Fix dependency and test case.

Author: dongjoon-hyun

dev/deps/spark-deps-hadoop-2.2

@@ -132,7 +132,7 @@ paranamer-2.3.jar
 parquet-column-1.9.0.jar
 parquet-common-1.9.0.jar
 parquet-encoding-1.9.0.jar
-parquet-format-2.3.0-incubating.jar
+parquet-format-2.3.1.jar
 parquet-hadoop-1.9.0.jar
 parquet-hadoop-bundle-1.6.0.jar
 parquet-jackson-1.9.0.jar

dev/deps/spark-deps-hadoop-2.3

@@ -139,7 +139,7 @@ paranamer-2.3.jar
 parquet-column-1.9.0.jar
 parquet-common-1.9.0.jar
 parquet-encoding-1.9.0.jar
-parquet-format-2.3.0-incubating.jar
+parquet-format-2.3.1.jar
 parquet-hadoop-1.9.0.jar
 parquet-hadoop-bundle-1.6.0.jar
 parquet-jackson-1.9.0.jar

dev/deps/spark-deps-hadoop-2.4

@@ -139,7 +139,7 @@ paranamer-2.3.jar
 parquet-column-1.9.0.jar
 parquet-common-1.9.0.jar
 parquet-encoding-1.9.0.jar
-parquet-format-2.3.0-incubating.jar
+parquet-format-2.3.1.jar
 parquet-hadoop-1.9.0.jar
 parquet-hadoop-bundle-1.6.0.jar
 parquet-jackson-1.9.0.jar

dev/deps/spark-deps-hadoop-2.6

@@ -147,7 +147,7 @@ paranamer-2.3.jar
 parquet-column-1.9.0.jar
 parquet-common-1.9.0.jar
 parquet-encoding-1.9.0.jar
-parquet-format-2.3.0-incubating.jar
+parquet-format-2.3.1.jar
 parquet-hadoop-1.9.0.jar
 parquet-hadoop-bundle-1.6.0.jar
 parquet-jackson-1.9.0.jar

dev/deps/spark-deps-hadoop-2.7

@@ -148,7 +148,7 @@ paranamer-2.3.jar
 parquet-column-1.9.0.jar
 parquet-common-1.9.0.jar
 parquet-encoding-1.9.0.jar
-parquet-format-2.3.0-incubating.jar
+parquet-format-2.3.1.jar
 parquet-hadoop-1.9.0.jar
 parquet-hadoop-bundle-1.6.0.jar
 parquet-jackson-1.9.0.jar

sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/parquet/ParquetSchemaSuite.scala

@@ -1424,7 +1424,7 @@ class ParquetSchemaSuite extends ParquetSchemaTest {
 
     catalystSchema = new StructType(),
 
-    expectedSchema = "message root {}")
+    expectedSchema = ParquetSchemaConverter.EMPTY_MESSAGE)
 
   testSchemaClipping(
     "disjoint field sets",