问题
When i write my JSON file with
dataframe.coalesce(1).write.format('json')
on pyspark im not able to change the name of file in the partition
Im writing my JSON like that:
dataframe.coalesce(1).write.format('json').mode('overwrite').save('path')
but im not able to change the name of file in the partition
I want the path like that:
/folder/my_name.json
where 'my_name.json' is a json file
回答1:
In spark we can't control name of the file written to the directory.
First write the data to the HDFS directory then For changing the name of file we need to use HDFS api.
Example:
In Pyspark:
l=[("a",1)]
ll=["id","sa"]
df=spark.createDataFrame(l,ll)
hdfs_dir = "/folder/" #your hdfs directory
new_filename="my_name.json" #new filename
df.coalesce(1).write.format("json").mode("overwrite").save(hdfs_dir)
fs = spark._jvm.org.apache.hadoop.fs.FileSystem.get(spark._jsc.hadoopConfiguration())
#list files in the directory
list_status = fs.listStatus(spark._jvm.org.apache.hadoop.fs.Path(hdfs_dir))
#filter name of the file starts with part-
file_name = [file.getPath().getName() for file in list_status if file.getPath().getName().startswith('part-')][0]
#rename the file
fs.rename(Path(hdfs_dir+''+file_name),Path(hdfs_dir+''+new_filename))
In case if you want to delete success files in the directory use fs.delete to delete _Success files.
In Scala:
val df=Seq(("a",1)).toDF("id","sa")
df.show(false)
import org.apache.hadoop.fs._
val hdfs_dir = "/folder/"
val new_filename="new_json.json"
df.coalesce(1).write.mode("overwrite").format("json").save(hdfs_dir)
val fs=FileSystem.get(sc.hadoopConfiguration)
val f=fs.globStatus(new Path(s"${hdfs_dir}" + "*")).filter(x => x.getPath.getName.toString.startsWith("part-")).map(x => x.getPath.getName).mkString
fs.rename(new Path(s"${hdfs_dir}${f}"),new Path(s"${hdfs_dir}${new_filename}"))
fs.delete(new Path(s"${hdfs_dir}" + "_SUCCESS"))
来源：https://stackoverflow.com/questions/58171365/how-to-rename-my-json-generated-by-pyspark