保存成json格式,并压缩
spark.range(20).repartition(1).write.mode('overwrite').json(path='out/20200715',compression='gzip')