帮助与文档

    搜索
    • 云服务器
    • 弹性IP
    • VPC
    • CDN
    • IAM

    文档中心

    对象存储(KS3)

    金山云 > 文档中心 > 对象存储(KS3) > SDK

    Python

    最近更新时间：2020-12-22 16:01:05

    KS3 SDK for python使用指南

    目录

    开发前准备

    安装依赖模块

    pip install six

    安装python sdk

    在线安装
    pip install ks3sdk
    本地安装

    1、通过git下载SDK到本地

    git clone https://github.com/ks3sdk/ks3-python-sdk.git

    2、进入ks3-python-sdk目录

    cd  ks3-python-sdk

    3、安装SDK

    python setup.py install

    初始化connection

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)

    参数说明

    • ak：金山云提供的ACCESS KEY ID

    • sk：金山云提供的SECRET KEY ID

    • host：金山云提供的各个Region的域名（例 ks3-cn-beijing.ksyun.com）,具体定义可参考 API接口文档-Region(区域)。也可以是用户自定义域名，如果是用户自定义域名，需要将domain_mode设置为True。

    • is_secure：是否通过HTTPS协议访问Ks3，True:启用 False:关闭。

    • domain_mode：是否使用自定义域名访问Ks3（host填写自定义域名），True:是 False:否。

    运行环境

    适用于2.6、2.7、3.3、3.4、3.5、3.6、3.7版本。

    SDK介绍及使用

    资源管理操作

    Service操作

    List Buckets：

    列出客户所有的 Bucket 信息

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)
buckets = c.get_all_buckets()
for b in buckets:
    print(b.name)

    Bucket操作

    Create Bucket：

    创建一个新的Bucket

    在建立了连接后，可以创建一个bucket。bucket在s3中是一个用于储存key/value的容器。用户可以将所有的数据存储在一个bucket里，也可以为不同种类数据创建相应的bucket。

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)
bucket_name = "YOUR_BUCKET_NAME"
b = c.create_bucket(bucket_name)

    注：这里如果出现409 conflict错误，说明请求的bucket name有冲突，因为bucket name是全局唯一的

    Delete Bucket:

    删除指定Bucket

    删除一个bucket可以通过delete_bucket方法实现。

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)
bucket_name = "YOUR_BUCKET_NAME" 
c.delete_bucket(bucket_name)

    如果bucket下面存在key，那么需要首先删除所有key

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)
bucket_name = "YOUR_BUCKET_NAME"
b = c.get_bucket(bucket_name)
for k in b.list():
    k.delete()
c.delete_bucket(bucket_name)

    Get Bucket ACL:

    获取Bucket的ACL

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)
bucket_name = "YOUR_BUCKET_NAME"
b = c.get_bucket(bucket_name)
policy = b.get_acl()
for grant in policy.acl.grants:
    print(grant.permission, grant.display_name, grant.email_address, grant.id)

    Put Bucket ACL:

    设置Bucket的ACL

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)

bucket_name = "YOUR_BUCKET_NAME"
b = c.get_bucket(bucket_name)
#设置bucket的权限, private or public-read or public-read-write
b.set_acl("public-read")

    Put Bucket Crr：

    为Bucket设置跨区域复制规则

    对bucket配置跨region复制规则，支持添加、查询、删除

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False) 
bucket_name = "YOUR_BUCKET_NAME"
bucket = c.get_bucket(bucket_name)
bucket.set_bucket_crr('emmacrra', 'Disabled')
#bucket_crr = bucket.get_bucket_crr() 
#bucket.delete_bucket_crr()

    Object操作

    Head Object:

    获取Object元信息

    获取Object元数据信息（大小、最后更新时间等）

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)

bucket_name = "YOUR_BUCKET_NAME"
key_name = "YOUR_KEY_NAME"
b = c.get_bucket(bucket_name)
try:
    k = b.get_key(key_name)
    if k:
         print(k.name, k.size, k.last_modified)
except:
    pass #异常处理

    Get Object：

    下载该Object数据

    下载object，并且作为字符串返回

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)

bucket_name = "YOUR_BUCKET_NAME"
key_name = "YOUR_KEY_NAME"
b = c.get_bucket(bucket_name)
try:
    k = b.get_key(key_name)
    s = k.get_contents_as_string().decode()
    print(s)
except:
    pass #异常处理

    下载object，并且保存到文件中

    #保存到文件
k.get_contents_to_filename("SAVED_FILE_PATH")

    Put Object

    上传Object数据

    将指定目录下某一个文件上传，同时可以指定文件ACL

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)  

bucket_name = "YOUR_BUCKET_NAME"
key_name = "YOUR_KEY_NAME"
try: 
    b = c.get_bucket(bucket_name)
    k = b.new_key(key_name)
    #object policy : 'private' or 'public-read'
    ret=k.set_contents_from_filename("YOUR_SOURCE_FILE_PATH", policy="private")
    if ret and ret.status == 200:
        print("上传成功")
except:
    pass #异常处理

    将字符串作为value上传

    k.set_contents_from_string('YOUR_FILE_CONTENTS')

    上传回调 需要设置x-kss-callbackurl和x-kss-callbackbody请求头，详见文档上传回调处理。

    headers = {"x-kss-callbackurl": "Callback_URL","x-kss-callbackbody":"objectKey=${key}&etag=${etag}&uid=123"}
ret = k.set_contents_from_filename(filepath, headers=headers)

    以低频存储方式上传

    //x-kss-storage-class有效值为"STANDARD"、"STANDARD_IA"。"STANDARD"表示标准存储，"STANDARD_IA"表示低频存储，如果不指定，默认为标准存储。
headers = {"x-kss-storage-class": "STANDARD_IA"}
resp = k.set_contents_from_filename(filepath, policy="private", headers=headers)

    抓取网络资源到KS3 content-type为非必填 x-kss-acl根据需要设定公开读 public-read，当未设置时为私密

    import ks3.auth as auth
class TestFetchObject(unittest.TestCase):
    # Create a object key in specified bucket
    def test_fetch_object(self):
        auth.qsa_of_interest.append('fetch')
        headerss = {'x-kss-sourceurl': 'http%3A%2F%2Fpic1.sc.chinaz.com%2Ffiles%2Fpic%2Fpic9%2F202006%2Fbpic20469.jpg',
                    'x-kss-acl': 'public-read', 'Content-Type': 'application/xml'}
        resp = conn.make_request('PUT', test_bucket, "testfetch", query_args='fetch', headers=headerss)

        print(resp.read())

    Put Object Copy:

    复制Object数据

    将指定Bucket下的文件复制到本Bucket下（需要对源Bucket下的文件具有读权限）

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False) 

bucket_name = "YOUR_DST_BUCKET_NAME"
key_name = "YOUR_DST_KEY_NAME"
try: 
    b = c.get_bucket(bucket_name)
    b.copy_key(key_name,"YOUR_SRC_BUCKET_NAME","YOUR_SRC_KEY_NAME")
except:
    pass #异常处理

    Delete Object:

    删除Object数据

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)  

bucket_name = "YOUR_BUCKET_NAME"
key_name = "YOUR_KEY_NAME"
try: 
    b=c.get_bucket(bucket_name)
    b.delete_key(key_name)
except:
    pass #异常处理

    List Object

    列举Bucket内的文件或者目录

    from ks3.connection import Connection
from ks3.prefix import Prefix
from ks3.key import Key
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False) 

bucket_name = "YOUR_BUCKET_NAME"
b = c.get_bucket(bucket_name)
keys = b.list(delimiter='/')
for k in keys:
    if isinstance(k, Key):
        print('file:%s' % k.name)
    elif isinstance(k, Prefix):
        print('dir:%s' % k.name)

    列举Bucket内指定前缀的文件

    keys = b.list(prefix="PREFIX")

    Get Object ACL

    获得Object的acl

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False) 

bucket_name = "YOUR_BUCKET_NAME"
key_name = "YOUR_KEY_NAME"  
b = c.get_bucket(bucket_name)
policy = b.get_acl(key_name)
print(policy.to_xml())

    Put Object ACL

    #object policy : 'private' or 'public-read'
b.set_acl("public-read", test_key)

    Upload Part：

    分块上传

    如果你想上传一个大文件，你可以将它分成几个小份，逐个上传，ks3会按照顺序把它们合成一个最终的object。通过SDK进行分块上传能够保障数据传到ks3的正确性，无需另外做数据校验。

    整个过程需要几步来完成，下面的demo程序是通过python的FileChunkIO模块来实现的。所以可能需要首先运行pip install FileChunkIO来安装。

    单线程示例
    import math, os
from ks3.connection import Connection
from filechunkio import FileChunkIO

# Connect to S3
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False)
bucket_name = "YOUR_BUCKET_NAME"
b = c.get_bucket(bucket_name)

# Get file info
source_path = 'SOURCE_FILE_PATH'
source_size = os.stat(source_path).st_size

# Create a multipart upload request
# 此处os.path.basename(source_path)可以替换为需要设置的objectKey
mp = b.initiate_multipart_upload(os.path.basename(source_path), policy="private")
# Use a chunk size of 50 MiB (feel free to change this)
chunk_size = 52428800
chunk_count = int(math.ceil(source_size*1.0 / chunk_size*1.0))

# Send the file parts, using FileChunkIO to create a file-like object
# that points to a certain byte range within the original file. We
# set bytes to never exceed the original file size.
try:
    for i in range(chunk_count):
        offset = chunk_size * i
        bytes = min(chunk_size, source_size - offset)
        with FileChunkIO(source_path, 'r', offset=offset, bytes=bytes) as fp:
            mp.upload_part_from_file(fp, part_num=i + 1)
    ret = mp.complete_upload()
    if ret and ret.status == 200:
        print("上传成功")
except:
    pass  # 异常处理

    说明 以低频存储方式上传，需要在initiate_multipart_upload阶段设置

    //x-kss-storage-class有效值为"STANDARD"、"STANDARD_IA"。"STANDARD"表示标准存储，"STANDARD_IA"表示低频存储，如果不指定，默认为标准存储。
headers = {"x-kss-storage-class": "STANDARD_IA"}
mp = b.initiate_multipart_upload(os.path.basename(source_path), policy="private", headers=headers)
    多线程分块上传示例
    multi_task_result = {}
def multi_put():
    import threadpool
    conn = Connection("AK","SK",host="ENDPOINT")
    b = conn.get_bucket("BUCKETNAME")
    ks3_key=""
    local_path=""
    retry_times=3
    thread_num=5
    f_size = os.stat(local_path).st_size
    mp = b.initiate_multipart_upload(ks3_key)
    if not mp:
        raise RuntimeError("%s init multiupload error" % ks3_key)
    print "%s begin multipart upload,uploadid=%s" % (ks3_key, mp.id)
    chunk_size = 104857600
    chunk_count = int(math.ceil(f_size / float(chunk_size)))
    pool_args_list = []
    try:
        for i in range(chunk_count):
            offset = chunk_size * i
            bs = min(chunk_size, f_size - offset)
            part_num = i + 1
            pool_args_list.append(
                ([mp, local_path, ks3_key, offset, bs, part_num, retry_times], None))
        pool = threadpool.ThreadPool(thread_num)
        requests = threadpool.makeRequests(upload_part_task, pool_args_list)
        [pool.putRequest(req) for req in requests]
        pool.wait()
        global multi_task_result
        if len(multi_task_result) != chunk_count:
            raise RuntimeError(
                "%s part miss,expect=%d,actual=%d" % (ks3_key, chunk_count, len(multi_task_result)))
        for item in multi_task_result.keys():
            if not multi_task_result[item]:
                raise RuntimeError("%s part upload has fail" % ks3_key)
        mp.complete_upload()
        print "%s multipart upload success" % ks3_key
    except BaseException, e:
        print "%s multipart upload fail" % ks3_key
        if mp:
            mp.cancel_upload()
        raise e
def upload_part_task(mp, local_path, ks3_key, offset, chunk_size, part_num, retry_times):
    global multi_task_result
    cur_task_ret = False
    try:
        for i in range(retry_times):
            try:
                with FileChunkIO(local_path, 'rb', offset=offset, bytes=chunk_size) as fp:
                    mp.upload_part_from_file(fp, part_num=part_num)
                cur_task_ret = True
                print "%s part %d upload success" % (ks3_key, part_num)
                break
            except BaseException, e:
                print "%s part %d fail,uploadid=%s,error=%s" % (
                    ks3_key, part_num, mp.id, traceback.format_exc())
                if i + 1 >= retry_times:
                    print "%s part %d upload fail" % (ks3_key, part_num)
                    raise e

    except BaseException, e:
        cur_task_ret = False
    finally:
        multi_task_result[part_num] = cur_task_ret

    获取已上传分块列表

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False) 

bucket_name = "YOUR_BUCKET_NAME" 
b = c.get_bucket(bucket_name)
for p in b.list_multipart_uploads():
    print('uploadId:%s,key:%s' % (p.id, p.key_name))
    for i in p:
        print(i.part_number, i.size, i.etag, i.last_modified)

    Generate URL

    生成下载外链地址

    对私密属性的文件生成下载外链（该链接具有时效性）

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT', is_secure=False, domain_mode=False) 

bucket_name = "YOUR_BUCKET_NAME"
key_name = "YOUR_KEY_NAME" 
b = c.get_bucket(bucket_name)
k = b.get_key(key_name)
if k:
    url = k.generate_url(60) # 60s 后该链接过期
    print(url)

    指定时间过期

    k.generate_url(1492073594,expires_in_absolute=True) # 1492073594为Unix Time

    Modify Object Meta

    修改文件元数据

    class ModifyObjectMeta(unittest.TestCase):
    #srcbucket == dstbucket, srckey == dstkey, x-kss-metadata-directive=REPLACE
    def modify_object_meta(self):
        b = conn.get_bucket(test_bucket)
        b.copy_key('1copykey.jpg', test_bucket, '1copykey.jpg', headers={'content-type':'text/plain', 'x-kss-metadata-directive':'REPLACE'})

    Modify Storage Class

    修改文件存储类型

    class ModifyStorageClass(unittest.TestCase):
    # srcbucket == dstbucket, srckey == dstkey, x-kss-storage-class=new storage class
    def modify_storage_class(self):
        b = conn.get_bucket(test_bucket)
        b.copy_key('1copykey.jpg', test_bucket, '1copykey.jpg', headers={'x-kss-storage-class':'STANDARD_IA'})

    客户端文件加密

    如果您有上传前先加密数据的需求，可以考虑使用加密模式。

    环境要求

    使用加密客户端需要安装pycrypto。

    1. 可选择使用pip install pycrypto进行安装
    2. 无法直接连接pypi服务的话，可选择下载pycrypto安装包手动进行安装

    配置密钥

    您可以使用自己的密钥文件，或者选择调用我们的密钥生成方法。

    1. 如果使用自己的密钥文件：请确保文件中密钥长度为16字节，如果不是16字节，程序将报错
    2. 如果调用密钥生成方法：
    from ks3.encryption import Crypts
Crypts.generate_key('your_path', 'key_name')

    请注意保管好您的key文件，KS3服务端将不会对客户端加密时使用的key文件进行保存，一旦丢失，文件将无法被解密。

    用法示例

    PUT、GET

    对put、get方法加密客户端和普通客户端用法基本一致，不同之处在于初始化Connection对象时需要多填两个参数：local_encrypt和local_key_path。local_encrypt设置为true，local_key_path设置为密钥存放在本地的地址。

    from ks3.connection import Connection
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT',is_secure=False, domain_mode=False, local_encrypt=True, 
        local_key_path="your_key_path")

b = c.get_bucket("your_bucket")
#put
kw = b.new_key("your_key")
ret = kw.set_contents_from_string("some string")
#get
get_key = b.get_key("your_key")
s = get_key.get_contents_as_string()
print("Result:",s)

    分块上传

    注意：加密客户端的分块上传不支持对文件的并行上传！分块上传时必须依次序上传每一块，否则数据将无法解密。

    示例1：使用FileChunkIO进行分块上传。与普通客户端的方法基本一致，和put一样只需在初始化时增加参数。

    c = Connection(ak, sk, host,is_secure=False, domain_mode=False, local_encrypt=True, local_key_path="your_key_path")
#继续普通分块上传方法

    示例2：自己切分文件进行分块上传。除了修改Connection的参数之外，还需在调用upload_part_from_file方法时指定is_last_part的值。

    from ks3.connection import Connection
from filechunkio import FileChunkIO
import math, os
ak = 'YOUR_ACCESS_KEY'
sk = 'YOUR_SECRET_KEY'
c = Connection(ak, sk, host='YOUR_REGION_ENDPOINT',is_secure=False, domain_mode=False, local_encrypt=True, 
       local_key_path="your_key_path")

b = c.get_bucket("your_bucket")
source_path = 'SOURCE_FILE_PATH'
source_size = os.stat(source_path).st_size
mp = bucket.initiate_multipart_upload(os.path.basename(source_path))
# Use a chunk size of 50 MiB (feel free to change this)
chunk_size = 52428800
chunk_count = int(math.ceil(source_size*1.0 / chunk_size*1.0))
try:
    for i in range(chunk_count):
        offset = chunk_size * i
        last = False
        bytes = min(chunk_size, source_size - offset)
        if i +1 == chunk_count:
            last = True
        with FileChunkIO(source_path, 'r', offset=offset,bytes=bytes) as fp:
            mp.upload_part_from_file(fp, part_num=i + 1, is_last_part=last)

    ret = mp.complete_upload()
    if ret and ret.status == 200:
        print("上传成功")
except:
    pass

    注意事项

    • 对于使用加密模式上传的数据，请使用加密模式下（local_encrypt=True）的get方法进行下载。未设置加密模式的get下载下来的这份数据是加密的，无法解读。
    • 加密上传默认进行MD5验证，以防止网络传输过程中的数据损坏。在文件较大的情况下，对加密后文件的MD5计算较为耗时（每500MB约耗时10s），如果不能接受这种额外耗时，可以在调用方法时设置calc_md5=False来关闭MD5校验功能。当然，我们不推荐您关闭MD5校验。
    #PUT上传取消MD5 CHECK：
ret = kw.set_contents_from_string(test_str, calc_encrypt_md5=False)
ret = kw.set_contents_from_filename(test_path, calc_encrypt_md5=False)
#分块上传时取消MD5 CHECK：
mp = b.initiate_multipart_upload(os.path.basename(path), calc_encrypt_md5=False)
    • 用户key的MD5值将作为自定义header放入元数据，方便您后续可能的验证操作。对key的MD5计算方法如下：
    import hashlib
import base64
md5_generator = hashlib.md5()
md5_generator.update("YOUR_KEY".encode())
print(base64.b64encode(md5_generator.hexdigest().encode()))
    • 如果需要在分块上传相关代码中加入重试逻辑，请将开始重试的part_num后的所有块都进行重试。比如上传8块，从第5块开始重试，则需要重新上传的块为5、6、7、8。
    • 对于空文件/空字符串的put请求，即使设置了加密模式也不会加密。

    金山云，开启您的云计算之旅

    免费注册