RAN function definitions are currently defined as string (UTF-8), but should be bytes as the original ASN1 type in E2AP is OCTET STRING and could very well include \0 characters.

https://github.com/o-ran-sc/ric-plt-nodeb-rnib/blob/master/entities/ran_function.proto

No sense in starting complicated things like base64 or some other encoding just to make it work as string.

Note that some other values, e.g., PLMN Identity could still be string even if originally defined as OCTET STRING (chapter 9.2.2.4 in TS 38.423) as for them they are only valid values that are strings. So we only need to define how to do the translation, e.g., in case of PLMN Identity what is used as "filler" in case of 2 digit MNCs.