BIND 10 #2108: redefine in-memory zone load()

[BIND 10 Development]

#2108: redefine in-memory zone load()
-------------------------------------+-------------------------------------
            Reporter:  jinmei        |                        Owner:
                Type:  task          |                       Status:  new
            Priority:  medium        |                    Milestone:  Next-
           Component:  data source   |  Sprint-Proposed
           Sensitive:  0             |                     Keywords:
         Sub-Project:  DNS           |              Defect Severity:  N/A
Estimated Difficulty:  0             |  Feature Depending on Ticket:
         Total Hours:  0             |  scalable inmemory
                                     |          Add Hours to Ticket:  0
                                     |                    Internal?:  0
-------------------------------------+-------------------------------------
 First, we should move it from `ZoneFinder` (finder is supposed to be
 read-only accessor, so it's been very awkward that it has a "load"
 method).  This will eventually need to be changed again with scalable
 loading, but for now my suggestion is to move it to
 `InMemoryDataSourceClient`.

 I also suggest making sure both an RRset and its RRSIG (when signed)
 are given at the time of constructing RdataSet.  Due to the encoding
 restriction, if this doesn't hold we need to destroy the old data and
 reconstruct everything.

 The load function/method will insert a tree node (if necessary),
 allocate/construct RdataSet, and add it to the node.  Except for the
 previous paragraph, this should be essentially an internal
 refactoring.

 The current additional-section shorcut won't be necessary; that part
 shouldn't be ported.

 Depends on #2107.

-- 
Ticket URL: <http://bind10.isc.org/ticket/2108>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development