SPARK-2407: Added Parser of SQL SUBSTR()

follow-up of #1359

Author: chutium <[email protected]>

Closes #1442 from chutium/master and squashes the following commits:

b49cc8a [chutium] SPARK-2407: Added Parser of SQL SUBSTRING() #1442
9a60ccf [chutium] SPARK-2407: Added Parser of SQL SUBSTR() #1442
06e933b [chutium] Merge https://github.com/apache/spark
c870172 [chutium] Merge https://github.com/apache/spark
094f773 [chutium] Merge https://github.com/apache/spark
88cb37d [chutium] Merge https://github.com/apache/spark
1de83a7 [chutium] SPARK-2407: Added Parse of SQL SUBSTR()

Author: chutium

sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/SqlParser.scala

@@ -120,7 +120,8 @@ class SqlParser extends StandardTokenParsers with PackratParsers {
   protected val WHERE = Keyword("WHERE")
   protected val INTERSECT = Keyword("INTERSECT")
   protected val EXCEPT = Keyword("EXCEPT")
-
+  protected val SUBSTR = Keyword("SUBSTR")
+  protected val SUBSTRING = Keyword("SUBSTRING")
 
   // Use reflection to find the reserved words defined in this class.
   protected val reservedWords =
@@ -316,6 +317,12 @@ class SqlParser extends StandardTokenParsers with PackratParsers {
     IF ~> "(" ~> expression ~ "," ~ expression ~ "," ~ expression <~ ")" ^^ {
       case c ~ "," ~ t ~ "," ~ f => If(c,t,f)
     } |
+    (SUBSTR | SUBSTRING) ~> "(" ~> expression ~ "," ~ expression <~ ")" ^^ {
+      case s ~ "," ~ p => Substring(s,p,Literal(Integer.MAX_VALUE))
+    } |
+    (SUBSTR | SUBSTRING) ~> "(" ~> expression ~ "," ~ expression ~ "," ~ expression <~ ")" ^^ {
+      case s ~ "," ~ p ~ "," ~ l => Substring(s,p,l)
+    } |
     ident ~ "(" ~ repsep(expression, ",") <~ ")" ^^ {
       case udfName ~ _ ~ exprs => UnresolvedFunction(udfName, exprs)
     }

sql/core/src/test/scala/org/apache/spark/sql/SQLQuerySuite.scala

@@ -36,6 +36,21 @@ class SQLQuerySuite extends QueryTest {
       "test")
   }
 
+  test("SPARK-2407 Added Parser of SQL SUBSTR()") {
+    checkAnswer(
+      sql("SELECT substr(tableName, 1, 2) FROM tableName"),
+      "te")
+    checkAnswer(
+      sql("SELECT substr(tableName, 3) FROM tableName"),
+      "st")
+    checkAnswer(
+      sql("SELECT substring(tableName, 1, 2) FROM tableName"),
+      "te")
+    checkAnswer(
+      sql("SELECT substring(tableName, 3) FROM tableName"),
+      "st")
+  }
+
   test("index into array") {
     checkAnswer(
       sql("SELECT data, data[0], data[0] + data[1], data[0 + 1] FROM arrayData"),