基于七牛 Python SDK 写的一个同步脚本(支持批量下载)

[heqingpan]

代码不是很多就贴出来吧

#!/usr/bin/env python
#-*- coding:utf-8 -*-
# 
# AUTHOR = "heqingpan"
# AUTHOR_EMAIL = "[e-mail]"
# URL = "http://git.oschina.net/hqp/qiniu_sync"

import qiniu
from qiniu import Auth
from qiniu import BucketManager
import os
import re
import urllib2
import sys

access_key = ''
secret_key = ''
bucket_name = ''
bucket_domain = ''

q = Auth (access_key, secret_key )
bucket = BucketManager (q )
basedir=os.path.realpath (os.path.dirname (__file__))
#同步目录
#basedir=""
filename=__file__
ignore_paths=[filename,"{0}c".format (filename )]
ignore_names=[".DS_Store",".git",".gitignore"]
charset="utf8"
diff_time=2*60


def list_all (bucket_name, bucket=None, prefix="", limit=100 ):
    rlist=[]
    if bucket is None:
        bucket = BucketManager (q )
    marker = None
    eof = False
    while eof is False:
        ret, eof, info = bucket.list (bucket_name, prefix=prefix, marker=marker, limit=limit )
        marker = ret.get ('marker', None )
        for item in ret['items']:
            rlist.append (item["key"])
    if eof is not True:
        # 错误处理
        #print "error"
        pass
    return rlist

def get_files (basedir="",fix="",rlist=None,ignore_paths=[],ignore_names=[]):
    if rlist is None:
        rlist=[]
    for subfile in os.listdir (basedir ):
        temp_path=os.path.join (basedir,subfile )
        tp=os.path.join (fix,subfile )
        if tp in ignore_names:
            continue
        if tp in ignore_paths:
            continue
        if os.path.isfile (temp_path ):
            rlist.append (tp )
        elif os.path.isdir (temp_path ):
            get_files (temp_path,tp,rlist,ignore_paths,ignore_names )
    return rlist

def get_valid_key_files (subdir=""):
    basedir=subdir or basedir
    files = get_files (basedir=basedir,ignore_paths=ignore_paths,ignore_names=ignore_names )
    return map (lambda f:(f.replace ("\\","/"),f ),files )


def sync ():
    qn_keys=list_all (bucket_name,bucket )
    qn_set=set (qn_keys )
    l_key_files=get_valid_key_files (basedir )
    k2f={}
    update_keys=[]
    u_count=500
    u_index=0
    for k,f in l_key_files:
        k2f[k]=f
        str_k=k
        if isinstance (k,str ):
            k=k.decode (charset )
        if k in qn_set:
            update_keys.append (str_k )
            u_index+=1
            if u_index > u_count:
                u_index-=u_count
                update_file (k2f,update_keys )
                update_keys=[]
        else:
            # upload
            upload_file (k,os.path.join (basedir,f ))
    if update_keys:
        update_file (k2f,update_keys )
    print "sync end"

def update_file (k2f,ulist ):
    ops=qiniu.build_batch_stat (bucket_name,ulist )
    rets,infos = bucket.batch (ops )
    for i in xrange (len (ulist )):
        k=ulist[i]
        f=k2f.get (k )
        ret=rets[i]["data"]
        size=ret.get ("fsize",None )
        put_time = int (ret.get ("putTime")/10000000 )
        local_size=os.path.getsize (f )
        local_time=int (os.path.getatime (f ))
        if local_size==size:
            continue
        if put_time >= local_time - diff_time:
            # is new
            continue
        # update
        upload_file (k,os.path.join (basedir,f ))

def upload_file (key,localfile ):
    print "upload_file:"
    print key
    token = q.upload_token (bucket_name, key )
    mime_type = get_mime_type (localfile )
    params = {'x:a': 'a'}
    progress_handler = lambda progress, total: progress
    ret, info = qiniu.put_file (token, key, localfile, params, mime_type, progress_handler=progress_handler )

def get_mime_type (path ):
    mime_type = "text/plain"
    return mime_type

def down_file (key,basedir="",is_private=1,expires=3600 ):
    if isinstance (key,unicode ):
        key=key.encode (charset )
    url = 'http://%s/%s' % (bucket_domain, key )
    if is_private:
        url=q.private_download_url (url, expires=expires )
    c=urllib2.urlopen (url )
    fpath=key.replace ("/",os.sep )
    savepath=os.path.join (basedir,fpath )
    dir_=os.path.dirname (savepath )
    if not os.path.isdir (dir_):
        os.makedirs (dir_)
    elif os.path.isfile (savepath ):
        os.remove (savepath )
    f = file (savepath, 'wb')
    f.write (c.read ())
    f.close ()

def down_all (prefix=""):
    import traceback
    for key in list_all (bucket_name,bucket,prefix=prefix ):
        try:
            down_file (key,basedir=basedir )
            print "down:\t"+key
        except:
            print "error down:\t"+key
            print traceback.format_exc ()
    print "down end"


def main ():
    if len (sys.argv )>1:
        if sys.argv[1]=="down":
            prefix=len (sys.argv )>2 and sys.argv[2] or ""
            down_all (prefix=prefix )
            return
    sync ()

if __name__=="__main__":
    main ()

使用方式

• 安装七牛 Python SDK
  pip install qiniu

• 填写脚本文件(qiniusync.py )的配置信息
  
access_key = ''
secret_key = ''
bucket_name = ''
bucket_domain = ''


  注册后可以拿到对应的信息

• 将脚本文件(qiniusync.py )拷贝到待同步根目录

• 运行脚本
  
python qiniusync.py
python qiniusync.py down
python qiniusync.py down [文件路径前缀]


脚本也有发到Git@OSC和github上，有兴趣的同学可以去看看。

[Ouyangan]

谢谢