[HUDI-2362] Add external config file support (#3416)

Co-authored-by: Wenning Ding <wenningd@amazon.com>
2021-11-18 01:59:26 -08:00
parent 8772cec4bd
commit 24def0b30d
25 changed files with 426 additions and 102 deletions
--- a/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/hudi/HoodieWriterUtils.scala
+++ b/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/hudi/HoodieWriterUtils.scala
@@ -22,7 +22,7 @@ import java.util.Properties
 import org.apache.hudi.DataSourceOptionsHelper.allAlternatives
 import org.apache.hudi.DataSourceWriteOptions._
 import org.apache.hudi.common.config.HoodieMetadataConfig.ENABLE
-import org.apache.hudi.common.config.{HoodieConfig, TypedProperties}
+import org.apache.hudi.common.config.{DFSPropertiesConfiguration, HoodieConfig, TypedProperties}
 import org.apache.hudi.common.table.HoodieTableConfig
 import org.apache.hudi.exception.HoodieException
 import org.apache.spark.sql.SparkSession
@@ -47,6 +47,7 @@ object HoodieWriterUtils {
    * @return
    */
  def parametersWithWriteDefaults(parameters: Map[String, String]): Map[String, String] = {
+    val globalProps = DFSPropertiesConfiguration.getGlobalProps.asScala
    Map(OPERATION.key -> OPERATION.defaultValue,
      TABLE_TYPE.key -> TABLE_TYPE.defaultValue,
      PRECOMBINE_FIELD.key -> PRECOMBINE_FIELD.defaultValue,
@@ -81,7 +82,7 @@ object HoodieWriterUtils {
      ENABLE_ROW_WRITER.key -> ENABLE_ROW_WRITER.defaultValue,
      RECONCILE_SCHEMA.key -> RECONCILE_SCHEMA.defaultValue.toString,
      DROP_PARTITION_COLUMNS.key -> DROP_PARTITION_COLUMNS.defaultValue
-    ) ++ DataSourceOptionsHelper.translateConfigurations(parameters)
+    ) ++ globalProps ++ DataSourceOptionsHelper.translateConfigurations(parameters)
  }

  def toProperties(params: Map[String, String]): TypedProperties = {
@@ -170,4 +171,23 @@ object HoodieWriterUtils {
      }
    }
  }
+
+  val sparkDatasourceConfigsToTableConfigsMap = Map(
+    TABLE_NAME -> HoodieTableConfig.NAME,
+    TABLE_TYPE -> HoodieTableConfig.TYPE,
+    PRECOMBINE_FIELD -> HoodieTableConfig.PRECOMBINE_FIELD,
+    PARTITIONPATH_FIELD -> HoodieTableConfig.PARTITION_FIELDS,
+    RECORDKEY_FIELD -> HoodieTableConfig.RECORDKEY_FIELDS,
+    PAYLOAD_CLASS_NAME -> HoodieTableConfig.PAYLOAD_CLASS_NAME
+  )
+  def mappingSparkDatasourceConfigsToTableConfigs(options: Map[String, String]): Map[String, String] = {
+    val includingTableConfigs = scala.collection.mutable.Map() ++ options
+    sparkDatasourceConfigsToTableConfigsMap.foreach(kv => {
+      if (options.containsKey(kv._1.key)) {
+        includingTableConfigs(kv._2.key) = options(kv._1.key)
+        includingTableConfigs.remove(kv._1.key)
+      }
+    })
+    includingTableConfigs.toMap
+  }
 }
--- a/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/HoodieSqlUtils.scala
+++ b/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/HoodieSqlUtils.scala
@@ -25,6 +25,7 @@ import org.apache.hadoop.conf.Configuration
 import org.apache.hadoop.fs.Path
 import org.apache.hudi.SparkAdapterSupport
 import org.apache.hudi.client.common.HoodieSparkEngineContext
+import org.apache.hudi.common.config.DFSPropertiesConfiguration
 import org.apache.hudi.common.config.HoodieMetadataConfig
 import org.apache.hudi.common.fs.FSUtils
 import org.apache.hudi.common.model.HoodieRecord
@@ -273,7 +274,7 @@ object HoodieSqlUtils extends SparkAdapterSupport {
   */
  def withSparkConf(spark: SparkSession, options: Map[String, String])
                   (baseConfig: Map[String, String] = Map.empty): Map[String, String] = {
-    baseConfig ++ // Table options has the highest priority
+    baseConfig ++ DFSPropertiesConfiguration.getGlobalProps.asScala ++ // Table options has the highest priority
      (spark.sessionState.conf.getAllConfs ++ HoodieOptionConfig.mappingSqlOptionToHoodieParam(options))
        .filterKeys(_.startsWith("hoodie."))
  }
--- a/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/AlterHoodieTableRenameCommand.scala
+++ b/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/AlterHoodieTableRenameCommand.scala
@@ -43,7 +43,7 @@ class AlterHoodieTableRenameCommand(
        .setConf(hadoopConf).build()
      // Init table with new name.
      HoodieTableMetaClient.withPropertyBuilder()
-        .fromProperties(metaClient.getTableConfig.getProps)
+        .fromProperties(metaClient.getTableConfig.getProps(true))
        .setTableName(newName.table)
        .initTable(hadoopConf, path)
      // Call AlterTableRenameCommand#run to rename table in meta.
--- a/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/CreateHoodieTableCommand.scala
+++ b/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/CreateHoodieTableCommand.scala
@@ -19,9 +19,9 @@ package org.apache.spark.sql.hudi.command

 import org.apache.hadoop.conf.Configuration
 import org.apache.hadoop.fs.Path
-
-import org.apache.hudi.{DataSourceWriteOptions, SparkAdapterSupport}
+import org.apache.hudi.{DataSourceWriteOptions, HoodieWriterUtils, SparkAdapterSupport}
 import org.apache.hudi.HoodieWriterUtils._
+import org.apache.hudi.common.config.DFSPropertiesConfiguration
 import org.apache.hudi.common.model.HoodieFileFormat
 import org.apache.hudi.common.table.{HoodieTableConfig, HoodieTableMetaClient}
 import org.apache.hudi.hadoop.HoodieParquetInputFormat
@@ -47,7 +47,7 @@ import org.apache.spark.sql.{AnalysisException, Row, SparkSession}
 import org.apache.spark.{SPARK_VERSION, SparkConf}

 import java.util.{Locale, Properties}
-
+import scala.collection.JavaConversions.mapAsJavaMap
 import scala.collection.JavaConverters._
 import scala.collection.mutable
 import scala.util.control.NonFatal
@@ -100,7 +100,10 @@ case class CreateHoodieTableCommand(table: CatalogTable, ignoreIfExists: Boolean
    // if CTAS, we treat the table we just created as nonexistent
    val isTableExists = if (ctas) false else tableExistsInPath(path, conf)
    var existingTableConfig = Map.empty[String, String]
-    val sqlOptions = HoodieOptionConfig.withDefaultSqlOptions(tblProperties)
+    val globalProps = DFSPropertiesConfiguration.getGlobalProps.asScala.toMap
+    val globalSqlProps = HoodieOptionConfig.mappingTableConfigToSqlOption(
+      HoodieWriterUtils.mappingSparkDatasourceConfigsToTableConfigs(globalProps))
+    val sqlOptions = HoodieOptionConfig.withDefaultSqlOptions(globalSqlProps ++ tblProperties)
    val catalogTableProps = HoodieOptionConfig.mappingSqlOptionToTableConfig(tblProperties)

    // get final schema and parameters
@@ -341,7 +344,7 @@ case class CreateHoodieTableCommand(table: CatalogTable, ignoreIfExists: Boolean
      }
    } else {
      extraConfig(HoodieTableConfig.HIVE_STYLE_PARTITIONING_ENABLE.key) = "true"
-      extraConfig(HoodieTableConfig.URL_ENCODE_PARTITIONING.key) = HoodieTableConfig.URL_ENCODE_PARTITIONING.defaultValue()
+      extraConfig(HoodieTableConfig.URL_ENCODE_PARTITIONING.key) = HoodieTableConfig.URL_ENCODE_PARTITIONING.defaultValue
    }

    if (originTableConfig.contains(HoodieTableConfig.KEY_GENERATOR_CLASS_NAME.key)) {
@@ -374,7 +377,7 @@ case class CreateHoodieTableCommand(table: CatalogTable, ignoreIfExists: Boolean
    checkTableConfigEqual(originTableConfig, tableOptions, HoodieTableConfig.URL_ENCODE_PARTITIONING.key)
    checkTableConfigEqual(originTableConfig, tableOptions, HoodieTableConfig.HIVE_STYLE_PARTITIONING_ENABLE.key)
    // Save all the table config to the hoodie.properties.
-    val parameters = originTableConfig ++ tableOptions
+    val parameters = HoodieWriterUtils.mappingSparkDatasourceConfigsToTableConfigs(originTableConfig ++ tableOptions)
    val properties = new Properties()
    properties.putAll(parameters.asJava)
    HoodieTableMetaClient.withPropertyBuilder()
--- a/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/DeleteHoodieTableCommand.scala
+++ b/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/DeleteHoodieTableCommand.scala
@@ -17,7 +17,7 @@

 package org.apache.spark.sql.hudi.command

-import org.apache.hudi.DataSourceWriteOptions.{OPERATION, _}
+import org.apache.hudi.DataSourceWriteOptions._
 import org.apache.hudi.common.table.HoodieTableMetaClient
 import org.apache.hudi.config.HoodieWriteConfig
 import org.apache.hudi.config.HoodieWriteConfig.TBL_NAME
--- a/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/TruncateHoodieTableCommand.scala
+++ b/hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/command/TruncateHoodieTableCommand.scala
@@ -45,7 +45,7 @@ class TruncateHoodieTableCommand(
      // Create MetaClient
      val metaClient = HoodieTableMetaClient.builder().setBasePath(path)
        .setConf(hadoopConf).build()
-      Some(metaClient.getTableConfig.getProps)
+      Some(metaClient.getTableConfig.getProps(true))
    } else {
      None
    }