encode Vs struct

[kjm]

I'm trying to get my head around the difference between

>>> struct.pack('4s', 'ping')
b'ping'

and

>>> 'ping'.encode()
b'ping'

is it just an a case of doing the same thing 2 different ways or is there some subtle difference I'm missing? I'm trying to figure out if there is a way to reduce the payload size of an epoch (10digits) sent over rawlora as per https://forum.pycom.io/topic/5555/lopy4-to-lopy4-over-lora-raw-pack-string/5.

[robert-hh]

@kjm I epoch is a number < 2^32, you can pack that as binary value using struct.pack("L", epoch). That will result in a 4 bytes object of 4 bytes length.
Larger numbers may have to be split.

[kjm]

@robert-hh So if I want to pack two 0-15 range variables into a byte, is there a better way than

import struct
slot=15; rtu=14
slotrtu=slot+rtu*16 
packet=struct.pack('b', slotrtu)
print(packet)

to do it?

[robert-hh]

@kjm That's a good way doing it. As a matter of preference & PEP8 coding style I would write it as:

import struct
slot = 15
rtu = 14
slotrtu = (slot & 0x0f)  | (rtu << 4) 
packet = struct.pack('B', slotrtu)

Using binary operators instead of arithmetic ones results in the same value but expresses more clearly the intention. And using slot & 0x0f is more a 'belt and straps' attitude.
The only important change is using B instead of b, indicating that the number is unsigned, which however mostly matters on unpack.