Ensure Content-Length header instead of TE: chunked

[malud]

Since we know the length of the given body bytes for each direction we should ensure sending the Content-Length header instead of sending those bytes with Transfer-Encoding: chunked.

Depends on the io.Reader. StringReader for example triggers an implicit set...

[filex]

Sending T/E chunked also leads to problems with the oauth2 function when used to obtain a token from the MS Graph API (https://login.microsoftonline.com): The API responds with a 404 for the token endpoint. Seems like a content-length is mandatory.

we should at least switch to content-length for all (upstream) requests. because chunked encoding is rather unusual for uploads.

for responses, I wouldn't mind to use chunked encoding.