Python requests module, need to encode the body, why?

Question

I was struggling with sending a request to a web service using the requests module. I'm using Python 3.5.1-32 bit on Windows.

Whenever I request something that contains, for example, a latin accented character such as "á" then I need to encode my string variable explicitly as utf-8. Can someone explain why I need to do this?

Here is my code example:

import requests

headers = { "content-type" : "text/xml;charset=UTF-8" }
url = '....'
body = 'á'
requests.post(url, data=body, headers=headers)

This doesn't work, the server receives an escaped version of the character, but if I add this before the request:

body = body.encode(encoding='utf-8')

It works.

Can somebody explain why this is necessary? My Python-Fu is such that it took me a good while to figure this out.

Answer 1

requests doc says:

:param data: (optional) Dictionary, bytes, or file-like object to send in the body of the :class:`Request`.

In Python 3 body='á' is a unicode string. With the method encode() you are converting it to bytes.

For a unicode string, requests understands it as file content and encodes with octet-stream/base64. The false content will be transferred.

P/S: It's only for Python 3. str of Python 2 isn't like str in Python 3