The translate method is not supported on MicroPython, neither is maketrans.

Good to know. Shouldn't we resolve it here then? As mentioned, it's a simply a copy.

• This whole b64encode function is superfluous. The altchars parameter is never used by the rest of the module and the only other actual functionality left then is being a wrapper for ubinascii.b2a_base64. This can be done in encode_basic_auth directly.
• Stripping the trailing newline is better done with .rstrip(b'\n').

Very good remark. Seems like I wasn't paying much attention to the functionality of b64encode. Simply knew I needed it ;-)

Several problems with this function too:

• Should take user and password as separate params, not as a single auth param, to increase clarity.
• Decoding the return value of b64encode is not necessary, since the header dict should contain byte strings anyway.
• Non-pep-8-compliant whitespace.

Suggested version:

def encode_basic_auth(user, password):
 from ubinascii import b2a_base64
 auth_encoded = b2a_base64(b"%s:%s" % (user, password)).rstrip(b'\n')
 return {b'Authorization': b'Basic %s' % auth_encoded}

Using the % string formatting operator here allows for user and password to be passed as either str or bytes, which is a bit unclean and wouldn't work in CPython, but works in MicroPython.

• An auth parameter needs to be added to the request function signature (defaulting to None).

• If following my suggested changes for encode_basic_auth, this should be changed to the following then:

  if auth:
   headers.update(encode_basic_auth(auth[0], auth[1]))
  

  (Also changing if auth is not None to if auth so that passing an empty tuple works as well.)

It might be also a good idea to allow passing a callable for auth (but this would be an extension not supported by the original requests.request().

if auth:
 headers.update(auth() if callable(auth) else encode_basic_auth(auth[0], auth[1]))

Obviously that mistake had to happen to me 🤦
Seems like I messed up when developing...

... this would be an extension not supported by the original requests.request().

How are we dealing with that? Do we actually want to extend the original or simply replicate it's functionality? Is there any guidelines for that?