代码之家 › 专栏 › 技术社区 › ÊÉÄ±u

Python告诉ftp传输何时完成

network-programming ftp python

ÊÉÄ±u · 技术社区 · 16 年前

我必须从FTP服务器下载一些文件。看起来很平淡。但是,此服务器的行为方式是,如果文件非常大,当下载表面上完成时,连接将挂起。

我如何在python中使用ftplib优雅地处理这个问题?

python代码示例:

from ftplib import FTP

...

ftp = FTP(host)
ftp.login(login, passwd)
files=ftp.nlst()
ftp.set_debuglevel(2)

for fname in files:
    ret_status = ftp.retrbinary('RETR ' + fname, open(fname, 'wb').write)

调试上述输出:

*cmd* 'TYPE I'
*put* 'TYPE I\r\n'
*get* '200 Type set to I.\r\n'
*resp* '200 Type set to I.'
*cmd* 'PASV'
*put* 'PASV\r\n'
*get* '227 Entering Passive Mode (0,0,0,0,10,52).\r\n'
*resp* '227 Entering Passive Mode (0,0,0,0,10,52).'
*cmd* 'RETR some_file'
*put* 'RETR some_file\r\n'
*get* '125 Data connection already open; Transfer starting.\r\n'
*resp* '125 Data connection already open; Transfer starting.'
[just sits there indefinitely]

当我尝试使用curl-v进行相同的下载时,它看起来是这样的:

* About to connect() to some_server port 21 (#0)
*   Trying some_ip... connected
* Connected to some_server (some_ip) port 21 (#0)
< 220 Microsoft FTP Service
> USER some_user
< 331 Password required for some_user.
> PASS some_password
< 230 User some_user logged in.
> PWD
< 257 "/some_dir" is current directory.
* Entry path is '/some_dir'
> EPSV
* Connect data stream passively
< 500 'EPSV': command not understood
* disabling EPSV usage
> PASV
< 227 Entering Passive Mode (0,0,0,0,11,116).
*   Trying some_ip... connected
* Connecting to some_ip (some_ip) port 2932
> TYPE I
< 200 Type set to I.
> SIZE some_file
< 213 229376897
> RETR some_file
< 125 Data connection already open; Transfer starting.
* Maxdownload = -1
* Getting file with size: 229376897
{ [data not shown]
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100  218M  100  218M    0     0   182k      0  0:20:28  0:20:28 --:--:--     0* FTP response timeout
* control connection looks dead
100  218M  100  218M    0     0   182k      0  0:20:29  0:20:29 --:--:--     0* Connection #0 to host some_server left intact

curl: (28) FTP response timeout
* Closing connection #0

--2009-07-09 11:32:23--  ftp://some_server/some_file
           => `some_file'
Resolving some_server... 0.0.0.0
Connecting to some_server|0.0.0.0|:21... connected.
Logging in as some_user ... Logged in!
==> SYST ... done.    ==> PWD ... done.
==> TYPE I ... done.  ==> CWD not needed.
==> SIZE some_file ... 229376897
==> PASV ... done.    ==> RETR some_file ... done.
Length: 229376897 (219M)

100%[==========================================================>] 229,376,897  387K/s   in 18m 54s 

2009-07-09 11:51:17 (198 KB/s) - Control connection closed.
Retrying.

--2009-07-09 12:06:18--  ftp://some_server/some_file
  (try: 2) => `some_file'
Connecting to some_server|0.0.0.0|:21... connected.
Logging in as some_user ... Logged in!
==> SYST ... done.    ==> PWD ... done.
==> TYPE I ... done.  ==> CWD not needed.
==> SIZE some_file ... 229376897
==> PASV ... done.    ==> REST 229376897 ... done.    
==> RETR some_file ... done.
Length: 229376897 (219M), 0 (0) remaining

100%[+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++] 229,376,897 --.-K/s   in 0s      

2009-07-09 12:06:18 (0.00 B/s) - `some_file' saved [229376897]

2 回复 | 直到 16 年前

John Fouhy 16 年前

我认为一些调试可能会有用。你能把下面的类折叠到你的代码中吗?(我自己没有这样做,因为我知道这个版本有效,不想冒出错的风险。你应该能够把类放在文件的顶部,用我在#loop body之后写的东西替换循环的主体)

class CounterFile():
    def __init__(self, file, maxsize):
        self.file = file
        self.count = 0
        self.maxsize = maxsize

    def write(self, bytes):
        self.count += len(bytes)
        print "total %d bytes / %d"%(self.count, self.maxsize)
        if self.count == self.maxsize:
            print "   Should be complete"
        self.file.write(bytes)


from ftplib import FTP
ftp = FTP('ftp.gimp.org')
ftp.login('ftp', 'thouis@gmail.com')
ftp.set_debuglevel(2)

ftp.cwd('/pub/gimp/v2.6/')
fname = 'gimp-2.6.2.tar.bz2'

# LOOP BODY
sz = ftp.size(fname)
if sz is None:
    print "Could not get size!"
    sz = 0
ret_status = ftp.retrbinary('RETR ' + fname, CounterFile(open(fname, 'wb'), sz).write)

user227667 user227667 16 年前

我从未使用过ftplib,但也许你可以这样做:

获取所需文件的名称和大小。
启动一个新的守护线程来下载文件。
在主线程中,每隔几秒钟检查一次磁盘上的文件大小是否等于目标大小。
当它发生时,等待几秒钟,让连接有机会很好地关闭,然后退出程序。