Skip to content

[SPARK-25425][SQL] Extra options should override session options in DataSource V2 #22413

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
MaxGekk wants to merge 9 commits into from
Closed

[SPARK-25425][SQL] Extra options should override session options in DataSource V2 #22413

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
13729b5
Session options shouldn't override extra options
MaxGekk Sep 13, 2018
14498fb
Extra options should overwrite session options in save
MaxGekk Sep 13, 2018
a443054
Remove a line
MaxGekk Sep 13, 2018
48d4cef
Test for load()
MaxGekk Sep 14, 2018
83789c6
Renaming DataSourceWithSessionConfV2 to DataSourceV2WithSessionConfigs
MaxGekk Sep 15, 2018
eba46d9
Code de-duplication in tests
MaxGekk Sep 15, 2018
96990c7
Add a test case for write path.
dongjoon-hyun Sep 15, 2018
d35d01b
rename
dongjoon-hyun Sep 15, 2018
325b9c4
Merge pull request #8 from dongjoon-hyun/SPARK-25425
MaxGekk Sep 15, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion sql/core/src/main/scala/org/apache/spark/sql/DataFrameReader.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -202,7 +202,7 @@ class DataFrameReader private[sql](sparkSession: SparkSession) extends Logging {
DataSourceOptions.PATHS_KEY -> objectMapper.writeValueAsString(paths.toArray)
}
Dataset.ofRows(sparkSession, DataSourceV2Relation.create(
ds, extraOptions.toMap ++ sessionOptions + pathsOption,
ds, sessionOptions ++ extraOptions.toMap + pathsOption,
Copy link
Member

@dongjoon-hyun dongjoon-hyun Sep 13, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Also, cc @rdblue since this is introduced at aadf953#diff-f70bda59304588cc3abfa3a9840653f4R197 .

Copy link
Member Author

@MaxGekk MaxGekk Sep 13, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@dongjoon-hyun The commit didn't change semantic actually. Before it was:

(extraOptions ++
        DataSourceV2Utils.extractSessionConfigs(
          ds = ds.asInstanceOf[DataSourceV2],
          conf = sparkSession.sessionState.conf))

Copy link
Member

@dongjoon-hyun dongjoon-hyun Sep 13, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Oh. It has more history. Thanks, @MaxGekk . Could you trace down when it started? We need to mark the affected version correctly in order to know the backport candidates.

Copy link
Member

@dongjoon-hyun dongjoon-hyun Sep 13, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Also, please add a test case for this. If this has a long history than SPARK-23203 (fixed at 2.4.0), we need to verify this during backporting.

Copy link
Member Author

@MaxGekk MaxGekk Sep 14, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The changes were added in #19861

Copy link
Member

@dongjoon-hyun dongjoon-hyun Sep 15, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you. Then, it was 2.3.0.

userSpecifiedSchema = userSpecifiedSchema))
} else {
loadV1Source(paths: _*)
Expand Down
8 changes: 5 additions & 3 deletions sql/core/src/main/scala/org/apache/spark/sql/DataFrameWriter.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -241,10 +241,12 @@ final class DataFrameWriter[T] private[sql](ds: Dataset[T]) {
val source = cls.newInstance().asInstanceOf[DataSourceV2]
source match {
case provider: BatchWriteSupportProvider =>
val options = extraOptions ++
DataSourceV2Utils.extractSessionConfigs(source, df.sparkSession.sessionState.conf)
val sessionOptions = DataSourceV2Utils.extractSessionConfigs(
source,
df.sparkSession.sessionState.conf)
val options = sessionOptions ++ extraOptions

val relation = DataSourceV2Relation.create(source, options.toMap)
val relation = DataSourceV2Relation.create(source, options)
Copy link
Member

@dongjoon-hyun dongjoon-hyun Sep 13, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Both read/write-side tests.

Copy link
Member

@HyukjinKwon HyukjinKwon Sep 14, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

+1 for tests

Copy link
Member Author

@MaxGekk MaxGekk Sep 14, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I wrote a test for read path since I was able to grab options propagated to DataSource but I have no idea so far for write path, only mocking probably. Does it make sense to do that?

Copy link
Member

@dongjoon-hyun dongjoon-hyun Sep 15, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

For write path, I think we can use the traditional way instead of introducing mocking. Let me try.

if (mode == SaveMode.Append) {
runCommand(df.sparkSession, "save") {
AppendData.byName(relation, df.logicalPlan)
Expand Down