[SPARK-13535][SQL] Fix Analysis Exceptions when Using Backticks in Tr…

…ansform Clause

#### What changes were proposed in this pull request?
```SQL
FROM
(FROM test SELECT TRANSFORM(key, value) USING 'cat' AS (`thing1` int, thing2 string)) t
SELECT thing1 + 1
```
This query returns an analysis error, like:
```
Failed to analyze query: org.apache.spark.sql.AnalysisException: cannot resolve '`thing1`' given input columns: [`thing1`, thing2]; line 3 pos 7
'Project [unresolvedalias(('thing1 + 1), None)]
+- SubqueryAlias t
   +- ScriptTransformation [key#2,value#3], cat, [`thing1`#6,thing2#7], HiveScriptIOSchema(List(),List(),Some(org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe),Some(org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe),List((field.delim,	)),List((field.delim,	)),Some(org.apache.hadoop.hive.ql.exec.TextRecordReader),Some(org.apache.hadoop.hive.ql.exec.TextRecordWriter),false)
      +- SubqueryAlias test
         +- Project [_1#0 AS key#2,_2#1 AS value#3]
            +- LocalRelation [_1#0,_2#1], [[1,1],[2,2],[3,3],[4,4],[5,5]]
```

The backpacks of \`thing1\` should be cleaned before entering Parser/Analyzer. This PR fixes this issue.

#### How was this patch tested?

Added a test case and modified an existing test case

Author: gatorsmile <[email protected]>

Closes apache#11415 from gatorsmile/scriptTransform.

sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveQl.scala

@@ -583,11 +583,11 @@ private[hive] class HiveQl(conf: ParserConf) extends SparkQl(conf) with Logging
 
       val (output, schemaLess) = outputClause match {
         case Token("TOK_ALIASLIST", aliases) :: Nil =>
-          (aliases.map { case Token(name, Nil) => AttributeReference(name, StringType)() },
-            false)
+          (aliases.map { case Token(name, Nil) =>
+            AttributeReference(cleanIdentifier(name), StringType)() }, false)
         case Token("TOK_TABCOLLIST", attributes) :: Nil =>
           (attributes.map { case Token("TOK_TABCOL", Token(name, Nil) :: dataType :: Nil) =>
-            AttributeReference(name, nodeToDataType(dataType))() }, false)
+            AttributeReference(cleanIdentifier(name), nodeToDataType(dataType))() }, false)
         case Nil =>
           (List(AttributeReference("key", StringType)(),
             AttributeReference("value", StringType)()), true)

sql/hive/src/test/scala/org/apache/spark/sql/hive/HiveQlSuite.scala

@@ -200,4 +200,12 @@ class HiveQlSuite extends SparkFunSuite with BeforeAndAfterAll {
 
     assert(plan.children.head.asInstanceOf[Generate].generator.isInstanceOf[JsonTuple])
   }
+
+  test("use backticks in output of Script Transform") {
+    val plan = parser.parsePlan(
+      """SELECT `t`.`thing1`
+        |FROM (SELECT TRANSFORM (`parquet_t1`.`key`, `parquet_t1`.`value`)
+        |USING 'cat' AS (`thing1` int, `thing2` string) FROM `default`.`parquet_t1`) AS t
+      """.stripMargin)
+  }
 }

sql/hive/src/test/scala/org/apache/spark/sql/hive/execution/SQLQuerySuite.scala

@@ -731,7 +731,7 @@ class SQLQuerySuite extends QueryTest with SQLTestUtils with TestHiveSingleton {
     data.toDF("key", "value").registerTempTable("test")
     checkAnswer(
       sql("""FROM
-          |(FROM test SELECT TRANSFORM(key, value) USING 'cat' AS (thing1 int, thing2 string)) t
+          |(FROM test SELECT TRANSFORM(key, value) USING 'cat' AS (`thing1` int, thing2 string)) t
           |SELECT thing1 + 1
         """.stripMargin), (2 to 6).map(i => Row(i)))
   }