carbondata-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From kumarvishal09 <...@git.apache.org>
Subject [GitHub] carbondata pull request #1865: [CARBONDATA-2088][CARBONDATA-1516] Optimize s...
Date Mon, 29 Jan 2018 10:01:30 GMT
Github user kumarvishal09 commented on a diff in the pull request:

    https://github.com/apache/carbondata/pull/1865#discussion_r164381622
  
    --- Diff: integration/spark2/src/main/scala/org/apache/spark/sql/execution/command/datamap/CarbonCreateDataMapCommand.scala
---
    @@ -79,32 +79,27 @@ case class CarbonCreateDataMapCommand(
           }
           createPreAggregateTableCommands.flatMap(_.processMetadata(sparkSession))
         } else {
    -      val dataMapSchema = new DataMapSchema(dataMapName, dmClassName)
    -      dataMapSchema.setProperties(new java.util.HashMap[String, String](dmproperties.asJava))
    -      val dbName = CarbonEnv.getDatabaseName(tableIdentifier.database)(sparkSession)
    -      // upadting the parent table about dataschema
    -      PreAggregateUtil.updateMainTable(dbName, tableIdentifier.table, dataMapSchema,
sparkSession)
    +      throw new UnsupportedDataMapException(dmClassName)
         }
         LOGGER.audit(s"DataMap $dataMapName successfully added to Table ${ tableIdentifier.table
}")
         Seq.empty
       }
     
       override def processData(sparkSession: SparkSession): Seq[Row] = {
    -    if (dmClassName.equals("org.apache.carbondata.datamap.AggregateDataMapHandler") ||
    --- End diff --
    
    I think process meta will handle exception part if class name mentioned in create data
map statement is not valid. When call will come to processData or undoMeta validation is already
passed so no need to add class validation again


---

Mime
View raw message