Search code examples
pythonpython-requestsputhttplib2

Python requests PUT method creates a zero byte file


I am trying to upload a file with PUT using the requests module in python.

my code is this:

with open(file, 'rb') as payload:
    r = requests.put(url, data=payload, auth=('username', 'password'))

The file is created, I am getting a response 200, but it has 0 bytes. If I am not doing something wrong I suspect that I have met the bug here.

Is this the case? If yes is there any workaround that I can try?

I've tried also the same with the httplib2 library

with open(file, 'rb') as payload:
    h = httplib2.Http(".cache")
    h.add_credentials('user', 'pass')
    resp, content = h.request(url, "PUT", body=payload)

But the request stays hanging for ever (again a 0 size file is created). Could it also be the same problem with the requests module?

[EDIT] Some extra info.

The service receiving the PUT is running on a ESXi hypervisor. It has a function where if you make an authed PUT request it stores the file of the request in /tmp. The server side is working (tested it with a perl script which does the same job and also with curl).

The file uploaded is a .tgz file which lies on my local filesystem and the url is in the form of "http://esx-server/tmp/file.tgz".


Solution

  • So it seems that there is some kind of bug indeed.

    My solution was to use the urllib2. It isn't as "clean" as requests. But still better than what I thought it would be.

    My working code now is:

    import urllib2
    from base64 import b64encode
    
    with open(source, 'rb') as file:
        data = file.read()
    request = urllib2.Request(url)
    request.add_data(data)
    request.add_header('Authorization', 'Basic ' + b64encode(username + ':' + password))
    request.get_method = lambda: "PUT"
    r = urllib2.urlopen(request)