[SPARK-17068][SQL] Make view-usage visible during analysis

## What changes were proposed in this pull request?
This PR adds a field to subquery alias in order to make the usage of views in a resolved `LogicalPlan` more visible (and more understandable).

For example, the following view and query:
```sql
create view constants as select 1 as id union all select 1 union all select 42
select * from constants;
```
...now yields the following analyzed plan:
```
Project [id#39]
+- SubqueryAlias c, `default`.`constants`
   +- Project [gen_attr_0#36 AS id#39]
      +- SubqueryAlias gen_subquery_0
         +- Union
            :- Union
            :  :- Project [1 AS gen_attr_0#36]
            :  :  +- OneRowRelation$
            :  +- Project [1 AS gen_attr_1#37]
            :     +- OneRowRelation$
            +- Project [42 AS gen_attr_2#38]
               +- OneRowRelation$
```
## How was this patch tested?
Added tests for the two code paths in `SessionCatalogSuite` (sql/core) and `HiveMetastoreCatalogSuite` (sql/hive)

Author: Herman van Hovell <hvanhovell@databricks.com>

Closes #14657 from hvanhovell/SPARK-17068.

8 files changed, 38 insertions, 31 deletions

diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
index a2a022c247..bd4c19181f 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
@@ -138,7 +138,7 @@ class Analyzer(
         case u : UnresolvedRelation =>
           val substituted = cteRelations.find(x => resolver(x._1, u.tableIdentifier.table))
             .map(_._2).map { relation =>
-              val withAlias = u.alias.map(SubqueryAlias(_, relation))
+              val withAlias = u.alias.map(SubqueryAlias(_, relation, None))
               withAlias.getOrElse(relation)
             }
           substituted.getOrElse(u)
@@ -2057,7 +2057,7 @@ class Analyzer(
  */
 object EliminateSubqueryAliases extends Rule[LogicalPlan] {
   def apply(plan: LogicalPlan): LogicalPlan = plan transformUp {
-    case SubqueryAlias(_, child) => child
+    case SubqueryAlias(_, child, _) => child
   }
 }
 
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/CheckAnalysis.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/CheckAnalysis.scala
index 41b7e62d8c..e07e9194be 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/CheckAnalysis.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/CheckAnalysis.scala
@@ -141,8 +141,8 @@ trait CheckAnalysis extends PredicateHelper {
 
             // Skip projects and subquery aliases added by the Analyzer and the SQLBuilder.
             def cleanQuery(p: LogicalPlan): LogicalPlan = p match {
-              case SubqueryAlias(_, child) => cleanQuery(child)
-              case Project(_, child) => cleanQuery(child)
+              case s: SubqueryAlias => cleanQuery(s.child)
+              case p: Project => cleanQuery(p.child)
               case child => child
             }
 
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/catalog/SessionCatalog.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/catalog/SessionCatalog.scala
index 00c3db0aac..62d0da076b 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/catalog/SessionCatalog.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/catalog/SessionCatalog.scala
@@ -411,27 +411,29 @@ class SessionCatalog(
   }
 
   /**
-   * Return a [[LogicalPlan]] that represents the given table.
+   * Return a [[LogicalPlan]] that represents the given table or view.
    *
-   * If a database is specified in `name`, this will return the table from that database.
-   * If no database is specified, this will first attempt to return a temporary table with
-   * the same name, then, if that does not exist, return the table from the current database.
+   * If a database is specified in `name`, this will return the table/view from that database.
+   * If no database is specified, this will first attempt to return a temporary table/view with
+   * the same name, then, if that does not exist, return the table/view from the current database.
+   *
+   * If the relation is a view, the relation will be wrapped in a [[SubqueryAlias]] which will
+   * track the name of the view.
    */
   def lookupRelation(name: TableIdentifier, alias: Option[String] = None): LogicalPlan = {
     synchronized {
       val db = formatDatabaseName(name.database.getOrElse(currentDb))
       val table = formatTableName(name.table)
-      val relation =
-        if (name.database.isDefined || !tempTables.contains(table)) {
-          val metadata = externalCatalog.getTable(db, table)
-          SimpleCatalogRelation(db, metadata)
-        } else {
-          tempTables(table)
+      val relationAlias = alias.getOrElse(table)
+      if (name.database.isDefined || !tempTables.contains(table)) {
+        val metadata = externalCatalog.getTable(db, table)
+        val view = Option(metadata.tableType).collect {
+          case CatalogTableType.VIEW => name
         }
-      val qualifiedTable = SubqueryAlias(table, relation)
-      // If an alias was specified by the lookup, wrap the plan in a subquery so that
-      // attributes are properly qualified with this alias.
-      alias.map(a => SubqueryAlias(a, qualifiedTable)).getOrElse(qualifiedTable)
+        SubqueryAlias(relationAlias, SimpleCatalogRelation(db, metadata), view)
+      } else {
+        SubqueryAlias(relationAlias, tempTables(table), Option(name))
+      }
     }
   }
 
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/dsl/package.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/dsl/package.scala
index 5181dcc786..9f54d709a0 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/dsl/package.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/dsl/package.scala
@@ -343,7 +343,7 @@ package object dsl {
           orderSpec: Seq[SortOrder]): LogicalPlan =
         Window(windowExpressions, partitionSpec, orderSpec, logicalPlan)
 
-      def subquery(alias: Symbol): LogicalPlan = SubqueryAlias(alias.name, logicalPlan)
+      def subquery(alias: Symbol): LogicalPlan = SubqueryAlias(alias.name, logicalPlan, None)
 
       def except(otherPlan: LogicalPlan): LogicalPlan = Except(logicalPlan, otherPlan)
 
@@ -367,7 +367,7 @@ package object dsl {
 
       def as(alias: String): LogicalPlan = logicalPlan match {
         case UnresolvedRelation(tbl, _) => UnresolvedRelation(tbl, Option(alias))
-        case plan => SubqueryAlias(alias, plan)
+        case plan => SubqueryAlias(alias, plan, None)
       }
 
       def repartition(num: Integer): LogicalPlan =
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/subquery.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/subquery.scala
index ac44f08897..ddbe937cba 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/subquery.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/subquery.scala
@@ -72,7 +72,7 @@ case class ScalarSubquery(
   override def dataType: DataType = query.schema.fields.head.dataType
   override def foldable: Boolean = false
   override def nullable: Boolean = true
-  override def plan: LogicalPlan = SubqueryAlias(toString, query)
+  override def plan: LogicalPlan = SubqueryAlias(toString, query, None)
   override def withNewPlan(plan: LogicalPlan): ScalarSubquery = copy(query = plan)
   override def toString: String = s"scalar-subquery#${exprId.id} $conditionString"
 }
@@ -100,7 +100,7 @@ case class PredicateSubquery(
   override lazy val resolved = childrenResolved && query.resolved
   override lazy val references: AttributeSet = super.references -- query.outputSet
   override def nullable: Boolean = nullAware
-  override def plan: LogicalPlan = SubqueryAlias(toString, query)
+  override def plan: LogicalPlan = SubqueryAlias(toString, query, None)
   override def withNewPlan(plan: LogicalPlan): PredicateSubquery = copy(query = plan)
   override def semanticEquals(o: Expression): Boolean = o match {
     case p: PredicateSubquery =>
@@ -153,7 +153,7 @@ case class ListQuery(query: LogicalPlan, exprId: ExprId = NamedExpression.newExp
   override def dataType: DataType = ArrayType(NullType)
   override def nullable: Boolean = false
   override def withNewPlan(plan: LogicalPlan): ListQuery = copy(query = plan)
-  override def plan: LogicalPlan = SubqueryAlias(toString, query)
+  override def plan: LogicalPlan = SubqueryAlias(toString, query, None)
   override def toString: String = s"list#${exprId.id}"
 }
 
@@ -174,6 +174,6 @@ case class Exists(query: LogicalPlan, exprId: ExprId = NamedExpression.newExprId
   override def children: Seq[Expression] = Seq.empty
   override def nullable: Boolean = false
   override def withNewPlan(plan: LogicalPlan): Exists = copy(query = plan)
-  override def plan: LogicalPlan = SubqueryAlias(toString, query)
+  override def plan: LogicalPlan = SubqueryAlias(toString, query, None)
   override def toString: String = s"exists#${exprId.id}"
 }
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala
index e34a478818..f97a78b411 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala
@@ -1862,7 +1862,7 @@ object RewriteCorrelatedScalarSubquery extends Rule[LogicalPlan] {
     // and Project operators, followed by an optional Filter, followed by an
     // Aggregate. Traverse the operators recursively.
     def evalPlan(lp : LogicalPlan) : Map[ExprId, Option[Any]] = lp match {
-      case SubqueryAlias(_, child) => evalPlan(child)
+      case SubqueryAlias(_, child, _) => evalPlan(child)
       case Filter(condition, child) =>
         val bindings = evalPlan(child)
         if (bindings.isEmpty) bindings
@@ -1920,7 +1920,7 @@ object RewriteCorrelatedScalarSubquery extends Rule[LogicalPlan] {
           topPart += p
           bottomPart = child
 
-        case s @ SubqueryAlias(_, child) =>
+        case s @ SubqueryAlias(_, child, _) =>
           topPart += s
           bottomPart = child
 
@@ -1991,8 +1991,8 @@ object RewriteCorrelatedScalarSubquery extends Rule[LogicalPlan] {
             topPart.reverse.foreach {
               case Project(projList, _) =>
                 subqueryRoot = Project(projList ++ havingInputs, subqueryRoot)
-              case s @ SubqueryAlias(alias, _) =>
-                subqueryRoot = SubqueryAlias(alias, subqueryRoot)
+              case s @ SubqueryAlias(alias, _, None) =>
+                subqueryRoot = SubqueryAlias(alias, subqueryRoot, None)
               case op => sys.error(s"Unexpected operator $op in corelated subquery")
             }
 
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala
index 09b650ce18..adf78396d7 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala
@@ -107,7 +107,7 @@ class AstBuilder extends SqlBaseBaseVisitor[AnyRef] with Logging {
    * This is only used for Common Table Expressions.
    */
   override def visitNamedQuery(ctx: NamedQueryContext): SubqueryAlias = withOrigin(ctx) {
-    SubqueryAlias(ctx.name.getText, plan(ctx.queryNoWith))
+    SubqueryAlias(ctx.name.getText, plan(ctx.queryNoWith), None)
   }
 
   /**
@@ -723,7 +723,7 @@ class AstBuilder extends SqlBaseBaseVisitor[AnyRef] with Logging {
    * Create an alias (SubqueryAlias) for a LogicalPlan.
    */
   private def aliasPlan(alias: ParserRuleContext, plan: LogicalPlan): LogicalPlan = {
-    SubqueryAlias(alias.getText, plan)
+    SubqueryAlias(alias.getText, plan, None)
   }
 
   /**
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/basicLogicalOperators.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/basicLogicalOperators.scala
index 2917d8d2a9..af1736e607 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/basicLogicalOperators.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/basicLogicalOperators.scala
@@ -19,6 +19,7 @@ package org.apache.spark.sql.catalyst.plans.logical
 
 import scala.collection.mutable.ArrayBuffer
 
+import org.apache.spark.sql.catalyst.TableIdentifier
 import org.apache.spark.sql.catalyst.analysis.MultiInstanceRelation
 import org.apache.spark.sql.catalyst.expressions._
 import org.apache.spark.sql.catalyst.expressions.aggregate.AggregateExpression
@@ -693,7 +694,11 @@ case class LocalLimit(limitExpr: Expression, child: LogicalPlan) extends UnaryNo
   }
 }
 
-case class SubqueryAlias(alias: String, child: LogicalPlan) extends UnaryNode {
+case class SubqueryAlias(
+    alias: String,
+    child: LogicalPlan,
+    view: Option[TableIdentifier])
+  extends UnaryNode {
 
   override def output: Seq[Attribute] = child.output.map(_.withQualifier(Some(alias)))
 }