From 34336b6250d99bcf009b082cbf83f326d6b00074 Mon Sep 17 00:00:00 2001
From: Andrew Or <andrew@databricks.com>
Date: Mon, 25 Apr 2016 15:30:18 -0700
Subject: [SPARK-14828][SQL] Start SparkSession in REPL instead of SQLContext

## What changes were proposed in this pull request?

```
Spark context available as 'sc' (master = local[*], app id = local-1461283768192).
Spark session available as 'spark'.
Welcome to
      ____              __
     / __/__  ___ _____/ /__
    _\ \/ _ \/ _ `/ __/  '_/
   /___/ .__/\_,_/_/ /_/\_\   version 2.0.0-SNAPSHOT
      /_/

Using Scala version 2.11.8 (Java HotSpot(TM) 64-Bit Server VM, Java 1.7.0_51)
Type in expressions to have them evaluated.
Type :help for more information.

scala> sql("SHOW TABLES").collect()
16/04/21 17:09:39 WARN ObjectStore: Version information not found in metastore. hive.metastore.schema.verification is not enabled so recording the schema version 1.2.0
16/04/21 17:09:39 WARN ObjectStore: Failed to get database default, returning NoSuchObjectException
res0: Array[org.apache.spark.sql.Row] = Array([src,false])

scala> sql("SHOW TABLES").collect()
res1: Array[org.apache.spark.sql.Row] = Array([src,false])

scala> spark.createDataFrame(Seq((1, 1), (2, 2), (3, 3)))
res2: org.apache.spark.sql.DataFrame = [_1: int, _2: int]
```

Hive things are loaded lazily.

## How was this patch tested?

Manual.

Author: Andrew Or <andrew@databricks.com>

Closes #12589 from andrewor14/spark-session-repl.
---
 .../scala/org/apache/spark/sql/SparkSession.scala  | 34 ++++++++++++++++++----
 1 file changed, 29 insertions(+), 5 deletions(-)

(limited to 'sql')

diff --git a/sql/core/src/main/scala/org/apache/spark/sql/SparkSession.scala b/sql/core/src/main/scala/org/apache/spark/sql/SparkSession.scala
index 131f28f98b..3561765642 100644
--- a/sql/core/src/main/scala/org/apache/spark/sql/SparkSession.scala
+++ b/sql/core/src/main/scala/org/apache/spark/sql/SparkSession.scala
@@ -907,16 +907,19 @@ class SparkSession private(
 
 object SparkSession {
 
+  private val HIVE_SHARED_STATE_CLASS_NAME = "org.apache.spark.sql.hive.HiveSharedState"
+  private val HIVE_SESSION_STATE_CLASS_NAME = "org.apache.spark.sql.hive.HiveSessionState"
+
   private def sharedStateClassName(conf: SparkConf): String = {
     conf.get(CATALOG_IMPLEMENTATION) match {
-      case "hive" => "org.apache.spark.sql.hive.HiveSharedState"
+      case "hive" => HIVE_SHARED_STATE_CLASS_NAME
       case "in-memory" => classOf[SharedState].getCanonicalName
     }
   }
 
   private def sessionStateClassName(conf: SparkConf): String = {
     conf.get(CATALOG_IMPLEMENTATION) match {
-      case "hive" => "org.apache.spark.sql.hive.HiveSessionState"
+      case "hive" => HIVE_SESSION_STATE_CLASS_NAME
       case "in-memory" => classOf[SessionState].getCanonicalName
     }
   }
@@ -938,10 +941,31 @@ object SparkSession {
     }
   }
 
-  // TODO: do we want to expose this?
+  /**
+   * Return true if Hive classes can be loaded, otherwise false.
+   */
+  private[spark] def hiveClassesArePresent: Boolean = {
+    try {
+      Utils.classForName(HIVE_SESSION_STATE_CLASS_NAME)
+      Utils.classForName(HIVE_SHARED_STATE_CLASS_NAME)
+      Utils.classForName("org.apache.hadoop.hive.conf.HiveConf")
+      true
+    } catch {
+      case _: ClassNotFoundException | _: NoClassDefFoundError => false
+    }
+  }
+
+  /**
+   * Create a new [[SparkSession]] with a catalog backed by Hive.
+   */
   def withHiveSupport(sc: SparkContext): SparkSession = {
-    sc.conf.set(CATALOG_IMPLEMENTATION.key, "hive")
-    new SparkSession(sc)
+    if (hiveClassesArePresent) {
+      sc.conf.set(CATALOG_IMPLEMENTATION.key, "hive")
+      new SparkSession(sc)
+    } else {
+      throw new IllegalArgumentException(
+        "Unable to instantiate SparkSession with Hive support because Hive classes are not found.")
+    }
   }
 
 }
-- 
cgit v1.2.3