通过 Python tcp 套接字传输标题为阿拉伯语的文件时出现 ValueError

Question

我正在构建一个套接字程序以在两台计算机之间传输文件。带有英文标题的文件已成功传输，但是当我尝试发送阿拉伯文标题文件（例如وثيق.docx）时，我得到一长串 ValueErrors，开头为：

invalid literal for int() with base 2: b'.docx000000000000000000010000001'

invalid literal for int() with base 2: b'10001PK\x03\x04\x14\x00\x08\x08\x08\x00\xe0'

我的代码是：服务器：

import socket

serversock = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
host = 'localhost'
port = 9000
serversock.bind((host,port))
filename = ""
serversock.listen(10)
print("Waiting for a connection.....")

clientsocket, addr = serversock.accept()
print("Got a connection from %s" % str(addr))

while True:
    try:
        size = clientsocket.recv(16) # Note that you limit your filename length to 255 bytes.
        if not size:
            clientsocket, addr = serversock.accept()
            print("Got a connection from %s" % str(addr))
            continue
        size = int(size, 2)
        print('SIZE', size)
        filename = clientsocket.recv(size)
        print('filename', filename)
        filesize = clientsocket.recv(32)
        print('FILESIZE', filesize, 'TYPE', type(filesize))
        filesize = int(filesize, 2) ##########
        file_to_write = open(filename, 'wb')
        chunksize = 4096
        while filesize > 0:
            if filesize < chunksize:
                chunksize = filesize
            data = clientsocket.recv(chunksize)
            file_to_write.write(data)
            filesize -= len(data)

        file_to_write.close()
        print('File (%s) received successfully' % filename.decode('utf-8'))
    except ValueError as verr:
        print(verr)
        #continue
    except FileNotFoundError:
        print('FileNotFoundError')
        #continue
serversock.close()

客户：

进口

 socket
import os
from file_walker import files_to_transmit

def transmit(host, port):
    s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
    try:
        s.connect((host, port))
        directory = files_to_transmit()
        for file in directory:
            filename = file
            size = len(filename.split('/')[-1]) # split() to get bare file name to transmit to server
            size = bin(size)[2:].zfill(16) # encode filename size as 16 bit binary
            s.send(size.encode('utf-8'))
            s.send(filename.split('/')[-1].encode('utf-8')) # split() to get bare file name

            filename = file
            filesize = os.path.getsize(filename)
            filesize = bin(filesize)[2:].zfill(32) # encode filesize as 32 bit binary
            s.send(filesize.encode('utf-8'))

            file_to_send = open(filename, 'rb')

            l = file_to_send.read()
            s.sendall(l)
            file_to_send.close()
            print('File Sent')

        s.close()
    except ConnectionRefusedError:
        print('ConnectionRefusedError: Server may not be running')

    except ValueError as e:
        print(e)

transmit('localhost', 9000)

这里有什么问题？请帮忙。

Answer 1

根据我的经验，遇到了类似的问题。我的建议是使用一种算法来编码和解码像 base64 这样的文件。

在你程序的某个地方，你正在读取文件以将他保存在内存中并在第二步发送。好吧，而不是只读编码也

import base64

with open("yourfile.ext", "rb") as image_file:
    encoded_string = base64.b64encode(image_file.read())

显然在另一边你需要解码并写入磁盘。您还可以对文件的标题进行编码和解码以避免 ValueError

Answer 2

您发送标题的unicode 字符大小，然后尝试以utf-8 编码发送。例如你的例子：

title = 'وثيقة.docx'
print(len(title), len(title.encode('utf8')))

给予

10 15

对方只会用10个字节作为文件名，剩下的5个字节作为文件大小的开始。并且会因为 .docx 不是二进制数的开头而窒息。会发生什么...

修复很容易，在计算长度之前构建字节字符串：

    ...
    for file in directory:
        filename = file.split('/')[-1].encode('utf_8') # split() to get bare file name
        size = len(filename)

        size = bin(size)[2:].zfill(16) # encode filename size as 16 bit binary
        s.send(size.encode('utf-8'))
        s.send(filename) 
        ...

通过 Python tcp 套接字传输标题为阿拉伯语的文件时出现 ValueError

ValueError when transmitting a file titled in Arabic over a Python tcp socket

python

sockets

arabic

valueerror

non-english