Re: [yang-doctors] Definition for base64

[Andy Bierman <[e-mail]>]

On Fri, Aug 17, 2018 at 12:03 PM, Juergen Schoenwaelder <
[e-mail]> wrote:

> On Fri, Aug 17, 2018 at 10:46:55AM -0700, Mahesh Jethanandani wrote:
> > I have a protocol (Babel) that defines router-id as a base64 string of 8
> octets. I could define it as a string, but isn’t it the case that base64
> includes only a set of printable characters defined by the regex below? Not
> all printable characters are included in the set.
>
> RFC 6126 section 4.1.2 says:
>
> 4.1.2.  Router-Id
>
>    A router-id is an arbitrary 8-octet value.  Router-ids SHOULD be
>    assigned in modified EUI-64 format [ADDRARCH].
>
> Where does it say that it is a string on 8 octets base64? Note, if you
> use binary with length of 8, you actually get a base64 encoding on the
> wire. But then there might be another doument that says the router-id
>

IMO the correct YANG data model would use type binary { length 8; }


Andy

is to rendered as base64 but then the base64 string obviously is
> longer than 8 octets. If you work off the information model, then I
> might understand that you are confused because
> draft-ietf-babel-information-model-03 says:
>
>    base64      An opaque array of bytes.
>
> What is that? Does this mean a base64 encoded opaque array of bytes?
> Or something else? I would check how babel people and implementors
> prefer to represent router-ids in textual format. Do they really
> want trailing characters?
>
>   import base64
>   print(base64.b64encode(bytes(8)))
>
> Gives me:
>
>   b'AAAAAAAAAAA='
>
> /js
>
> --
> Juergen Schoenwaelder           Jacobs University Bremen gGmbH
> Phone: [phone]         Campus Ring 1 | 28759 Bremen | Germany
> Fax:   [phone]         <https://www.jacobs-university.de/>
>