embulk-output-orc 0.3.4 → 0.3.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 725b869f19175110dd6e542619f5bfd9c8c0168b
-  data.tar.gz: 9c8016b015f380815cb2ad7127c192243ebfa45a
+  metadata.gz: 6ecb39bb650455937f641f073e9e0b13338f268b
+  data.tar.gz: 393ef796dfdf47239a11186b33988466432b0d02
 SHA512:
-  metadata.gz: 197da7bd52b19ccc2c3f30ad0d2c2c36446fa00e48426de9bb333b0ef2a8eddc213f0db947e50215297333a933e5d6d3b63ad0c0388f78e19d984fae6810bb99
-  data.tar.gz: ccf9bc5b8f73c8d0f48e7d17e16dd4b113b3f9b7a1b9178045109ef3476f4cb47f58e0699e4a9516ab81830d59a4bae3be2fba8f936c4cef0f705f6acc384f2d
+  metadata.gz: ee733e3cca10bfff236c7ff24d3249a1f8a30629ba314a9b840a2de4d6b552412fa1a6cb9555979dbad499259152b5a24439586105ce0417f629079b26775e9b
+  data.tar.gz: 41f8059f0af1f7eb1accccb18e33c111f7e75b7a69ffba97b26cd74de7349922ebdd852b87ea31f40c257cc57543b88cada9e32aea828f7f8d852adf20d3f328

data/.github/workflows/gradle.yml

@@ -0,0 +1,25 @@
+name: Java CI
+
+on: [push]
+
+jobs:
+  build:
+
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v1
+    - name: Set up JDK 1.8
+      uses: actions/setup-java@v1
+      with:
+        java-version: 1.8
+    - name: Build with Gradle
+      run:
+        ./gradlew build
+    - name: Checkstyle & static check
+      run: |
+        ./gradlew --info checkstyle
+        ./gradlew --info check
+#    - name: Spotbugs
+#      run: |
+#        ./gradlew spotbugsMain spotbugsTest

data/README.md

@@ -1,6 +1,6 @@
 # Orc output plugin for Embulk
 
-[![Build Status](https://travis-ci.org/yuokada/embulk-output-orc.svg?branch=master)](https://travis-ci.org/yuokada/embulk-output-orc)
+[![Build Status](https://github.com/yuokada/embulk-output-orc/workflows/Java%20CI/badge.svg)](https://github.com/yuokada/embulk-output-orc/actions)
 [![Gem Version](https://badge.fury.io/rb/embulk-output-orc.svg)](https://badge.fury.io/rb/embulk-output-orc)
 
 ## Overview

data/build.gradle

@@ -2,7 +2,9 @@ plugins {
     id "com.jfrog.bintray" version "1.1"
     id "com.github.jruby-gradle.base" version "1.5.0"
     id "java"
+    id "scala"
     id "checkstyle"
+//    id "com.github.spotbugs" version "3.0.1"
     id "org.sonarqube" version "2.5"
 }
 import com.github.jrubygradle.JRubyExec
@@ -18,14 +20,15 @@ configurations {
     runtime.exclude group: "org.slf4j", module: "slf4j-log4j12"
 }
 
-version = "0.3.4"
+version = "0.3.5"
 
 sourceCompatibility = 1.8
 targetCompatibility = 1.8
 
 dependencies {
-    compile  "org.embulk:embulk-core:0.8.34"
-    provided "org.embulk:embulk-core:0.8.34"
+    compile  "org.embulk:embulk-core:0.9.23"
+    provided "org.embulk:embulk-core:0.9.23"
+    compile  "org.scala-lang:scala-library:2.12.+"
 
     compile "org.apache.orc:orc:1.5.4"
     compile "org.apache.orc:orc-core:1.5.4"
@@ -34,14 +37,24 @@ dependencies {
     compile 'org.embulk.input.s3:embulk-util-aws-credentials:0.2.8'
     compile "com.amazonaws:aws-java-sdk-s3:1.10.33"
     compile "org.apache.hadoop:hadoop-aws:2.7.5"
-    compile 'com.google.guava:guava:24.1-jre'
 
     testCompile 'org.jmockit:jmockit:1.38'
     // testCompile "junit:junit:4.+"
     testCompile 'org.hamcrest:hamcrest-core:1.3'
     testCompile 'org.testng:testng:6.14.2'
-    testCompile "org.embulk:embulk-core:0.8.34:tests"
-    testCompile "org.embulk:embulk-standards:0.8.34"
+    testCompile "org.embulk:embulk-core:0.8.39:tests"
+    testCompile "org.embulk:embulk-standards:0.8.39"
+}
+
+sourceSets {
+    main {
+        scala {
+            srcDirs = ['src/main/scala', 'src/main/java']
+        }
+        java {
+            srcDirs = []
+        }
+    }
 }
 
 task classpath(type: Copy, dependsOn: ["jar"]) {

data/gradle/wrapper/gradle-wrapper.properties

@@ -1,6 +1,5 @@
-#Wed Jan 09 23:08:09 JST 2019
 distributionBase=GRADLE_USER_HOME
 distributionPath=wrapper/dists
+distributionUrl=https\://services.gradle.org/distributions/gradle-5.6.4-bin.zip
 zipStoreBase=GRADLE_USER_HOME
 zipStorePath=wrapper/dists
-distributionUrl=https\://services.gradle.org/distributions/gradle-4.10-all.zip

data/gradlew

@@ -1,5 +1,21 @@
 #!/usr/bin/env sh
 
+#
+# Copyright 2015 the original author or authors.
+#
+# 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
+#
+#      https://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.
+#
+
 ##############################################################################
 ##
 ##  Gradle start up script for UN*X
@@ -28,7 +44,7 @@ APP_NAME="Gradle"
 APP_BASE_NAME=`basename "$0"`
 
 # Add default JVM options here. You can also use JAVA_OPTS and GRADLE_OPTS to pass JVM options to this script.
-DEFAULT_JVM_OPTS=""
+DEFAULT_JVM_OPTS='"-Xmx64m" "-Xms64m"'
 
 # Use the maximum available, or set MAX_FD != -1 to use that value.
 MAX_FD="maximum"

data/gradlew.bat

@@ -1,3 +1,19 @@
+@rem
+@rem Copyright 2015 the original author or authors.
+@rem
+@rem Licensed under the Apache License, Version 2.0 (the "License");
+@rem you may not use this file except in compliance with the License.
+@rem You may obtain a copy of the License at
+@rem
+@rem      https://www.apache.org/licenses/LICENSE-2.0
+@rem
+@rem Unless required by applicable law or agreed to in writing, software
+@rem distributed under the License is distributed on an "AS IS" BASIS,
+@rem WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+@rem See the License for the specific language governing permissions and
+@rem limitations under the License.
+@rem
+
 @if "%DEBUG%" == "" @echo off
 @rem ##########################################################################
 @rem
@@ -14,7 +30,7 @@ set APP_BASE_NAME=%~n0
 set APP_HOME=%DIRNAME%
 
 @rem Add default JVM options here. You can also use JAVA_OPTS and GRADLE_OPTS to pass JVM options to this script.
-set DEFAULT_JVM_OPTS=
+set DEFAULT_JVM_OPTS="-Xmx64m" "-Xms64m"
 
 @rem Find java.exe
 if defined JAVA_HOME goto findJavaFromJavaHome

data/src/main/scala/org/embulk/output/orc/OrcColumnVisitor.scala

@@ -0,0 +1,42 @@
+package org.embulk.output.orc
+
+import java.nio.charset.StandardCharsets
+
+import org.apache.hadoop.hive.ql.exec.vector._
+import org.embulk.spi.{Column, ColumnVisitor, PageReader}
+
+class OrcColumnVisitor(val reader: PageReader, val batch: VectorizedRowBatch, val i: Integer) extends ColumnVisitor {
+  override def booleanColumn(column: Column): Unit = if (reader.isNull(column)) {
+    batch.cols(column.getIndex).noNulls = false
+    batch.cols(column.getIndex).isNull(i) = true
+  }
+  else if (reader.getBoolean(column)) batch.cols(column.getIndex).asInstanceOf[LongColumnVector].vector(i) = 1
+  else batch.cols(column.getIndex).asInstanceOf[LongColumnVector].vector(i) = 0
+
+  override def longColumn(column: Column): Unit = if (reader.isNull(column)) {
+    batch.cols(column.getIndex).noNulls = false
+    batch.cols(column.getIndex).isNull(i) = true
+  }
+  else batch.cols(column.getIndex).asInstanceOf[LongColumnVector].vector(i) = reader.getLong(column)
+
+  override def doubleColumn(column: Column): Unit = if (reader.isNull(column)) {
+    batch.cols(column.getIndex).noNulls = false
+    batch.cols(column.getIndex).isNull(i) = true
+  }
+  else batch.cols(column.getIndex).asInstanceOf[DoubleColumnVector].vector(i) = reader.getDouble(column)
+
+  override def stringColumn(column: Column): Unit = if (!reader.isNull(column)) batch.cols(column.getIndex).asInstanceOf[BytesColumnVector].setVal(i, reader.getString(column).getBytes(StandardCharsets.UTF_8))
+  else {
+    batch.cols(column.getIndex).noNulls = false
+    batch.cols(column.getIndex).isNull(i) = true
+  }
+
+  override def timestampColumn(column: Column): Unit = if (reader.isNull(column)) batch.cols(column.getIndex).asInstanceOf[TimestampColumnVector].setNullValue(i)
+  else {
+    val timestamp = reader.getTimestamp(column)
+    val ts = new java.sql.Timestamp(timestamp.getEpochSecond * 1000)
+    batch.cols(column.getIndex).asInstanceOf[TimestampColumnVector].set(i, ts)
+  }
+
+  override def jsonColumn(column: Column) = throw new UnsupportedOperationException("orc output plugin does not support json type")
+}

data/src/main/scala/org/embulk/output/orc/OrcOutputPlugin.scala

@@ -0,0 +1,156 @@
+package org.embulk.output.orc
+
+import java.io.IOException
+import java.util
+
+import org.apache.hadoop.conf.Configuration
+import org.apache.hadoop.fs.{LocalFileSystem, Path}
+import org.apache.hadoop.hdfs.DistributedFileSystem
+import org.apache.hadoop.util.VersionInfo
+import org.apache.orc.{CompressionKind, MemoryManager, OrcFile, TypeDescription, Writer}
+import org.embulk.config.{ConfigSource, TaskReport, TaskSource}
+import org.embulk.spi.util.Timestamps
+import org.embulk.spi.{Exec, OutputPlugin, PageReader, Schema}
+import org.embulk.util.aws.credentials.AwsCredentials
+
+object OrcOutputPlugin {
+  private[orc] def getSchema(schema: Schema) = {
+    val oschema = TypeDescription.createStruct
+    for (i <- 0 until schema.size) {
+      val column = schema.getColumn(i)
+      val `type` = column.getType
+      `type`.getName match {
+        case "long" =>
+          oschema.addField(column.getName, TypeDescription.createLong)
+        case "double" =>
+          oschema.addField(column.getName, TypeDescription.createDouble)
+        case "boolean" =>
+          oschema.addField(column.getName, TypeDescription.createBoolean)
+        case "string" =>
+          oschema.addField(column.getName, TypeDescription.createString)
+        case "timestamp" =>
+          oschema.addField(column.getName, TypeDescription.createTimestamp)
+        case _ =>
+          System.out.println("Unsupported type")
+      }
+    }
+    oschema
+  }
+
+  // We avoid using orc.MemoryManagerImpl since it is not threadsafe, but embulk is multi-threaded.
+  // Embulk creates and uses multiple instances of TransactionalPageOutput in worker threads.
+  // As a workaround, WriterLocalMemoryManager is bound to a single orc.Writer instance, and
+  // notifies checkMemory() only to that instance.
+  private class WriterLocalMemoryManager extends MemoryManager {
+    final private[orc] val rowsBetweenChecks = 10000
+    private var rowsAddedSinceCheck = 0
+    private[orc] var boundCallback: MemoryManager.Callback = _
+
+    @throws[IOException]
+    override def addWriter(path: Path, requestedAllocation: Long, callback: MemoryManager.Callback): Unit = {
+      if (boundCallback != null) {
+        throw new IllegalStateException("WriterLocalMemoryManager should be bound to a single orc.Writer instance.")
+      } else {
+        boundCallback = callback
+      }
+    }
+
+    @throws[IOException]
+    override def removeWriter(path: Path): Unit = boundCallback = null
+
+    @throws[IOException]
+    override def addedRow(rows: Int): Unit = {
+      rowsAddedSinceCheck += rows
+      if (rowsAddedSinceCheck > rowsBetweenChecks) {
+        boundCallback.checkMemory(1)
+        rowsAddedSinceCheck = 0
+      }
+    }
+  }
+
+}
+
+class OrcOutputPlugin extends OutputPlugin {
+  override def transaction(config: ConfigSource, schema: Schema, taskCount: Int, control: OutputPlugin.Control) = {
+    val task = config.loadConfig(classOf[PluginTask])
+    // retryable (idempotent) output:
+    // return resume(task.dump(), schema, taskCount, control);
+    // non-retryable (non-idempotent) output:
+    control.run(task.dump)
+    Exec.newConfigDiff
+  }
+
+  override def resume(taskSource: TaskSource, schema: Schema, taskCount: Int, control: OutputPlugin.Control) = throw new UnsupportedOperationException("orc output plugin does not support resuming")
+
+  override def cleanup(taskSource: TaskSource, schema: Schema, taskCount: Int, successTaskReports: util.List[TaskReport]): Unit = {
+  }
+
+  override def open(taskSource: TaskSource, schema: Schema, taskIndex: Int) = {
+    val task = taskSource.loadTask(classOf[PluginTask])
+    if (task.getOverwrite) {
+      val credentials = AwsCredentials.getAWSCredentialsProvider(task).getCredentials
+      OrcOutputPluginHelper.removeOldFile(buildPath(task, taskIndex), task)
+    }
+    val reader = new PageReader(schema)
+    val writer = createWriter(task, schema, taskIndex)
+    new OrcTransactionalPageOutput(reader, writer, task)
+  }
+
+  private def buildPath(task: PluginTask, processorIndex: Int): String = {
+    val pathPrefix = task.getPathPrefix
+    val pathSuffix = task.getFileNameExtension
+    val sequenceFormat = task.getSequenceFormat
+    val fmt = java.lang.String.format(sequenceFormat, processorIndex.asInstanceOf[AnyRef])
+    pathPrefix + fmt + pathSuffix
+  }
+
+  private def getHadoopConfiguration(task: PluginTask) = {
+    val conf = new Configuration
+    // see: https://stackoverflow.com/questions/17265002/hadoop-no-filesystem-for-scheme-file
+    conf.set("fs.hdfs.impl", classOf[DistributedFileSystem].getName)
+    conf.set("fs.file.impl", classOf[LocalFileSystem].getName)
+    // see: https://stackoverflow.com/questions/20833444/how-to-set-objects-in-hadoop-configuration
+    AwsCredentials.getAWSCredentialsProvider(task)
+    if (task.getAccessKeyId.isPresent) {
+      conf.set("fs.s3a.access.key", task.getAccessKeyId.get)
+      conf.set("fs.s3n.awsAccessKeyId", task.getAccessKeyId.get)
+    }
+    if (task.getSecretAccessKey.isPresent) {
+      conf.set("fs.s3a.secret.key", task.getSecretAccessKey.get)
+      conf.set("fs.s3n.awsSecretAccessKey", task.getSecretAccessKey.get)
+    }
+    if (task.getEndpoint.isPresent) {
+      conf.set("fs.s3a.endpoint", task.getEndpoint.get)
+      conf.set("fs.s3n.endpoint", task.getEndpoint.get)
+    }
+    conf
+  }
+
+  private def createWriter(task: PluginTask, schema: Schema, processorIndex: Int): Writer = {
+    val timestampFormatters = Timestamps.newTimestampColumnFormatters(task, schema, task.getColumnOptions)
+    val conf = getHadoopConfiguration(task)
+    val oschema = OrcOutputPlugin.getSchema(schema)
+    // see: https://groups.google.com/forum/#!topic/vertx/lLb-slzpWVg
+    Thread.currentThread.setContextClassLoader(classOf[VersionInfo].getClassLoader)
+
+    var writer: Writer = null
+    try { // Make writerOptions
+      val writerOptions = createWriterOptions(task, conf)
+      // see: https://stackoverflow.com/questions/9256733/how-to-connect-hive-in-ireport
+      // see: https://community.hortonworks.com/content/kbentry/73458/connecting-dbvisualizer-and-datagrip-to-hive-with.html
+      writer = OrcFile.createWriter(new Path(buildPath(task, processorIndex)), writerOptions.setSchema(oschema).memory(new OrcOutputPlugin.WriterLocalMemoryManager).version(OrcFile.Version.V_0_12))
+    } catch {
+      case e: IOException => throw e
+    }
+    writer
+  }
+
+  private def createWriterOptions(task: PluginTask, conf: Configuration) = {
+    val bufferSize = task.getBufferSize
+    val stripSize = task.getStripSize
+    val blockSize = task.getBlockSize
+    val kindString = task.getCompressionKind
+    val kind = CompressionKind.valueOf(kindString)
+    OrcFile.writerOptions(conf).bufferSize(bufferSize).blockSize(blockSize.toLong).stripeSize(stripSize.toLong).compress(kind)
+  }
+}

data/src/main/scala/org/embulk/output/orc/OrcOutputPluginHelper.scala

@@ -0,0 +1,57 @@
+package org.embulk.output.orc
+
+import java.io.IOException
+import java.nio.file.{Files, Paths}
+
+import com.amazonaws.auth.profile.ProfileCredentialsProvider
+import com.amazonaws.services.s3.AmazonS3Client
+import com.amazonaws.services.s3.model.DeleteObjectRequest
+
+import scala.beans.BeanProperty
+
+object OrcOutputPluginHelper {
+  def removeOldFile(fpath: String, task: PluginTask): Unit = {
+    // NOTE: Delete a file if local-filesystem, not HDFS or S3.
+    val schema = getSchema(fpath)
+    if (isDeleteTarget(schema)) schema match {
+      case "file" =>
+        try Files.deleteIfExists(Paths.get(fpath))
+        catch {
+          case e: IOException => throw e
+        }
+      case "s3" | "s3n" | "s3a" =>
+        val s3Url = parseS3Url(fpath)
+        val s3client = new AmazonS3Client(new ProfileCredentialsProvider)
+        if (task.getEndpoint.isPresent) s3client.setEndpoint(task.getEndpoint.get)
+        s3client.deleteObject(new DeleteObjectRequest(s3Url.bucket, s3Url.key))
+      case _ =>
+    }
+  }
+
+  def isDeleteTarget(schema: String): Boolean = schema match {
+    case "file" => true
+    case "s3" | "s3n" | "s3a" => true
+    case _ => false
+  }
+
+  def getSchema(fpath: String): String = {
+    val schema = fpath.split("://").toList.head
+    schema match {
+      case "s3" | "s3a" | "s3n" => schema
+      case _ => {
+        val path = Paths.get(fpath)
+        path.getFileSystem.provider.getScheme
+      }
+    }
+  }
+
+  def parseS3Url(s3url: String): AmazonS3URILikeObject = {
+    val parts = s3url.split("(://|/)").toList
+    val bucket = parts.apply(1)
+    val key = parts.slice(2, parts.size).mkString("/")
+    OrcOutputPluginHelper.AmazonS3URILikeObject(bucket, key)
+  }
+
+  case class AmazonS3URILikeObject(@BeanProperty bucket: String, @BeanProperty key: String)
+
+}

data/src/main/scala/org/embulk/output/orc/OrcTransactionalPageOutput.scala

@@ -0,0 +1,52 @@
+package org.embulk.output.orc
+
+import java.io.IOException
+
+import org.apache.orc.Writer
+import org.embulk.config.TaskReport
+import org.embulk.spi.{Exec, Page, PageReader, TransactionalPageOutput}
+
+class OrcTransactionalPageOutput(val reader: PageReader, val writer: Writer, val task: PluginTask) extends TransactionalPageOutput {
+  override def add(page: Page): Unit = synchronized {
+    try {
+      // int size = page.getStringReferences().size();
+      val schema = OrcOutputPlugin.getSchema(reader.getSchema)
+      val batch = schema.createRowBatch
+      // batch.size = size;
+      reader.setPage(page)
+      while ( {
+        reader.nextRecord
+      }) {
+        val row = {
+          batch.size += 1;
+          batch.size - 1
+        }
+        reader.getSchema.visitColumns(new OrcColumnVisitor(reader, batch, row))
+        if (batch.size >= batch.getMaxSize) {
+          writer.addRowBatch(batch)
+          batch.reset()
+        }
+      }
+      if (batch.size != 0) {
+        writer.addRowBatch(batch)
+        batch.reset()
+      }
+    } catch {
+      case e: IOException =>
+        e.printStackTrace()
+    }
+  }
+
+  override def finish(): Unit = {
+    try writer.close()
+    catch {
+      case e: IOException => throw e
+    }
+  }
+
+  override def close(): Unit = {}
+
+  override def abort(): Unit = {}
+
+  override def commit: TaskReport = Exec.newTaskReport
+}

data/src/main/scala/org/embulk/output/orc/PluginTask.scala

@@ -0,0 +1,56 @@
+package org.embulk.output.orc
+
+import java.util
+
+import com.google.common.base.Optional
+import org.embulk.config.{Config, ConfigDefault, Task}
+import org.embulk.spi.time.TimestampFormatter
+import org.embulk.util.aws.credentials.AwsCredentialsTask
+import org.joda.time.DateTimeZone
+
+trait PluginTask extends Task with TimestampFormatter.Task with AwsCredentialsTask {
+  @Config("path_prefix")
+  def getPathPrefix: String
+
+  @Config("file_ext")
+  @ConfigDefault("\".orc\"")
+  def getFileNameExtension: String
+
+  @Config("column_options")
+  @ConfigDefault("{}")
+  def getColumnOptions: util.Map[String, TimestampColumnOption]
+
+  @Config("sequence_format")
+  @ConfigDefault("\".%03d\"")
+  def getSequenceFormat: String
+
+  // see: https://orc.apache.org/docs/hive-config.html
+  // ORC File options
+  @Config("strip_size")
+  @ConfigDefault("67108864") // 64MB
+  def getStripSize: Integer
+
+  @Config("buffer_size")
+  @ConfigDefault("262144") // 256KB
+  def getBufferSize: Integer
+
+  @Config("block_size")
+  @ConfigDefault("268435456") // 256MB
+  def getBlockSize: Integer
+
+  @Config("compression_kind")
+  @ConfigDefault("ZLIB")
+  def getCompressionKind: String
+
+  @Config("overwrite")
+  @ConfigDefault("false")
+  def getOverwrite: Boolean
+
+  @Config("default_from_timezone")
+  @ConfigDefault("\"UTC\"")
+  def getDefaultFromTimeZone: DateTimeZone
+
+  @Config("endpoint")
+  @ConfigDefault("null")
+  def getEndpoint: Optional[String]
+}

data/src/main/scala/org/embulk/output/orc/TimestampColumnOption.scala

@@ -0,0 +1,32 @@
+package org.embulk.output.orc
+
+import java.util
+
+import com.google.common.base.Optional
+import org.embulk.config.{Config, ConfigDefault, Task}
+import org.embulk.spi.time.TimestampFormatter
+import org.joda.time.DateTimeZone
+
+/*
+public interface TimestampColumnOption
+  extends Task, TimestampFormatter.TimestampColumnOption
+{
+    @Config("from_timezone")
+    @ConfigDefault("null")
+    Optional<DateTimeZone> getFromTimeZone();
+
+    @Config("from_format")
+    @ConfigDefault("null")
+    Optional<List<String>> getFromFormat();
+}
+ */
+
+trait TimestampColumnOption extends Task with TimestampFormatter.TimestampColumnOption {
+  @Config("from_timezone")
+  @ConfigDefault("null")
+  def getFromTimeZone: Optional[DateTimeZone]
+
+  @Config("from_format")
+  @ConfigDefault("null")
+  def getFromFormat: Optional[util.List[String]]
+}

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: embulk-output-orc
 version: !ruby/object:Gem::Version
-  version: 0.3.4
+  version: 0.3.5
 platform: ruby
 authors:
 - yuokada
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2019-02-25 00:00:00.000000000 Z
+date: 2020-08-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   requirement: !ruby/object:Gem::Requirement
@@ -45,13 +45,12 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- ".github/workflows/gradle.yml"
 - ".gitignore"
-- ".travis.yml"
 - LICENSE.txt
 - README.md
 - build.gradle
 - classpath/aircompressor-0.10.jar
-- classpath/animal-sniffer-annotations-1.14.jar
 - classpath/apacheds-i18n-2.0.0-M15.jar
 - classpath/apacheds-kerberos-codec-2.0.0-M15.jar
 - classpath/api-asn1-api-1.0.0-M20.jar
@@ -62,7 +61,6 @@ files:
 - classpath/aws-java-sdk-core-1.10.33.jar
 - classpath/aws-java-sdk-kms-1.10.33.jar
 - classpath/aws-java-sdk-s3-1.10.33.jar
-- classpath/checker-compat-qual-2.0.0.jar
 - classpath/commons-beanutils-1.7.0.jar
 - classpath/commons-cli-1.2.jar
 - classpath/commons-codec-1.6.jar
@@ -80,11 +78,9 @@ files:
 - classpath/curator-client-2.7.1.jar
 - classpath/curator-framework-2.7.1.jar
 - classpath/curator-recipes-2.7.1.jar
-- classpath/embulk-output-orc-0.3.4.jar
+- classpath/embulk-output-orc-0.3.5.jar
 - classpath/embulk-util-aws-credentials-0.2.8.jar
-- classpath/error_prone_annotations-2.1.3.jar
 - classpath/gson-2.2.4.jar
-- classpath/guava-24.1-jre.jar
 - classpath/hadoop-annotations-2.7.5.jar
 - classpath/hadoop-auth-2.7.5.jar
 - classpath/hadoop-aws-2.7.5.jar
@@ -94,7 +90,6 @@ files:
 - classpath/htrace-core-3.1.0-incubating.jar
 - classpath/httpclient-4.3.6.jar
 - classpath/httpcore-4.3.3.jar
-- classpath/j2objc-annotations-1.1.jar
 - classpath/jackson-core-asl-1.9.13.jar
 - classpath/jackson-jaxrs-1.8.3.jar
 - classpath/jackson-mapper-asl-1.9.13.jar
@@ -123,6 +118,7 @@ files:
 - classpath/orc-shims-1.5.4.jar
 - classpath/paranamer-2.3.jar
 - classpath/protobuf-java-2.5.0.jar
+- classpath/scala-library-2.12.12.jar
 - classpath/servlet-api-2.5-20081211.jar
 - classpath/servlet-api-2.5.jar
 - classpath/snappy-java-1.0.4.1.jar
@@ -139,11 +135,12 @@ files:
 - gradlew
 - gradlew.bat
 - lib/embulk/output/orc.rb
-- src/main/java/org/embulk/output/orc/OrcColumnVisitor.java
-- src/main/java/org/embulk/output/orc/OrcOutputPlugin.java
-- src/main/java/org/embulk/output/orc/OrcOutputPluginHelper.java
-- src/main/java/org/embulk/output/orc/PluginTask.java
-- src/main/java/org/embulk/output/orc/TimestampColumnOption.java
+- src/main/scala/org/embulk/output/orc/OrcColumnVisitor.scala
+- src/main/scala/org/embulk/output/orc/OrcOutputPlugin.scala
+- src/main/scala/org/embulk/output/orc/OrcOutputPluginHelper.scala
+- src/main/scala/org/embulk/output/orc/OrcTransactionalPageOutput.scala
+- src/main/scala/org/embulk/output/orc/PluginTask.scala
+- src/main/scala/org/embulk/output/orc/TimestampColumnOption.scala
 - src/test/java/org/embulk/output/orc/OrcOutputPluginHelperTest.java
 - src/test/resources/example-null.yml
 - src/test/resources/example.yml

data/.travis.yml

@@ -1,14 +0,0 @@
-language: java
-jdk:
-  - oraclejdk8
-
-cache:
-  directories:  # run "travis cache --delete" to delete caches
-    - $HOME/.gradle
-
-sudo: false
-script:
-  - ./gradlew --info checkstyle
-  - ./gradlew --info check
-
-after_success: ./gradlew sonarqube -Dsonar.organization=yuokada-github -Dsonar.host.url=https://sonarcloud.io -Dsonar.login=${SONAR_TOKEN}

data/src/main/java/org/embulk/output/orc/OrcColumnVisitor.java

@@ -1,101 +0,0 @@
-package org.embulk.output.orc;
-
-import org.apache.hadoop.hive.ql.exec.vector.BytesColumnVector;
-import org.apache.hadoop.hive.ql.exec.vector.DoubleColumnVector;
-import org.apache.hadoop.hive.ql.exec.vector.LongColumnVector;
-import org.apache.hadoop.hive.ql.exec.vector.TimestampColumnVector;
-import org.apache.hadoop.hive.ql.exec.vector.VectorizedRowBatch;
-import org.embulk.spi.Column;
-import org.embulk.spi.ColumnVisitor;
-import org.embulk.spi.PageReader;
-import org.embulk.spi.time.Timestamp;
-
-import java.nio.charset.StandardCharsets;
-
-public class OrcColumnVisitor
-        implements ColumnVisitor
-{
-    private final PageReader reader;
-    private final VectorizedRowBatch batch;
-    private final Integer i;
-
-    public OrcColumnVisitor(PageReader pageReader, VectorizedRowBatch rowBatch, Integer i)
-    {
-        this.reader = pageReader;
-        this.batch = rowBatch;
-        this.i = i;
-    }
-
-    @Override
-    public void booleanColumn(Column column)
-    {
-        if (reader.isNull(column)) {
-            batch.cols[column.getIndex()].noNulls = false;
-            batch.cols[column.getIndex()].isNull[i] = true;
-        }
-        else {
-            if (reader.getBoolean(column)) {
-                ((LongColumnVector) batch.cols[column.getIndex()]).vector[i] = 1;
-            }
-            else {
-                ((LongColumnVector) batch.cols[column.getIndex()]).vector[i] = 0;
-            }
-        }
-    }
-
-    @Override
-    public void longColumn(Column column)
-    {
-        if (reader.isNull(column)) {
-            batch.cols[column.getIndex()].noNulls = false;
-            batch.cols[column.getIndex()].isNull[i] = true;
-        }
-        else {
-            ((LongColumnVector) batch.cols[column.getIndex()]).vector[i] = reader.getLong(column);
-        }
-    }
-
-    @Override
-    public void doubleColumn(Column column)
-    {
-        if (reader.isNull(column)) {
-            batch.cols[column.getIndex()].noNulls = false;
-            batch.cols[column.getIndex()].isNull[i] = true;
-        }
-        else {
-            ((DoubleColumnVector) batch.cols[column.getIndex()]).vector[i] = reader.getDouble(column);
-        }
-    }
-
-    @Override
-    public void stringColumn(Column column)
-    {
-        if (!reader.isNull(column)) {
-            ((BytesColumnVector) batch.cols[column.getIndex()])
-                    .setVal(i, reader.getString(column).getBytes(StandardCharsets.UTF_8));
-        }
-        else {
-            batch.cols[column.getIndex()].noNulls = false;
-            batch.cols[column.getIndex()].isNull[i] = true;
-        }
-    }
-
-    @Override
-    public void timestampColumn(Column column)
-    {
-        if (reader.isNull(column)) {
-            ((TimestampColumnVector) batch.cols[column.getIndex()]).setNullValue(i);
-        }
-        else {
-            Timestamp timestamp = reader.getTimestamp(column);
-            java.sql.Timestamp ts = new java.sql.Timestamp(timestamp.getEpochSecond() * 1000);
-            ((TimestampColumnVector) batch.cols[column.getIndex()]).set(i, ts);
-        }
-    }
-
-    @Override
-    public void jsonColumn(Column column)
-    {
-        throw new UnsupportedOperationException("orc output plugin does not support json type");
-    }
-}

data/src/main/java/org/embulk/output/orc/OrcOutputPlugin.java

@@ -1,298 +0,0 @@
-package org.embulk.output.orc;
-
-import com.amazonaws.auth.AWSCredentials;
-import com.google.common.base.Throwables;
-import org.apache.hadoop.conf.Configuration;
-import org.apache.hadoop.fs.LocalFileSystem;
-import org.apache.hadoop.fs.Path;
-import org.apache.hadoop.hdfs.DistributedFileSystem;
-import org.apache.hadoop.hive.ql.exec.vector.VectorizedRowBatch;
-import org.apache.hadoop.util.VersionInfo;
-import org.apache.orc.CompressionKind;
-import org.apache.orc.MemoryManager;
-import org.apache.orc.OrcFile;
-import org.apache.orc.TypeDescription;
-import org.apache.orc.Writer;
-import org.embulk.config.ConfigDiff;
-import org.embulk.config.ConfigSource;
-import org.embulk.config.TaskReport;
-import org.embulk.config.TaskSource;
-import org.embulk.spi.Column;
-import org.embulk.spi.Exec;
-import org.embulk.spi.OutputPlugin;
-import org.embulk.spi.Page;
-import org.embulk.spi.PageReader;
-import org.embulk.spi.Schema;
-import org.embulk.spi.TransactionalPageOutput;
-import org.embulk.spi.time.TimestampFormatter;
-import org.embulk.spi.type.Type;
-import org.embulk.spi.util.Timestamps;
-import org.embulk.util.aws.credentials.AwsCredentials;
-
-import java.io.IOException;
-import java.util.List;
-
-public class OrcOutputPlugin
-        implements OutputPlugin
-{
-    @Override
-    public ConfigDiff transaction(ConfigSource config,
-            Schema schema, int taskCount,
-            OutputPlugin.Control control)
-    {
-        PluginTask task = config.loadConfig(PluginTask.class);
-
-        // retryable (idempotent) output:
-        // return resume(task.dump(), schema, taskCount, control);
-
-        // non-retryable (non-idempotent) output:
-        control.run(task.dump());
-        return Exec.newConfigDiff();
-    }
-
-    @Override
-    public ConfigDiff resume(TaskSource taskSource,
-            Schema schema, int taskCount,
-            OutputPlugin.Control control)
-    {
-        throw new UnsupportedOperationException("orc output plugin does not support resuming");
-    }
-
-    @Override
-    public void cleanup(TaskSource taskSource,
-            Schema schema, int taskCount,
-            List<TaskReport> successTaskReports)
-    {
-    }
-
-    @Override
-    public TransactionalPageOutput open(TaskSource taskSource, Schema schema, int taskIndex)
-    {
-        PluginTask task = taskSource.loadTask(PluginTask.class);
-
-        if (task.getOverwrite()) {
-            AWSCredentials credentials = AwsCredentials.getAWSCredentialsProvider(task).getCredentials();
-            OrcOutputPluginHelper.removeOldFile(buildPath(task, taskIndex), task);
-        }
-
-        final PageReader reader = new PageReader(schema);
-        Writer writer = createWriter(task, schema, taskIndex);
-
-        return new OrcTransactionalPageOutput(reader, writer, task);
-    }
-
-    private String buildPath(PluginTask task, int processorIndex)
-    {
-        final String pathPrefix = task.getPathPrefix();
-        final String pathSuffix = task.getFileNameExtension();
-        final String sequenceFormat = task.getSequenceFormat();
-        return pathPrefix + String.format(sequenceFormat, processorIndex) + pathSuffix;
-    }
-
-    private TypeDescription getSchema(Schema schema)
-    {
-        TypeDescription oschema = TypeDescription.createStruct();
-        for (int i = 0; i < schema.size(); i++) {
-            Column column = schema.getColumn(i);
-            Type type = column.getType();
-            switch (type.getName()) {
-                case "long":
-                    oschema.addField(column.getName(), TypeDescription.createLong());
-                    break;
-                case "double":
-                    oschema.addField(column.getName(), TypeDescription.createDouble());
-                    break;
-                case "boolean":
-                    oschema.addField(column.getName(), TypeDescription.createBoolean());
-                    break;
-                case "string":
-                    oschema.addField(column.getName(), TypeDescription.createString());
-                    break;
-                case "timestamp":
-                    oschema.addField(column.getName(), TypeDescription.createTimestamp());
-                    break;
-                default:
-                    System.out.println("Unsupported type");
-                    break;
-            }
-        }
-        return oschema;
-    }
-
-    private Configuration getHadoopConfiguration(PluginTask task)
-    {
-        Configuration conf = new Configuration();
-
-        // see: https://stackoverflow.com/questions/17265002/hadoop-no-filesystem-for-scheme-file
-        conf.set("fs.hdfs.impl", DistributedFileSystem.class.getName());
-        conf.set("fs.file.impl", LocalFileSystem.class.getName());
-        // see: https://stackoverflow.com/questions/20833444/how-to-set-objects-in-hadoop-configuration
-
-        AwsCredentials.getAWSCredentialsProvider(task);
-        if (task.getAccessKeyId().isPresent()) {
-            conf.set("fs.s3a.access.key", task.getAccessKeyId().get());
-            conf.set("fs.s3n.awsAccessKeyId", task.getAccessKeyId().get());
-        }
-        if (task.getSecretAccessKey().isPresent()) {
-            conf.set("fs.s3a.secret.key", task.getSecretAccessKey().get());
-            conf.set("fs.s3n.awsSecretAccessKey", task.getSecretAccessKey().get());
-        }
-        if (task.getEndpoint().isPresent()) {
-            conf.set("fs.s3a.endpoint", task.getEndpoint().get());
-            conf.set("fs.s3n.endpoint", task.getEndpoint().get());
-        }
-        return conf;
-    }
-
-    private Writer createWriter(PluginTask task, Schema schema, int processorIndex)
-    {
-        final TimestampFormatter[] timestampFormatters = Timestamps
-                .newTimestampColumnFormatters(task, schema, task.getColumnOptions());
-
-        Configuration conf = getHadoopConfiguration(task);
-        TypeDescription oschema = getSchema(schema);
-
-        // see: https://groups.google.com/forum/#!topic/vertx/lLb-slzpWVg
-        Thread.currentThread().setContextClassLoader(VersionInfo.class.getClassLoader());
-
-        Writer writer = null;
-        try {
-            // Make writerOptions
-            OrcFile.WriterOptions writerOptions = createWriterOptions(task, conf);
-            // see: https://stackoverflow.com/questions/9256733/how-to-connect-hive-in-ireport
-            // see: https://community.hortonworks.com/content/kbentry/73458/connecting-dbvisualizer-and-datagrip-to-hive-with.html
-            writer = OrcFile.createWriter(
-                    new Path(buildPath(task, processorIndex)),
-                    writerOptions.setSchema(oschema)
-                            .memory(new WriterLocalMemoryManager())
-                            .version(OrcFile.Version.V_0_12)
-            );
-        }
-        catch (IOException e) {
-            Throwables.propagate(e);
-        }
-        return writer;
-    }
-
-    private OrcFile.WriterOptions createWriterOptions(PluginTask task, Configuration conf)
-    {
-        final Integer bufferSize = task.getBufferSize();
-        final Integer stripSize = task.getStripSize();
-        final Integer blockSize = task.getBlockSize();
-        final String kindString = task.getCompressionKind();
-        CompressionKind kind = CompressionKind.valueOf(kindString);
-        return OrcFile.writerOptions(conf)
-                .bufferSize(bufferSize)
-                .blockSize(blockSize)
-                .stripeSize(stripSize)
-                .compress(kind);
-    }
-
-    class OrcTransactionalPageOutput
-            implements TransactionalPageOutput
-    {
-        private final PageReader reader;
-        private final Writer writer;
-
-        public OrcTransactionalPageOutput(PageReader reader, Writer writer, PluginTask task)
-        {
-            this.reader = reader;
-            this.writer = writer;
-        }
-
-        @Override
-        public void add(Page page)
-        {
-            try {
-                // int size = page.getStringReferences().size();
-                final TypeDescription schema = getSchema(reader.getSchema());
-                final VectorizedRowBatch batch = schema.createRowBatch();
-                // batch.size = size;
-
-                reader.setPage(page);
-                while (reader.nextRecord()) {
-                    final int row = batch.size++;
-                    reader.getSchema().visitColumns(
-                            new OrcColumnVisitor(reader, batch, row)
-                    );
-                    if (batch.size >= batch.getMaxSize()) {
-                        writer.addRowBatch(batch);
-                        batch.reset();
-                    }
-                }
-                if (batch.size != 0) {
-                    writer.addRowBatch(batch);
-                    batch.reset();
-                }
-            }
-            catch (IOException e) {
-                e.printStackTrace();
-            }
-        }
-
-        @Override
-        public void finish()
-        {
-            try {
-                writer.close();
-            }
-            catch (IOException e) {
-                Throwables.propagate(e);
-            }
-        }
-
-        @Override
-        public void close()
-        {
-        }
-
-        @Override
-        public void abort()
-        {
-        }
-
-        @Override
-        public TaskReport commit()
-        {
-            return Exec.newTaskReport();
-        }
-    }
-
-    // We avoid using orc.MemoryManagerImpl since it is not threadsafe, but embulk is multi-threaded.
-    // Embulk creates and uses multiple instances of TransactionalPageOutput in worker threads.
-    // As a workaround, WriterLocalMemoryManager is bound to a single orc.Writer instance, and
-    // notifies checkMemory() only to that instance.
-    private static class WriterLocalMemoryManager implements MemoryManager
-    {
-        final long rowsBetweenChecks = 10000;
-
-        private int rowsAddedSinceCheck = 0;
-        Callback boundCallback = null;
-
-        @Override
-        public void addWriter(Path path, long requestedAllocation, Callback callback) throws IOException
-        {
-            if (boundCallback != null) {
-                throw new IllegalStateException("WriterLocalMemoryManager should be bound to a single orc.Writer instance.");
-            }
-
-            boundCallback = callback;
-        }
-
-        @Override
-        public void removeWriter(Path path) throws IOException
-        {
-            boundCallback = null;
-        }
-
-        @Override
-        public void addedRow(int rows) throws IOException
-        {
-            rowsAddedSinceCheck += rows;
-            if (rowsAddedSinceCheck > rowsBetweenChecks) {
-                boundCallback.checkMemory(1);
-                rowsAddedSinceCheck = 0;
-            }
-        }
-    }
-}

data/src/main/java/org/embulk/output/orc/OrcOutputPluginHelper.java

@@ -1,111 +0,0 @@
-package org.embulk.output.orc;
-
-import com.amazonaws.auth.profile.ProfileCredentialsProvider;
-import com.amazonaws.services.s3.AmazonS3;
-import com.amazonaws.services.s3.AmazonS3Client;
-import com.amazonaws.services.s3.model.DeleteObjectRequest;
-import com.google.common.base.Joiner;
-import com.google.common.base.Splitter;
-import com.google.common.base.Throwables;
-
-import java.io.IOException;
-import java.nio.file.Files;
-import java.nio.file.Path;
-import java.nio.file.Paths;
-import java.util.Arrays;
-import java.util.List;
-
-class OrcOutputPluginHelper
-{
-    protected OrcOutputPluginHelper()
-    {
-        throw new UnsupportedOperationException();
-    }
-
-    static void removeOldFile(String fpath, PluginTask task)
-    {
-        // NOTE: Delete a file if local-filesystem, not HDFS or S3.
-        String schema = getSchema(fpath);
-        if (isDeleteTarget(schema)) {
-            switch (schema) {
-                case "file":
-                    try {
-                        Files.deleteIfExists(Paths.get(fpath));
-                    }
-                    catch (IOException e) {
-                        Throwables.propagate(e);
-                    }
-                    break;
-                case "s3":
-                case "s3n":
-                case "s3a":
-                    AmazonS3URILikeObject s3Url = parseS3Url(fpath);
-                    AmazonS3 s3client = new AmazonS3Client(new ProfileCredentialsProvider());
-                    if (task.getEndpoint().isPresent()) {
-                        s3client.setEndpoint(task.getEndpoint().get());
-                    }
-                    s3client.deleteObject(new DeleteObjectRequest(s3Url.getBucket(), s3Url.getKey()));
-                default:
-                    // TODO: Unsupported
-            }
-        }
-    }
-
-    public static boolean isDeleteTarget(String schema)
-    {
-        switch (schema) {
-            case "file":
-                return true;
-            case "s3":
-            case "s3a":
-            case "s3n":
-                return true;
-            default:
-                return false;
-        }
-    }
-
-    static String getSchema(String fpath)
-    {
-        String schema = Splitter.on("://")
-                .splitToList(fpath).get(0);
-        if (schema.equals("s3a") || schema.equals("s3n") || schema.equals("s3")) {
-            return schema;
-        }
-        else {
-            Path path = Paths.get(fpath);
-            return path.getFileSystem().provider().getScheme();
-        }
-    }
-
-    static AmazonS3URILikeObject parseS3Url(String s3url)
-    {
-        List<String> parts = Arrays.asList(
-                s3url.split("(://|/)"));
-        String bucket = parts.get(1);
-        String key = Joiner.on("/").join(parts.subList(2, parts.size()));
-        return new AmazonS3URILikeObject(bucket, key);
-    }
-
-    static class AmazonS3URILikeObject
-    {
-        String bucket;
-        String key;
-
-        public AmazonS3URILikeObject(String bucket, String key)
-        {
-            this.bucket = bucket;
-            this.key = key;
-        }
-
-        public String getBucket()
-        {
-            return bucket;
-        }
-
-        public String getKey()
-        {
-            return key;
-        }
-    }
-}

data/src/main/java/org/embulk/output/orc/PluginTask.java

@@ -1,60 +0,0 @@
-package org.embulk.output.orc;
-
-import com.google.common.base.Optional;
-import org.embulk.config.Config;
-import org.embulk.config.ConfigDefault;
-import org.embulk.config.Task;
-import org.embulk.spi.time.TimestampFormatter;
-import org.embulk.util.aws.credentials.AwsCredentialsTask;
-import org.joda.time.DateTimeZone;
-
-import java.util.Map;
-
-public interface PluginTask
-        extends Task, TimestampFormatter.Task, AwsCredentialsTask
-{
-    @Config("path_prefix")
-    String getPathPrefix();
-
-    @Config("file_ext")
-    @ConfigDefault("\".orc\"")
-    String getFileNameExtension();
-
-    @Config("column_options")
-    @ConfigDefault("{}")
-    Map<String, TimestampColumnOption> getColumnOptions();
-
-    @Config("sequence_format")
-    @ConfigDefault("\".%03d\"")
-    String getSequenceFormat();
-
-    // see: https://orc.apache.org/docs/hive-config.html
-    // ORC File options
-    @Config("strip_size")
-    @ConfigDefault("67108864") // 64MB
-    Integer getStripSize();
-
-    @Config("buffer_size")
-    @ConfigDefault("262144") // 256KB
-    Integer getBufferSize();
-
-    @Config("block_size")
-    @ConfigDefault("268435456") // 256MB
-    Integer getBlockSize();
-
-    @Config("compression_kind")
-    @ConfigDefault("ZLIB")
-    public String getCompressionKind();
-
-    @Config("overwrite")
-    @ConfigDefault("false")
-    boolean getOverwrite();
-
-    @Config("default_from_timezone")
-    @ConfigDefault("\"UTC\"")
-    DateTimeZone getDefaultFromTimeZone();
-
-    @Config("endpoint")
-    @ConfigDefault("null")
-    Optional<String> getEndpoint();
-}

data/src/main/java/org/embulk/output/orc/TimestampColumnOption.java

@@ -1,22 +0,0 @@
-package org.embulk.output.orc;
-
-import com.google.common.base.Optional;
-import org.embulk.config.Config;
-import org.embulk.config.ConfigDefault;
-import org.embulk.config.Task;
-import org.embulk.spi.time.TimestampFormatter;
-import org.joda.time.DateTimeZone;
-
-import java.util.List;
-
-public interface TimestampColumnOption
-        extends Task, TimestampFormatter.TimestampColumnOption
-{
-    @Config("from_timezone")
-    @ConfigDefault("null")
-    Optional<DateTimeZone> getFromTimeZone();
-
-    @Config("from_format")
-    @ConfigDefault("null")
-    Optional<List<String>> getFromFormat();
-}