[SPARK-15211][SQL] Select features column from LibSVMRelation causes failure

## What changes were proposed in this pull request?

We need to use `requiredSchema` in `LibSVMRelation` to project the fetch required columns when loading data from this data source. Otherwise, when users try to select `features` column, it will cause failure.

## How was this patch tested?
`LibSVMRelationSuite`.

Author: Liang-Chi Hsieh <simonh@tw.ibm.com>

Closes #12986 from viirya/fix-libsvmrelation.

mllib/src/main/scala/org/apache/spark/ml/source/libsvm/LibSVMRelation.scala

@@ -203,10 +203,18 @@ class DefaultSource extends FileFormat with DataSourceRegister {
           }
 
       val converter = RowEncoder(dataSchema)
+      val fullOutput = dataSchema.map { f =>
+        AttributeReference(f.name, f.dataType, f.nullable, f.metadata)()
+      }
+      val requiredOutput = fullOutput.filter { a =>
+        requiredSchema.fieldNames.contains(a.name)
+      }
+
+      val requiredColumns = GenerateUnsafeProjection.generate(requiredOutput, fullOutput)
 
       points.map { pt =>
         val features = if (sparse) pt.features.toSparse else pt.features.toDense
-        converter.toRow(Row(pt.label, features))
+        requiredColumns(converter.toRow(Row(pt.label, features)))
       }
     }
   }

mllib/src/test/scala/org/apache/spark/ml/source/libsvm/LibSVMRelationSuite.scala

@@ -108,5 +108,6 @@ class LibSVMRelationSuite extends SparkFunSuite with MLlibTestSparkContext {
   test("select features from libsvm relation") {
     val df = sqlContext.read.format("libsvm").load(path)
     df.select("features").rdd.map { case Row(d: Vector) => d }.first
+    df.select("features").collect
   }
 }