Spark2.0SQL语句执行过程

simba.sql("Select * from b") 调用SparkSession.scala的sql()方法：

  def sql(sqlText: String): DataFrame = {
    Dataset.ofRows(self, sessionState.sqlParser.parsePlan(sqlText))
  }

Dataset.ofRows()方法:

 def ofRows(sparkSession: SparkSession, logicalPlan: LogicalPlan): DataFrame = {
    val qe = sparkSession.sessionState.executePlan(logicalPlan)
    qe.assertAnalyzed()
    new Dataset[Row](sparkSession, qe, RowEncoder(qe.analyzed.schema))

  }

SessionState.scala下面的executePlan()方法:
  def executePlan(plan: LogicalPlan): QueryExecution = new QueryExecution(sparkSession, plan)

至此，定义了一个QueryExecution()方法，用来直接相关的SparkSQL查询。

def assertAnalyzed(): Unit = {
    try sparkSession.sessionState.analyzer.checkAnalysis(analyzed) catch {
      case e: AnalysisException =>
        val ae = new AnalysisException(e.message, e.line, e.startPosition, Some(analyzed))
        ae.setStackTrace(e.getStackTrace)
        throw ae
    }
  }

检测分析是否发生异常

new Dataset[Row](sparkSession, qe, RowEncoder(qe.analyzed.schema))引用

class Dataset[T] private[sql](
    @transient val sparkSession: SparkSession,
    @DeveloperApi @InterfaceStability.Unstable @transient val queryExecution: QueryExecution,
    encoder: Encoder[T])

获得查询数据，至此结束。