Re: [Cbor] Question about CBOR padding

[Thiago Macieira <[e-mail]>]

On Tuesday, 6 November 2018 13:58:18 PST Matt Vollrath wrote:
> Hello list,
> 
> I'm finding CBOR extremely useful for moving large binary blobs
> (specifically: ROS point clouds) into browsers via WebSocket. While
> hacking on a CBOR integration, I ran into something that the spec
> doesn't seem to cover.
> 
> How should CBOR data be padded, if I were to do something weird like
> pack it into a PNG?  Believe it or not, the previous implementation
> packed JSON into PNG with \n padding to complete the last pixel.  I
> didn't see anything in the spec about padding, except the mention that
> UBJSON has an explicit noop code.

CBOR has no need for padding and it does not have any byte sequence that can 
fill in for padding.

Some CBOR decoders will simply stop when they've finished reading the one 
element, ignoring anything remaining in the buffer. If you know this is your 
case, then simply append anything you want until you reach the boundary you 
ant.

Other decoders will report garbage at the end. If this is a possibility for 
your file format, you should design your container so you know where the 
actual CBOR data ends and so you can feed the decoder no padding.

-- 
Thiago Macieira - thiago.macieira (AT) intel.com
  Software Architect - Intel Open Source Technology Center