将包含HashMap数据列表的RDD转换为dataFrame

Question

我正在编写一个代码来在没有com.databricks实用程序的情况下在spark中展平XML。

我编写了一个将XML转换为List＆gt;的函数。稍后将其存储到RDD。示例代码如下所示

  JavaPairRDD<Integer, List> ones=  lines.mapToPair(new PairFunction<Tuple2<LongWritable,Text>, Integer, List>() {

        private static final long serialVersionUID = 1L;
        void PairFunction(){}
        public Tuple2<Integer, List> call(
                Tuple2<LongWritable, Text> t) throws Exception {

                        List<Map<String,String>> List_Vals = new ArrayList<Map<String,String>>();
                        Map<String,String> vals = new LinkedHashMap<String,String>();
                        vals.put("Sequence", "1234");
                        vals.put("customer", "1234");
                        temp++;
                        List_Vals.add(vals);
                        Map<String,String> val2 = new LinkedHashMap<String,String>();
                        val2.put("Sequence", "5678");
                        val2.put("customer", "ABCDE");
                        List_Vals.add(val2);
                    return new Tuple2<Integer, List>(temp, List_Vals);  
        }
    });

如果我做ones.collect（）;我得到以下输出

 [(1,[{Sequence=1234, customer=1234}, {Sequence=5678, customer=ABCDE}])

但我希望以

格式获取输出

  SrN, Sequence, customer
  1,   1234    , 1234
  1,   5678    , ABCDE

我无法确定获得此输出的方法。请帮助