val dfList = List(("Hadoop", "Java,SQL,Hive,HBase,MySQL"), ("Spark", "Scala,SQL,DataSet,MLlib,GraphX"))
dfList: List[(String, String)] = List((Hadoop,Java,SQL,Hive,HBase,MySQL), (Spark,Scala,SQL,DataSet,MLlib,GraphX))
case class Book(title: String, words: String)
val df=dfList.map{p=>Book(p._1,p._2)}.toDS()
df: org.apache.spark.sql.Dataset[Book] = [title: string, words: string]
df.show
+------+--------------------+
| title| words|
+------+--------------------+
|Hadoop|Java,SQL,Hive,HBa...|
| Spark|Scala,SQL,DataSet...|
+------+--------------------+
df.flatMap(_.words.split(",")).show
+-------+
| value|
+-------+
| Java|
| SQL|
| Hive|
| HBase|
| MySQL|
| Scala|
| SQL|
|DataSet|
| MLlib|
| GraphX|
+-------+
