Python Adds An Extra CR At The End Of The Received Lines

Question

A Java application sends an XML to a Python application. They are both on the same machine. When I open the received file I can see extra lines (because of extra CRs). What could be the reason for this?

This is the receiver:

f = open('c:/python/python.xml', 'w')
while 1:
    print("xxx")        
    data = socket.recv(recv_frame)
    remain_byte = remain_byte - len(data)
    print(remain_byte)
    f.write(data)
    if (something):
        break

This is the sender:

    while ((bytesRead = file_inp.read(buffer)) > 0) {
        output_local.write(buffer, 0, bytesRead);
    }

This is the original file:

<root><CR><LF>
    <SONG><CR><LF>
        <ARTIST>Coldplay</ARTIST><CR><LF>
    </SONG><CR><LF>
</root><CR><LF>

This is the received:

<root><CR>
<CR><LF>
    <SONG><CR>
<CR><LF>
        <ARTIST>Coldplay</ARTIST><CR>
<CR><LF>
    </SONG><CR>
<CR><LF>
</root><CR>
<CR><LF>

Halfwitted

What operating system are the applications running on? Is it windows? – Velez

Are you looking at the file "c:/python/python.xml"? On windows, some of the C libraries do carriage-return/line-feed processing to make UNIX files compatible with windows files. Unix only uses one character for line separation (\n), but Windows uses cr+lf. So it is just-about possible that two sets of processing (Java and Python) are expanding one line separator to two, – Velez

Yes, the file I'm checking is correct. – Halfwitted

Change filemode from 'w' to 'wb'. – Grant

IIRC use "wb" when opening files for write, and that suppresses Python doing line separator expansion. – Velez

wb solved the issue! great, thanks! – Halfwitted

Answer 1

Change filemode from 'w' to 'wb', otherwise Python converts any newlines ('\n') into the platform specific representation ('\r\n' for Windows). Binary mode suppresses this conversion.

Grant