Mailing-List: contact issues-help@carbondata.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@carbondata.apache.org
From: ravipesala <git@git.apache.org>
To: issues@carbondata.apache.org
Reply-To: issues@carbondata.apache.org
References: <git-pr-1626-carbondata@git.apache.org>
In-Reply-To: <git-pr-1626-carbondata@git.apache.org>
Subject: [GitHub] carbondata pull request #1626: [CARBONDATA-1519][PreAgg-Timeseries] Support ...
Content-Type: text/plain
Message-Id: <20171206160227.49244E0219@git1-us-west.apache.org>
Date: Wed,  6 Dec 2017 16:02:27 +0000 (UTC)
archived-at: Wed, 06 Dec 2017 16:02:29 -0000

Github user ravipesala commented on a diff in the pull request:

    https://github.com/apache/carbondata/pull/1626#discussion_r155279382
  
    --- Diff: integration/spark2/src/main/scala/org/apache/spark/sql/execution/command/preaaggregate/PreAggregateUtil.scala ---
    @@ -493,4 +492,102 @@ object PreAggregateUtil {
         updatedPlan
       }
     
    +  /**
    +   * Below method will be used to get the select query when rollup policy is
    +   * applied in case of timeseries table
    +   * @param tableSchema
    +   *                    main data map schema
    +   * @param selectedDataMapSchema
    +   *                              selected data map schema for rollup
    +   * @return select query based on rolloup
    +   */
    +  def createTimeseriesSelectQueryForRollup(
    +      tableSchema: TableSchema,
    +      selectedDataMapSchema: AggregationDataMapSchema): String = {
    +    val aggregateColumns = scala.collection.mutable.ArrayBuffer.empty[String]
    +    val groupingExpressions = scala.collection.mutable.ArrayBuffer.empty[String]
    +    tableSchema.getListOfColumns.asScala.foreach {
    +      a => if (a.getAggFunction.nonEmpty) {
    +        aggregateColumns += s"${a.getAggFunction match {
    +          case "count" => "sum"
    +          case others@_ => others}}(${selectedDataMapSchema.getAggChildColByParent(
    +          a.getParentColumnTableRelations.get(0).getColumnName, a.getAggFunction).getColumnName})"
    +      } else if (a.getTimeSeriesFunction.nonEmpty) {
    +        groupingExpressions += s"timeseries(${
    +          selectedDataMapSchema
    +            .getNonAggNonTimeChildColBasedByParent(a.getParentColumnTableRelations.
    +              get(0).getColumnName).getColumnName
    +        } , '${ a.getTimeSeriesFunction }')"
    +      } else {
    +        groupingExpressions += selectedDataMapSchema
    +          .getNonAggNonTimeChildColBasedByParent(a.getParentColumnTableRelations.
    +            get(0).getColumnName).getColumnName
    +      }
    +    }
    +    s"select ${ groupingExpressions.mkString(",") },${ aggregateColumns.mkString(",")
    +    } from ${selectedDataMapSchema.getChildSchema.getTableName } " +
    +    s"group by ${ groupingExpressions.mkString(",") }"
    +  }
    +
    +  /**
    +   * Below method will be used to creating select query for timeseries
    +   * for lowest level for aggergation like second level, in that case it will
    +   * hit the maintable
    +   * @param tableSchema
    +   *                    data map schema
    +   * @param parentTableName
    +   *                        parent schema
    +   * @return select query for loading
    +   */
    +  def createTimeSeriesSelectQueryFromMain(tableSchema: TableSchema,
    +      parentTableName: String): String = {
    +    val aggregateColumns = scala.collection.mutable.ArrayBuffer.empty[String]
    +    val groupingExpressions = scala.collection.mutable.ArrayBuffer.empty[String]
    +    tableSchema.getListOfColumns.asScala.foreach {
    +      a =>
    +        if (a.getAggFunction.nonEmpty) {
    +          aggregateColumns +=
    +          s"${ a.getAggFunction }(${ a.getParentColumnTableRelations.get(0).getColumnName })"
    +        } else if (a.getTimeSeriesFunction.nonEmpty) {
    +          groupingExpressions +=
    +          s"timeseries(${ a.getParentColumnTableRelations.get(0).getColumnName },'${
    +            a
    +              .getTimeSeriesFunction
    +          }')"
    --- End diff --
    
    format properly


---