spark DataStreamWriter 源码

spark DataStreamWriter 代码

文件路径：/sql/core/src/main/scala/org/apache/spark/sql/streaming/DataStreamWriter.scala

/*
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You 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 org.apache.spark.sql.streaming

import java.util.Locale
import java.util.concurrent.TimeoutException

import scala.collection.JavaConverters._

import org.apache.hadoop.fs.Path

import org.apache.spark.annotation.Evolving
import org.apache.spark.api.java.function.VoidFunction2
import org.apache.spark.sql._
import org.apache.spark.sql.catalyst.analysis.UnresolvedIdentifier
import org.apache.spark.sql.catalyst.catalog.{CatalogTable, CatalogTableType}
import org.apache.spark.sql.catalyst.plans.logical.{CreateTable, TableSpec}
import org.apache.spark.sql.catalyst.streaming.InternalOutputModes
import org.apache.spark.sql.catalyst.util.CaseInsensitiveMap
import org.apache.spark.sql.connector.catalog.{Identifier, SupportsWrite, Table, TableCatalog, TableProvider, V1Table, V2TableWithV1Fallback}
import org.apache.spark.sql.connector.catalog.TableCapability._
import org.apache.spark.sql.errors.QueryCompilationErrors
import org.apache.spark.sql.execution.command.DDLUtils
import org.apache.spark.sql.execution.datasources.DataSource
import org.apache.spark.sql.execution.datasources.v2.{DataSourceV2Utils, FileDataSourceV2}
import org.apache.spark.sql.execution.streaming._
import org.apache.spark.sql.execution.streaming.sources._
import org.apache.spark.sql.util.CaseInsensitiveStringMap
import org.apache.spark.util.Utils

/**
 * Interface used to write a streaming `Dataset` to external storage systems (e.g. file systems,
 * key-value stores, etc). Use `Dataset.writeStream` to access this.
 *
 * @since 2.0.0
 */
@Evolving
final class DataStreamWriter[T] private[sql](ds: Dataset[T]) {
  import DataStreamWriter._

  private val df = ds.toDF()

  /**
   * Specifies how data of a streaming DataFrame/Dataset is written to a streaming sink.
   * <ul>
   * <li> `OutputMode.Append()`: only the new rows in the streaming DataFrame/Dataset will be
   * written to the sink.</li>
   * <li> `OutputMode.Complete()`: all the rows in the streaming DataFrame/Dataset will be written
   * to the sink every time there are some updates.</li>
   * <li> `OutputMode.Update()`: only the rows that were updated in the streaming
   * DataFrame/Dataset will be written to the sink every time there are some updates.
   * If the query doesn't contain aggregations, it will be equivalent to
   * `OutputMode.Append()` mode.</li>
   * </ul>
   *
   * @since 2.0.0
   */
  def outputMode(outputMode: OutputMode): DataStreamWriter[T] = {
    this.outputMode = outputMode
    this
  }

  /**
   * Specifies how data of a streaming DataFrame/Dataset is written to a streaming sink.
   * <ul>
   * <li> `append`: only the new rows in the streaming DataFrame/Dataset will be written to
   * the sink.</li>
   * <li> `complete`: all the rows in the streaming DataFrame/Dataset will be written to the sink
   * every time there are some updates.</li>
   * <li> `update`: only the rows that were updated in the streaming DataFrame/Dataset will
   * be written to the sink every time there are some updates. If the query doesn't
   * contain aggregations, it will be equivalent to `append` mode.</li>
   * </ul>
   *
   * @since 2.0.0
   */
  def outputMode(outputMode: String): DataStreamWriter[T] = {
    this.outputMode = InternalOutputModes(outputMode)
    this
  }

  /**
   * Set the trigger for the stream query. The default value is `ProcessingTime(0)` and it will run
   * the query as fast as possible.
   *
   * Scala Example:
   * {{{
   *   df.writeStream.trigger(ProcessingTime("10 seconds"))
   *
   *   import scala.concurrent.duration._
   *   df.writeStream.trigger(ProcessingTime(10.seconds))
   * }}}
   *
   * Java Example:
   * {{{
   *   df.writeStream().trigger(ProcessingTime.create("10 seconds"))
   *
   *   import java.util.concurrent.TimeUnit
   *   df.writeStream().trigger(ProcessingTime.create(10, TimeUnit.SECONDS))
   * }}}
   *
   * @since 2.0.0
   */
  def trigger(trigger: Trigger): DataStreamWriter[T] = {
    this.trigger = trigger
    this
  }

  /**
   * Specifies the name of the [[StreamingQuery]] that can be started with `start()`.
   * This name must be unique among all the currently active queries in the associated SQLContext.
   *
   * @since 2.0.0
   */
  def queryName(queryName: String): DataStreamWriter[T] = {
    this.extraOptions += ("queryName" -> queryName)
    this
  }

  /**
   * Specifies the underlying output data source.
   *
   * @since 2.0.0
   */
  def format(source: String): DataStreamWriter[T] = {
    this.source = source
    this
  }

  /**
   * Partitions the output by the given columns on the file system. If specified, the output is
   * laid out on the file system similar to Hive's partitioning scheme. As an example, when we
   * partition a dataset by year and then month, the directory layout would look like:
   *
   * <ul>
   * <li> year=2016/month=01/</li>
   * <li> year=2016/month=02/</li>
   * </ul>
   *
   * Partitioning is one of the most widely used techniques to optimize physical data layout.
   * It provides a coarse-grained index for skipping unnecessary data reads when queries have
   * predicates on the partitioned columns. In order for partitioning to work well, the number
   * of distinct values in each column should typically be less than tens of thousands.
   *
   * @since 2.0.0
   */
  @scala.annotation.varargs
  def partitionBy(colNames: String*): DataStreamWriter[T] = {
    this.partitioningColumns = Option(colNames)
    this
  }

  /**
   * Adds an output option for the underlying data source.
   *
   * @since 2.0.0
   */
  def option(key: String, value: String): DataStreamWriter[T] = {
    this.extraOptions += (key -> value)
    this
  }

  /**
   * Adds an output option for the underlying data source.
   *
   * @since 2.0.0
   */
  def option(key: String, value: Boolean): DataStreamWriter[T] = option(key, value.toString)

  /**
   * Adds an output option for the underlying data source.
   *
   * @since 2.0.0
   */
  def option(key: String, value: Long): DataStreamWriter[T] = option(key, value.toString)

  /**
   * Adds an output option for the underlying data source.
   *
   * @since 2.0.0
   */
  def option(key: String, value: Double): DataStreamWriter[T] = option(key, value.toString)

  /**
   * (Scala-specific) Adds output options for the underlying data source.
   *
   * @since 2.0.0
   */
  def options(options: scala.collection.Map[String, String]): DataStreamWriter[T] = {
    this.extraOptions ++= options
    this
  }

  /**
   * Adds output options for the underlying data source.
   *
   * @since 2.0.0
   */
  def options(options: java.util.Map[String, String]): DataStreamWriter[T] = {
    this.options(options.asScala)
    this
  }

  /**
   * Starts the execution of the streaming query, which will continually output results to the given
   * path as new data arrives. The returned [[StreamingQuery]] object can be used to interact with
   * the stream.
   *
   * @since 2.0.0
   */
  def start(path: String): StreamingQuery = {
    if (!df.sparkSession.sessionState.conf.legacyPathOptionBehavior &&
        extraOptions.contains("path")) {
      throw QueryCompilationErrors.setPathOptionAndCallWithPathParameterError("start")
    }
    startInternal(Some(path))
  }

  /**
   * Starts the execution of the streaming query, which will continually output results to the given
   * path as new data arrives. The returned [[StreamingQuery]] object can be used to interact with
   * the stream. Throws a `TimeoutException` if the following conditions are met:
   *  - Another run of the same streaming query, that is a streaming query
   *    sharing the same checkpoint location, is already active on the same
   *    Spark Driver
   *  - The SQL configuration `spark.sql.streaming.stopActiveRunOnRestart`
   *    is enabled
   *  - The active run cannot be stopped within the timeout controlled by
   *    the SQL configuration `spark.sql.streaming.stopTimeout`
   *
   * @since 2.0.0
   */
  @throws[TimeoutException]
  def start(): StreamingQuery = startInternal(None)

  /**
   * Starts the execution of the streaming query, which will continually output results to the given
   * table as new data arrives. The returned [[StreamingQuery]] object can be used to interact with
   * the stream.
   *
   * For v1 table, partitioning columns provided by `partitionBy` will be respected no matter the
   * table exists or not. A new table will be created if the table not exists.
   *
   * For v2 table, `partitionBy` will be ignored if the table already exists. `partitionBy` will be
   * respected only if the v2 table does not exist. Besides, the v2 table created by this API lacks
   * some functionalities (e.g., customized properties, options, and serde info). If you need them,
   * please create the v2 table manually before the execution to avoid creating a table with
   * incomplete information.
   *
   * @since 3.1.0
   */
  @Evolving
  @throws[TimeoutException]
  def toTable(tableName: String): StreamingQuery = {
    this.tableName = tableName

    import df.sparkSession.sessionState.analyzer.CatalogAndIdentifier

    import org.apache.spark.sql.connector.catalog.CatalogV2Implicits._
    val originalMultipartIdentifier = df.sparkSession.sessionState.sqlParser
      .parseMultipartIdentifier(tableName)
    val CatalogAndIdentifier(catalog, identifier) = originalMultipartIdentifier

    // Currently we don't create a logical streaming writer node in logical plan, so cannot rely
    // on analyzer to resolve it. Directly lookup only for temp view to provide clearer message.
    // TODO (SPARK-27484): we should add the writing node before the plan is analyzed.
    if (df.sparkSession.sessionState.catalog.isTempView(originalMultipartIdentifier)) {
      throw QueryCompilationErrors.tempViewNotSupportStreamingWriteError(tableName)
    }

    if (!catalog.asTableCatalog.tableExists(identifier)) {
      import org.apache.spark.sql.connector.catalog.CatalogV2Implicits._
      /**
       * Note, currently the new table creation by this API doesn't fully cover the V2 table.
       * TODO (SPARK-33638): Full support of v2 table creation
       */
      val tableSpec = TableSpec(
        Map.empty[String, String],
        Some(source),
        Map.empty[String, String],
        extraOptions.get("path"),
        None,
        None,
        false)
      val cmd = CreateTable(
        UnresolvedIdentifier(originalMultipartIdentifier),
        df.schema.asNullable,
        partitioningColumns.getOrElse(Nil).asTransforms.toSeq,
        tableSpec,
        ignoreIfExists = false)
      Dataset.ofRows(df.sparkSession, cmd)
    }

    val tableInstance = catalog.asTableCatalog.loadTable(identifier)

    def writeToV1Table(table: CatalogTable): StreamingQuery = {
      if (table.tableType == CatalogTableType.VIEW) {
        throw QueryCompilationErrors.streamingIntoViewNotSupportedError(tableName)
      }
      require(table.provider.isDefined)
      if (source != table.provider.get) {
        throw QueryCompilationErrors.inputSourceDiffersFromDataSourceProviderError(
          source, tableName, table)
      }
      format(table.provider.get).startInternal(
        Some(new Path(table.location).toString), catalogTable = Some(table))
    }

    import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Implicits._
    tableInstance match {
      case t: SupportsWrite if t.supports(STREAMING_WRITE) =>
        startQuery(t, extraOptions, catalogAndIdent = Some(catalog.asTableCatalog, identifier))
      case t: V2TableWithV1Fallback =>
        writeToV1Table(t.v1Table)
      case t: V1Table =>
        writeToV1Table(t.v1Table)
      case t => throw QueryCompilationErrors.tableNotSupportStreamingWriteError(tableName, t)
    }
  }

  private def startInternal(
      path: Option[String],
      catalogTable: Option[CatalogTable] = None): StreamingQuery = {
    if (source.toLowerCase(Locale.ROOT) == DDLUtils.HIVE_PROVIDER) {
      throw QueryCompilationErrors.cannotOperateOnHiveDataSourceFilesError("write")
    }

    if (source == SOURCE_NAME_MEMORY) {
      assertNotPartitioned(SOURCE_NAME_MEMORY)
      if (extraOptions.get("queryName").isEmpty) {
        throw QueryCompilationErrors.queryNameNotSpecifiedForMemorySinkError()
      }
      val sink = new MemorySink()
      val resultDf = Dataset.ofRows(df.sparkSession, new MemoryPlan(sink, df.schema.toAttributes))
      val recoverFromCheckpoint = outputMode == OutputMode.Complete()
      val query = startQuery(sink, extraOptions, recoverFromCheckpoint = recoverFromCheckpoint,
        catalogTable = catalogTable)
      resultDf.createOrReplaceTempView(query.name)
      query
    } else if (source == SOURCE_NAME_FOREACH) {
      assertNotPartitioned(SOURCE_NAME_FOREACH)
      val sink = ForeachWriterTable[T](foreachWriter, ds.exprEnc)
      startQuery(sink, extraOptions, catalogTable = catalogTable)
    } else if (source == SOURCE_NAME_FOREACH_BATCH) {
      assertNotPartitioned(SOURCE_NAME_FOREACH_BATCH)
      if (trigger.isInstanceOf[ContinuousTrigger]) {
        throw QueryCompilationErrors.sourceNotSupportedWithContinuousTriggerError(source)
      }
      val sink = new ForeachBatchSink[T](foreachBatchWriter, ds.exprEnc)
      startQuery(sink, extraOptions, catalogTable = catalogTable)
    } else {
      val cls = DataSource.lookupDataSource(source, df.sparkSession.sessionState.conf)
      val disabledSources =
        Utils.stringToSeq(df.sparkSession.sqlContext.conf.disabledV2StreamingWriters)
      val useV1Source = disabledSources.contains(cls.getCanonicalName) ||
        // file source v2 does not support streaming yet.
        classOf[FileDataSourceV2].isAssignableFrom(cls)

      val optionsWithPath = if (path.isEmpty) {
        extraOptions
      } else {
        extraOptions + ("path" -> path.get)
      }

      val sink = if (classOf[TableProvider].isAssignableFrom(cls) && !useV1Source) {
        val provider = cls.getConstructor().newInstance().asInstanceOf[TableProvider]
        val sessionOptions = DataSourceV2Utils.extractSessionConfigs(
          source = provider, conf = df.sparkSession.sessionState.conf)
        val finalOptions = sessionOptions.filterKeys(!optionsWithPath.contains(_)).toMap ++
          optionsWithPath.originalMap
        val dsOptions = new CaseInsensitiveStringMap(finalOptions.asJava)
        // If the source accepts external table metadata, here we pass the schema of input query
        // to `getTable`. This is for avoiding schema inference, which can be very expensive.
        // If the query schema is not compatible with the existing data, the behavior is undefined.
        val outputSchema = if (provider.supportsExternalMetadata()) {
          Some(df.schema)
        } else {
          None
        }
        val table = DataSourceV2Utils.getTableFromProvider(
          provider, dsOptions, userSpecifiedSchema = outputSchema)
        import org.apache.spark.sql.execution.datasources.v2.DataSourceV2Implicits._
        table match {
          case table: SupportsWrite if table.supports(STREAMING_WRITE) =>
            table
          case _ => createV1Sink(optionsWithPath)
        }
      } else {
        createV1Sink(optionsWithPath)
      }

      startQuery(sink, optionsWithPath, catalogTable = catalogTable)
    }
  }

  private def startQuery(
      sink: Table,
      newOptions: CaseInsensitiveMap[String],
      recoverFromCheckpoint: Boolean = true,
      catalogAndIdent: Option[(TableCatalog, Identifier)] = None,
      catalogTable: Option[CatalogTable] = None): StreamingQuery = {
    val useTempCheckpointLocation = SOURCES_ALLOW_ONE_TIME_QUERY.contains(source)

    df.sparkSession.sessionState.streamingQueryManager.startQuery(
      newOptions.get("queryName"),
      newOptions.get("checkpointLocation"),
      df,
      newOptions.originalMap,
      sink,
      outputMode,
      useTempCheckpointLocation = useTempCheckpointLocation,
      recoverFromCheckpointLocation = recoverFromCheckpoint,
      trigger = trigger,
      catalogAndIdent = catalogAndIdent,
      catalogTable = catalogTable)
  }

  private def createV1Sink(optionsWithPath: CaseInsensitiveMap[String]): Sink = {
    val ds = DataSource(
      df.sparkSession,
      className = source,
      options = optionsWithPath.originalMap,
      partitionColumns = normalizedParCols.getOrElse(Nil))
    ds.createSink(outputMode)
  }

  /**
   * Sets the output of the streaming query to be processed using the provided writer object.
   * object. See [[org.apache.spark.sql.ForeachWriter]] for more details on the lifecycle and
   * semantics.
   * @since 2.0.0
   */
  def foreach(writer: ForeachWriter[T]): DataStreamWriter[T] = {
    this.source = SOURCE_NAME_FOREACH
    this.foreachWriter = if (writer != null) {
      ds.sparkSession.sparkContext.clean(writer)
    } else {
      throw new IllegalArgumentException("foreach writer cannot be null")
    }
    this
  }

  /**
   * :: Experimental ::
   *
   * (Scala-specific) Sets the output of the streaming query to be processed using the provided
   * function. This is supported only in the micro-batch execution modes (that is, when the
   * trigger is not continuous). In every micro-batch, the provided function will be called in
   * every micro-batch with (i) the output rows as a Dataset and (ii) the batch identifier.
   * The batchId can be used to deduplicate and transactionally write the output
   * (that is, the provided Dataset) to external systems. The output Dataset is guaranteed
   * to be exactly the same for the same batchId (assuming all operations are deterministic
   * in the query).
   *
   * @since 2.4.0
   */
  @Evolving
  def foreachBatch(function: (Dataset[T], Long) => Unit): DataStreamWriter[T] = {
    this.source = SOURCE_NAME_FOREACH_BATCH
    if (function == null) throw new IllegalArgumentException("foreachBatch function cannot be null")
    this.foreachBatchWriter = function
    this
  }

  /**
   * :: Experimental ::
   *
   * (Java-specific) Sets the output of the streaming query to be processed using the provided
   * function. This is supported only in the micro-batch execution modes (that is, when the
   * trigger is not continuous). In every micro-batch, the provided function will be called in
   * every micro-batch with (i) the output rows as a Dataset and (ii) the batch identifier.
   * The batchId can be used to deduplicate and transactionally write the output
   * (that is, the provided Dataset) to external systems. The output Dataset is guaranteed
   * to be exactly the same for the same batchId (assuming all operations are deterministic
   * in the query).
   *
   * @since 2.4.0
   */
  @Evolving
  def foreachBatch(function: VoidFunction2[Dataset[T], java.lang.Long]): DataStreamWriter[T] = {
    foreachBatch((batchDs: Dataset[T], batchId: Long) => function.call(batchDs, batchId))
  }

  private def normalizedParCols: Option[Seq[String]] = partitioningColumns.map { cols =>
    cols.map(normalize(_, "Partition"))
  }

  /**
   * The given column name may not be equal to any of the existing column names if we were in
   * case-insensitive context. Normalize the given column name to the real one so that we don't
   * need to care about case sensitivity afterwards.
   */
  private def normalize(columnName: String, columnType: String): String = {
    val validColumnNames = df.logicalPlan.output.map(_.name)
    validColumnNames.find(df.sparkSession.sessionState.analyzer.resolver(_, columnName))
      .getOrElse(throw QueryCompilationErrors.columnNotFoundInExistingColumnsError(
        columnType, columnName, validColumnNames))
  }

  private def assertNotPartitioned(operation: String): Unit = {
    if (partitioningColumns.isDefined) {
      throw QueryCompilationErrors.operationNotSupportPartitioningError(operation)
    }
  }

  ///////////////////////////////////////////////////////////////////////////////////////
  // Builder pattern config options
  ///////////////////////////////////////////////////////////////////////////////////////

  private var source: String = df.sparkSession.sessionState.conf.defaultDataSourceName

  private var tableName: String = null

  private var outputMode: OutputMode = OutputMode.Append

  private var trigger: Trigger = Trigger.ProcessingTime(0L)

  private var extraOptions = CaseInsensitiveMap[String](Map.empty)

  private var foreachWriter: ForeachWriter[T] = null

  private var foreachBatchWriter: (Dataset[T], Long) => Unit = null

  private var partitioningColumns: Option[Seq[String]] = None
}

object DataStreamWriter {
  val SOURCE_NAME_MEMORY = "memory"
  val SOURCE_NAME_FOREACH = "foreach"
  val SOURCE_NAME_FOREACH_BATCH = "foreachBatch"
  val SOURCE_NAME_CONSOLE = "console"
  val SOURCE_NAME_TABLE = "table"
  val SOURCE_NAME_NOOP = "noop"

  // these writer sources are also used for one-time query, hence allow temp checkpoint location
  val SOURCES_ALLOW_ONE_TIME_QUERY = Seq(SOURCE_NAME_MEMORY, SOURCE_NAME_FOREACH,
    SOURCE_NAME_FOREACH_BATCH, SOURCE_NAME_CONSOLE, SOURCE_NAME_NOOP)
}

相关信息

spark 源码目录

相关文章

spark DataStreamReader 源码

spark GroupState 源码

spark StreamingQuery 源码

spark StreamingQueryException 源码

spark StreamingQueryListener 源码

spark StreamingQueryManager 源码

spark StreamingQueryStatus 源码

spark TestGroupState 源码

spark progress 源码