Skip to content

[SPARK-31238][SQL] Rebase dates to/from Julian calendar in write/read for ORC datasource #28016

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 7 commits into from
Closed

[SPARK-31238][SQL] Rebase dates to/from Julian calendar in write/read for ORC datasource #28016

Show file tree
Hide file tree
Changes from 4 commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
a2a081a
Add a test
MaxGekk Mar 25, 2020
121b4c6
Fix v2.3 OrcColumnVector
MaxGekk Mar 25, 2020
ecef05d
Fix v1.2 OrcColumnVector
MaxGekk Mar 25, 2020
47d8588
Rebase in write
MaxGekk Mar 25, 2020
29d7966
Copy DaysWritable to v1.2
MaxGekk Mar 26, 2020
3b1b791
Remove `index` variable in OrcColumnVector
MaxGekk Mar 26, 2020
c8a897a
Added a comment to `DaysWritable` v1.2
MaxGekk Mar 26, 2020
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
10 changes: 8 additions & 2 deletions .../apache/spark/sql/hive/DaysWritable.scala → .../execution/datasources/DaysWritable.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@
* limitations under the License.
*/

package org.apache.spark.sql.hive
package org.apache.spark.sql.execution.datasources

import java.io.{DataInput, DataOutput, IOException}
import java.sql.Date
Expand All @@ -35,11 +35,12 @@ import org.apache.spark.sql.catalyst.util.DateTimeUtils.{rebaseGregorianToJulian
* @param julianDays The number of days since the epoch 1970-01-01 in
* Julian calendar.
*/
private[hive] class DaysWritable(
class DaysWritable(
var gregorianDays: Int,
var julianDays: Int)
extends DateWritable {

def this() = this(0, 0)
Copy link
Member Author

@MaxGekk MaxGekk Mar 25, 2020

Choose a reason for hiding this comment

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

I assume that gregorianDays and julianDays will be set later via the set method.

def this(gregorianDays: Int) =
this(gregorianDays, rebaseGregorianToJulianDays(gregorianDays))
def this(dateWritable: DateWritable) = {
Expand All @@ -55,6 +56,11 @@ private[hive] class DaysWritable(
override def getDays: Int = julianDays
override def get(): Date = new Date(DateWritable.daysToMillis(julianDays))

override def set(d: Int): Unit = {
gregorianDays = d
julianDays = rebaseGregorianToJulianDays(d)
}

@throws[IOException]
override def write(out: DataOutput): Unit = {
WritableUtils.writeVInt(out, julianDays)
Expand Down
Binary file added sql/core/src/test/resources/test-data/before_1582_date_v2_4.snappy.orc
View file
Open in desktop
Binary file not shown.
28 changes: 27 additions & 1 deletion sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/orc/OrcSourceSuite.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,7 +19,7 @@ package org.apache.spark.sql.execution.datasources.orc

import java.io.File
import java.nio.charset.StandardCharsets.UTF_8
import java.sql.Timestamp
import java.sql.{Date, Timestamp}
import java.util.Locale

import org.apache.hadoop.conf.Configuration
Expand Down Expand Up @@ -482,6 +482,32 @@ abstract class OrcSuite extends OrcTest with BeforeAndAfterAll {
}
}
}

test("SPARK-31238: compatibility with Spark 2.4 in reading dates") {
Seq(false, true).foreach { vectorized =>
withSQLConf(SQLConf.ORC_VECTORIZED_READER_ENABLED.key -> vectorized.toString) {
checkAnswer(
readResourceOrcFile("test-data/before_1582_date_v2_4.snappy.orc"),
Row(java.sql.Date.valueOf("1200-01-01")))
}
}
}

test("SPARK-31238: rebasing dates in write") {
withTempPath { dir =>
val path = dir.getAbsolutePath
Seq("1001-01-01").toDF("dateS")
.select($"dateS".cast("date").as("date"))
.write
.orc(path)

Seq(false, true).foreach { vectorized =>
withSQLConf(SQLConf.ORC_VECTORIZED_READER_ENABLED.key -> vectorized.toString) {
checkAnswer(spark.read.orc(path), Row(Date.valueOf("1001-01-01")))
}
}
}
}
}

class OrcSourceSuite extends OrcSuite with SharedSparkSession {
Expand Down
5 changes: 5 additions & 0 deletions sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/orc/OrcTest.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -133,4 +133,9 @@ abstract class OrcTest extends QueryTest with FileBasedDataSourceTest with Befor
throw new AnalysisException("Can not match OrcTable in the query.")
}
}

protected def readResourceOrcFile(name: String): DataFrame = {
val url = Thread.currentThread().getContextClassLoader.getResource(name)
spark.read.orc(url.toString)
}
}
16 changes: 15 additions & 1 deletion ...re/v1.2/src/main/java/org/apache/spark/sql/execution/datasources/orc/OrcColumnVector.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,6 +23,7 @@

import org.apache.spark.sql.catalyst.util.DateTimeUtils;
import org.apache.spark.sql.types.DataType;
import org.apache.spark.sql.types.DateType;
import org.apache.spark.sql.types.Decimal;
import org.apache.spark.sql.types.TimestampType;
import org.apache.spark.sql.vectorized.ColumnarArray;
Expand All @@ -42,6 +43,7 @@ public class OrcColumnVector extends org.apache.spark.sql.vectorized.ColumnVecto
private DecimalColumnVector decimalData;
private TimestampColumnVector timestampData;
private final boolean isTimestamp;
private final boolean isDate;

private int batchSize;

Expand All @@ -54,6 +56,12 @@ public class OrcColumnVector extends org.apache.spark.sql.vectorized.ColumnVecto
isTimestamp = false;
}

if (type instanceof DateType) {
isDate = true;
} else {
isDate = false;
}

baseData = vector;
if (vector instanceof LongColumnVector) {
longData = (LongColumnVector) vector;
Expand Down Expand Up @@ -130,7 +138,13 @@ public short getShort(int rowId) {

@Override
public int getInt(int rowId) {
return (int) longData.vector[getRowIndex(rowId)];
int index = getRowIndex(rowId);
int value = (int) longData.vector[index];
Copy link
Contributor

@cloud-fan cloud-fan Mar 26, 2020

Choose a reason for hiding this comment

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

nit: int value = (int) longData.vector[getRowIndex(rowId)];

if (isDate) {
return DateTimeUtils.rebaseJulianToGregorianDays(value);
} else {
return value;
}
}

@Override
Expand Down
5 changes: 3 additions & 2 deletions ...ore/v1.2/src/main/scala/org/apache/spark/sql/execution/datasources/orc/OrcShimUtils.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,6 +26,7 @@ import org.apache.orc.storage.ql.io.sarg.PredicateLeaf.{Operator => OrcOperator}
import org.apache.orc.storage.serde2.io.{DateWritable, HiveDecimalWritable}

import org.apache.spark.sql.catalyst.expressions.SpecializedGetters
import org.apache.spark.sql.execution.datasources.DaysWritable
import org.apache.spark.sql.types.Decimal

/**
Expand All @@ -47,13 +48,13 @@ private[sql] object OrcShimUtils {

def getDateWritable(reuseObj: Boolean): (SpecializedGetters, Int) => DateWritable = {
if (reuseObj) {
val result = new DateWritable()
val result = new DaysWritable()
(getter, ordinal) =>
result.set(getter.getInt(ordinal))
result
} else {
(getter: SpecializedGetters, ordinal: Int) =>
new DateWritable(getter.getInt(ordinal))
new DaysWritable(getter.getInt(ordinal))
}
}

Expand Down
16 changes: 15 additions & 1 deletion ...re/v2.3/src/main/java/org/apache/spark/sql/execution/datasources/orc/OrcColumnVector.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,6 +23,7 @@

import org.apache.spark.sql.catalyst.util.DateTimeUtils;
import org.apache.spark.sql.types.DataType;
import org.apache.spark.sql.types.DateType;
import org.apache.spark.sql.types.Decimal;
import org.apache.spark.sql.types.TimestampType;
import org.apache.spark.sql.vectorized.ColumnarArray;
Expand All @@ -42,6 +43,7 @@ public class OrcColumnVector extends org.apache.spark.sql.vectorized.ColumnVecto
private DecimalColumnVector decimalData;
private TimestampColumnVector timestampData;
private final boolean isTimestamp;
private final boolean isDate;

private int batchSize;

Expand All @@ -54,6 +56,12 @@ public class OrcColumnVector extends org.apache.spark.sql.vectorized.ColumnVecto
isTimestamp = false;
}

if (type instanceof DateType) {
isDate = true;
} else {
isDate = false;
}

baseData = vector;
if (vector instanceof LongColumnVector) {
longData = (LongColumnVector) vector;
Expand Down Expand Up @@ -130,7 +138,13 @@ public short getShort(int rowId) {

@Override
public int getInt(int rowId) {
return (int) longData.vector[getRowIndex(rowId)];
int index = getRowIndex(rowId);
int value = (int) longData.vector[index];
if (isDate) {
return DateTimeUtils.rebaseJulianToGregorianDays(value);
} else {
return value;
}
}

@Override
Expand Down
5 changes: 3 additions & 2 deletions ...ore/v2.3/src/main/scala/org/apache/spark/sql/execution/datasources/orc/OrcShimUtils.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,6 +26,7 @@ import org.apache.hadoop.hive.ql.io.sarg.PredicateLeaf.{Operator => OrcOperator}
import org.apache.hadoop.hive.serde2.io.{DateWritable, HiveDecimalWritable}

import org.apache.spark.sql.catalyst.expressions.SpecializedGetters
import org.apache.spark.sql.execution.datasources.DaysWritable
Copy link
Member

@dongjoon-hyun dongjoon-hyun Mar 26, 2020

Choose a reason for hiding this comment

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

After moving DaysWritable.scala to here. I guess we don't need this line. Please try to remove.

import org.apache.spark.sql.types.Decimal

/**
Expand All @@ -47,13 +48,13 @@ private[sql] object OrcShimUtils {

def getDateWritable(reuseObj: Boolean): (SpecializedGetters, Int) => DateWritable = {
if (reuseObj) {
val result = new DateWritable()
val result = new DaysWritable()
(getter, ordinal) =>
result.set(getter.getInt(ordinal))
result
} else {
(getter: SpecializedGetters, ordinal: Int) =>
new DateWritable(getter.getInt(ordinal))
new DaysWritable(getter.getInt(ordinal))
}
}

Expand Down
1 change: 1 addition & 0 deletions sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveInspectors.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,6 +32,7 @@ import org.apache.spark.sql.AnalysisException
import org.apache.spark.sql.catalyst.InternalRow
import org.apache.spark.sql.catalyst.expressions._
import org.apache.spark.sql.catalyst.util._
import org.apache.spark.sql.execution.datasources.DaysWritable
Copy link
Contributor

@cloud-fan cloud-fan Mar 26, 2020

Choose a reason for hiding this comment

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

unnecessary change

Copy link
Member Author

@MaxGekk MaxGekk Mar 26, 2020

Choose a reason for hiding this comment

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

If I remove it, the build fails with:

Error:(1012, 44) not found: type DaysWritable
  private def getDateWritable(value: Any): DaysWritable =

I moved DaysWritable from sql/hive to sql/core to reuse it in ORC.

import org.apache.spark.sql.types
import org.apache.spark.sql.types._
import org.apache.spark.unsafe.types.UTF8String
Expand Down