Python系列之入门篇——HDFS

Python系列之入门篇——HDFS

简介

HDFS (Hadoop Distributed File System) Hadoop分布式文件系统，具有高容错性，适合部署在廉价的机器上。Python

提供了两种接口方式，分别是hdfscli(Restful Api Call)，pyhdfs(RPC Call)，这一节主要讲hdfscli的使用

代码示例

1. 安装

   pip install hdfs
   

2. 引入相关模块

   from hdfs import *
   

3. 创建客户端

   """
   It has two different kind of client, Client and InsecureClient.
   Client: cannot define file owner
   InsecureClient: can define file owner, default None
   """
   hdfs_root_path = 'http://localhost:50070'
   fs = Client(hdfs_root_path)
   fs = InsecureClient(hdfs_root_path, user='hdfs')
   

4. 创建目录

   """
   Change file permission to 777, default None
   """
   fs.makedirs('/test', permission=777)
   

5. 写文件

   """
   Write append or not depends on the file is exist or not
   strict: If `False`, return `None` rather than raise an exception if
         the path doesn't exist.
   """
   content = fs.content(hdfs_file_path, strict=False)
   if content is None:
       fs.write('/test/test.txt', data=data, permission=777)
   else:
       fs.write('/test/test.txt', data=data, append=True)
   

6. 上传文件

   """
   overwrite default False, if don't set True, when you upload the file which is exist
   in hdfs, it will raise File is exist Exception.
   """
   client.upload(hdfs_path, local_path, overwrite=True)
   

7. 总结
   
   还没有找到判断文件是否存在的方法，目前代码示例中用fs.content()来替换，如果大家有更好的方式，也麻烦分享给我