Skip to content

[SPARK-20783][SQL][Follow-up] Create ColumnVector to abstract existing compressed column #19508

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
viirya wants to merge 2 commits into from
Closed

[SPARK-20783][SQL][Follow-up] Create ColumnVector to abstract existing compressed column #19508

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5200ef0
Simplify dictionary initialization and remove unused method.
viirya Oct 16, 2017
25003cf
Revert dictionary change for avoiding boxing.
viirya Oct 19, 2017
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
3 changes: 0 additions & 3 deletions sql/core/src/main/scala/org/apache/spark/sql/execution/columnar/ColumnAccessor.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -63,9 +63,6 @@ private[columnar] abstract class BasicColumnAccessor[JvmType](
}

protected def underlyingBuffer = buffer

def getByteBuffer: ByteBuffer =

@kiszk kiszk Oct 18, 2017

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

Good catch, it is not used.

buffer.duplicate.order(ByteOrder.nativeOrder())
}

private[columnar] class NullColumnAccessor(buffer: ByteBuffer)
Expand Down
33 changes: 8 additions & 25 deletions ...c/main/scala/org/apache/spark/sql/execution/columnar/compression/compressionSchemes.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -452,31 +452,10 @@ private[columnar] case object DictionaryEncoding extends CompressionScheme {

class Decoder[T <: AtomicType](buffer: ByteBuffer, columnType: NativeColumnType[T])
extends compression.Decoder[T] {
val elementNum = ByteBufferHelper.getInt(buffer)
private val dictionary: Array[Any] = new Array[Any](elementNum)
private var intDictionary: Array[Int] = null
private var longDictionary: Array[Long] = null

columnType.dataType match {
case _: IntegerType =>
intDictionary = new Array[Int](elementNum)
for (i <- 0 until elementNum) {
val v = columnType.extract(buffer).asInstanceOf[Int]
intDictionary(i) = v
dictionary(i) = v
}
case _: LongType =>
longDictionary = new Array[Long](elementNum)
for (i <- 0 until elementNum) {
val v = columnType.extract(buffer).asInstanceOf[Long]
longDictionary(i) = v
dictionary(i) = v
}
case _: StringType =>
for (i <- 0 until elementNum) {
val v = columnType.extract(buffer).asInstanceOf[Any]
dictionary(i) = v
}

private val dictionary: Array[Any] = {
val elementNum = ByteBufferHelper.getInt(buffer)
Array.fill[Any](elementNum)(columnType.extract(buffer).asInstanceOf[Any])
}

override def next(row: InternalRow, ordinal: Int): Unit = {
Expand All @@ -495,6 +474,8 @@ private[columnar] case object DictionaryEncoding extends CompressionScheme {
columnType.dataType match {
case _: IntegerType =>
val dictionaryIds = columnVector.reserveDictionaryIds(capacity)
val intDictionary = dictionary.map(_.asInstanceOf[Int])

@kiszk kiszk Oct 18, 2017

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

@viirya Could you please see this comment to avoid unboxing? IIUC, this code still causes unboxing.

What do you think?

@viirya viirya Oct 18, 2017

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

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

I didn't notice that comment. Let me check it then. If we can't avoid, I will revert this and only remove the unused method. Thanks.

@viirya viirya Oct 19, 2017

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

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

Thanks @kiszk. I checked the comment and code and reverted this change.


columnVector.setDictionary(new ColumnDictionary(intDictionary))
while (pos < capacity) {
if (pos != nextNullIndex) {
Expand All @@ -508,6 +489,8 @@ private[columnar] case object DictionaryEncoding extends CompressionScheme {
}
case _: LongType =>
val dictionaryIds = columnVector.reserveDictionaryIds(capacity)
val longDictionary = dictionary.map(_.asInstanceOf[Long])

columnVector.setDictionary(new ColumnDictionary(longDictionary))
while (pos < capacity) {
if (pos != nextNullIndex) {
Expand Down