[core] Web Link Attribute parsing problems

[Jim Schaad <[e-mail]>]

I have been going through my code base looking to get
draft-ietf-core-links-json implemented as well as making sure that my
implementation of RFC 6690 (CoRE Link Format).  At this point in time, I
cannot see anyway to future proof my implementation of this code because of
how RFC 6690 decided to address a compression optimization.

My code currently does not do a good job of deciding which attributes are to
be split when they are space separated (i.e. 'rt') and which should be kept
as a single string (i.e. 'title').  I can easily get a list of the things
that should be treated in which direction right now, however going forward
into the future that will not be the case.  When the code sees a new link,
it will not know what the correct processing is in the event that a space
occurs in the value field.  

As an example of what I am looking at.  I get a new WebLink with the
attribute of 

<googoo>ct=0;bletch="foo bar"

I then get a discovery query

get  /.well-known/core?bletch="foo" 

is "googoo" supposed to be a match or not?  If it is a single string, ala
'title', then it is not a match.  If it is space separated attibutes, ala
'rt', then it is a match.

Is there a recommended way to proceed?

Jim