Report Beam Lineage from Parquet reads #5850
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,124 @@ | ||
| /* | ||
| * Copyright 2026 Spotify AB. | ||
| * | ||
| * Licensed 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 com.spotify.scio.parquet | ||
|
|
||
| import com.spotify.scio.ScioContext | ||
| import com.spotify.scio.coders.{Coder, CoderMaterializer} | ||
| import com.spotify.scio.util.ScioUtil | ||
| import com.spotify.scio.values.SCollection | ||
| import com.twitter.chill.ClosureCleaner | ||
| import org.apache.beam.sdk.io.FileSystems | ||
| import org.apache.beam.sdk.io.hadoop.format.HadoopFormatIO | ||
| import org.apache.beam.sdk.transforms.DoFn.ProcessElement | ||
| import org.apache.beam.sdk.transforms.{DoFn, SimpleFunction} | ||
| import org.apache.beam.sdk.values.TypeDescriptor | ||
| import org.apache.hadoop.conf.Configuration | ||
| import org.slf4j.LoggerFactory | ||
|
|
||
| import java.util.concurrent.atomic.AtomicBoolean | ||
| import scala.reflect.ClassTag | ||
|
|
||
| private[parquet] object HadoopParquet { | ||
|
|
||
| /** Read data from Hadoop format using HadoopFormatIO. */ | ||
| def readHadoopFormatIO[A: ClassTag, T: ClassTag]( | ||
| sc: ScioContext, | ||
| conf: Configuration, | ||
| projectionFn: Option[A => T], | ||
| skipValueClone: Option[Boolean] | ||
| )(implicit coder: Coder[T]): SCollection[T] = { | ||
| val inputType = ScioUtil.classOf[A] | ||
| val outputType = ScioUtil.classOf[T] | ||
| val bcoder = CoderMaterializer.beam(sc, coder) | ||
|
|
||
| val hadoop = HadoopFormatIO | ||
| .read[java.lang.Boolean, T]() | ||
| // Hadoop input always emit key-value, and `Void` causes NPE in Beam coder | ||
| .withKeyTranslation(new SimpleFunction[Void, java.lang.Boolean]() { | ||
| override def apply(input: Void): java.lang.Boolean = true | ||
| }) | ||
|
|
||
| val withSkipClone = skipValueClone.fold(hadoop)(skip => hadoop.withSkipValueClone(skip)) | ||
|
|
||
|
|
||
| val withValueTranslation = projectionFn.fold { | ||
| withSkipClone.withValueTranslation( | ||
| new SimpleFunction[T, T]() { | ||
| override def apply(input: T): T = input | ||
|
|
||
| override def getInputTypeDescriptor = TypeDescriptor.of(outputType) | ||
|
|
||
| override def getOutputTypeDescriptor = TypeDescriptor.of(outputType) | ||
| }, | ||
| bcoder | ||
| ) | ||
| } { fn => | ||
| val g = ClosureCleaner.clean(fn) // defeat closure | ||
| withSkipClone.withValueTranslation( | ||
| new SimpleFunction[A, T]() { | ||
| // Workaround for incomplete Avro objects | ||
| // `SCollection#map` might throw NPE on incomplete Avro objects when the runner tries | ||
| // to serialized them. Lifting the mapping function here fixes the problem. | ||
| override def apply(input: A): T = g(input) | ||
|
|
||
| override def getInputTypeDescriptor = TypeDescriptor.of(inputType) | ||
|
|
||
| override def getOutputTypeDescriptor = TypeDescriptor.of(outputType) | ||
| }, | ||
| bcoder | ||
| ) | ||
| } | ||
|
|
||
| sc.applyTransform(withValueTranslation.withConfiguration(conf)).map(_.getValue) | ||
| } | ||
| } | ||
|
|
||
| private[parquet] object LineageReportDoFn { | ||
| // Atomic flag to ensure lineage is reported only once per JVM | ||
| private val lineageReported = new AtomicBoolean(false) | ||
| } | ||
|
|
||
| /** | ||
| * DoFn that reports directory-level source lineage for legacy Parquet reads. | ||
| * | ||
| * @param filePattern | ||
| * The file pattern or path to report lineage for | ||
| */ | ||
| private[parquet] class LineageReportDoFn[T](filePattern: String) extends DoFn[T, T] { | ||
|
|
||
| @transient | ||
| private lazy val logger = LoggerFactory.getLogger(this.getClass) | ||
|
|
||
| @ProcessElement | ||
| def processElement(@DoFn.Element element: T, out: DoFn.OutputReceiver[T]): Unit = { | ||
| if (LineageReportDoFn.lineageReported.compareAndSet(false, true)) { | ||
| try { | ||
| val isDirectory = filePattern.endsWith("/") | ||
| val resourceId = FileSystems.matchNewResource(filePattern, isDirectory) | ||
| val directory = resourceId.getCurrentDirectory | ||
| FileSystems.reportSourceLineage(directory) | ||
| } catch { | ||
| case e: Exception => | ||
| logger.warn( | ||
| s"Error when reporting lineage for pattern: $filePattern", | ||
| e | ||
| ) | ||
| } | ||
| } | ||
| out.output(element) | ||
| } | ||
| } | ||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is just to reduce duplication or is there a functional change here?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
just to reduce, no new functionality, except I noticed that in some cases Scio's derived coder was not set to HadoopFormatIO transformation. Probably Beam auto-derives the same coder, but anyway it is better to set explicitly