DeleteSupport.java ---
@@ -0,0 +1,51 @@
+/*
+ * 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.catalyst.expressions.Expression;
+
+/**
+ * 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.
+ * 
+ * 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.
+   * 
+   * 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.
+   * 
+   * 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.
+   * 
+   * Implementations may throw {@link UnsupportedOperationException} if 
the delete operation is not
+   * supported because one of the filter expressions is not supported. 
Implementations should throw
+   * this exception with a clear error message that identifies the 
unsupported expression.
+   *
+   * @param filters filter expressions, used to select rows to delete when 
all expressions match
+   * @throws UnsupportedOperationException If one or more filter 
expressions is not supported
+   * @throws IllegalArgumentException If the delete is rejected due to 
required effort
+   */
+  void deleteWhere(Expression[] filters);
--- End diff --

Yes, this is a driver-side operation. That's why the source can reject the 
delete. Anything that requires a parallel operation should really be 
implemented as read, filter, and replace data.


---

-
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

[GitHub] spark pull request #21308: SPARK-24253: Add DeleteSupport mix-in for DataSou...

[bersprockets]

Github user bersprockets commented on a diff in the pull request:

https://github.com/apache/spark/pull/21308#discussion_r188392219
  
--- Diff: 
sql/core/src/main/java/org/apache/spark/sql/sources/v2/DeleteSupport.java ---
@@ -0,0 +1,51 @@
+/*
+ * 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.catalyst.expressions.Expression;
+
+/**
+ * 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.
+ * 
+ * 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.
+   * 
+   * 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.
+   * 
+   * 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.
+   * 
+   * Implementations may throw {@link UnsupportedOperationException} if 
the delete operation is not
+   * supported because one of the filter expressions is not supported. 
Implementations should throw
+   * this exception with a clear error message that identifies the 
unsupported expression.
+   *
+   * @param filters filter expressions, used to select rows to delete when 
all expressions match
+   * @throws UnsupportedOperationException If one or more filter 
expressions is not supported
+   * @throws IllegalArgumentException If the delete is rejected due to 
required effort
+   */
+  void deleteWhere(Expression[] filters);
--- End diff --

Does putting the delete method here (as opposed to say, in DataDeleters on 
some other thing parallel to to the DataWriters) imply that this is a 
driver-side operation only? I understand the use case is deleting partitions 
which is usually only a file system operation, but will that always be the case?


---

-
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org

[GitHub] spark pull request #21308: SPARK-24253: Add DeleteSupport mix-in for DataSou...

[rdblue]

GitHub user rdblue opened a pull request:

https://github.com/apache/spark/pull/21308

SPARK-24253: Add DeleteSupport mix-in for DataSourceV2.

## What changes were proposed in this pull request?

Adds `DeleteSupport` mix-in for `DataSourceV2`. This mix-in provides a 
method to delete data with catalyst expressions in support of `delete from` and 
overwrite logical operations.

## How was this patch tested?

No tests, this adds an interface.

You can merge this pull request into a Git repository by running:

$ git pull https://github.com/rdblue/spark SPARK-24253-add-v2-delete-support

Alternatively you can review and apply these changes as the patch at:

https://github.com/apache/spark/pull/21308.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

This closes #21308


commit c0243cd5807142a3c61e5615406842d9d97bf7de
Author: Ryan Blue 
Date:   2018-05-11T22:04:15Z

SPARK-24253: Add DeleteSupport mix-in for DataSourceV2.




---

-
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org