[SPARK-18947][SQL] SQLContext.tableNames should not call Catalog.listTables

## What changes were proposed in this pull request?

It's a huge waste to call `Catalog.listTables` in `SQLContext.tableNames`, which only need the table names, while `Catalog.listTables` will get the table metadata for each table name.

## How was this patch tested?

N/A

Author: Wenchen Fan <wenchen@databricks.com>

Closes #16352 from cloud-fan/minor.

sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala

@@ -747,7 +747,7 @@ class SQLContext private[sql](val sparkSession: SparkSession)
    * @since 1.3.0
    */
   def tableNames(): Array[String] = {
-    sparkSession.catalog.listTables().collect().map(_.name)
+    tableNames(sparkSession.catalog.currentDatabase)
   }
 
   /**
@@ -757,7 +757,7 @@ class SQLContext private[sql](val sparkSession: SparkSession)
    * @since 1.3.0
    */
   def tableNames(databaseName: String): Array[String] = {
-    sparkSession.catalog.listTables(databaseName).collect().map(_.name)
+    sessionState.catalog.listTables(databaseName).map(_.table).toArray
   }
 
   ////////////////////////////////////////////////////////////////////////////

sql/core/src/main/scala/org/apache/spark/sql/api/r/SQLUtils.scala

@@ -276,11 +276,12 @@ private[sql] object SQLUtils extends Logging {
   }
 
   def getTableNames(sparkSession: SparkSession, databaseName: String): Array[String] = {
-    databaseName match {
-      case n: String if n != null && n.trim.nonEmpty =>
-        sparkSession.catalog.listTables(n).collect().map(_.name)
+    val db = databaseName match {
+      case _ if databaseName != null && databaseName.trim.nonEmpty =>
+        databaseName
       case _ =>
-        sparkSession.catalog.listTables().collect().map(_.name)
+        sparkSession.catalog.currentDatabase
     }
+    sparkSession.sessionState.catalog.listTables(db).map(_.table).toArray
   }
 }