meituan / mssapi_python

美团云对象存储 Python SDK

MSS(Meituan Storage Service) SDK for python

This is MSS SDK for Python 2

Introduction

MSS服务介绍

美团云存储服务（Meituan Storage Service, 简称MSS)，是美团云对外提供的云存储服务，其具备高可靠，安全，低成本等特性，并且其API兼容S3。MSS适合存放非结构化的数据，比如图片，视频，文档，备份等。

MSS基本概念介绍

• MSS的API兼容S3, 其基本概念也和S3相同，主要包括Object, Bucket, Access Key, Secret Key等。

• Object对应一个文件，包括数据和元数据两部分。元数据以key-value的形式构成，它包含一些默认的元数据信息，比如Content-Type, Etag等，用户也可以自定义元数据。

• Bucket是object的容器，每个object都必须包含在一个bucket中。用户可以创建任意多个bucket。

• Access Key和Secret Key: 用户注册MSS时，系统会给用户分配一对Access Key和Secret Key, 用于标识用户，用户在使用API使用MSS服务时，需要使用这两个Key。请在美团云管理控制台查询AccessKey和SecretKey。

MSS访问域名

mtmss.com

Installation

git clone https://github.com/meituan/mssapi_python.git
cd mssapi_python
sudo python setup.py install

Quick Start

create S3 connection

import mssapi
from mssapi.s3.connection import S3Connection
from mssapi.s3.key import Key

conn = S3Connection(
    aws_access_key_id = access_key,
    aws_secret_access_key = access_secret,
    port = port,
    host = host,
)

handle bucket

create bucket

b0=conn.create_bucket('tmpbucket0')
b1=conn.create_bucket('tmpbucket1')

get buckets

bs = conn.get_all_buckets()
for b in bs:
    print b.name

get bucket

b1 = conn.get_bucket('tmpbucket1')

delete bucket

conn.delete_bucket(b1)

head bucket

conn.head_bucket('tmpbucket0')

bucket in

'tmpbucket0' in conn

get bucket keys

keys = b0.get_all_keys()
for k in keys:
    print k.name

handle Object

# First, you should get bucket instance
bucket = conn.get_bucket('tmpbucket0')

create Object

# Object are present as Key in following code
k0 = bucket.new_key('key0')
k0.set_contents_from_string('hello key0')

k1 = Key(bucket, 'key1')
k1.set_contents_from_filename('file_w1')

get Object

k0 = bucket.get_key('key0')
cont =  k0.get_contents_as_string()

k1 = Key(bucket, 'key1')
k1.get_contents_to_filename('file_r1')

delete Object

bucket.delete_key('key0')

lookup Object

bucket.lookup('key0')

generate temporary url

k1.generate_url(expires_in = 300)

handle multipart

# First, you need to init chunk_path and chunk_num

mp = bucket.initiate_multipart_upload('multipartkey')

for i in xrange(0, chunk_num):
    fp = open(chunk_path + str(i), 'r' )
    mp.upload_part_from_file(fp, part_num=i + 1)

mp.complete_upload()