There's no record validation in v2

[Betsy Luzader]

Validation needs to be added to the recordsets in V1 for each record type, similarly to what was done for records in v1. See api/v1/records for an example.

[Kiall Mac Innes (kiall)]

Relates to https://blueprints.launchpad.net/designate/+spec/validation-cleanup

[OpenStack Infra (hudson-openstack)]

Fix proposed to branch: master
Review: https://review.openstack.org/169044

[OpenStack Infra (hudson-openstack)]

Reviewed: https://review.openstack.org/169044
Committed: https://git.openstack.org/cgit/openstack/designate/commit/?id=707bc639ebe857b281702a1a711a0e961cffffe6
Submitter: Jenkins
Branch: master

commit 707bc639ebe857b281702a1a711a0e961cffffe6
Author: Graham Hayes <email address hidden>
Date: Mon Mar 30 19:31:29 2015 +0100

    Enable Record Data Validation in v2 API

    This enables the validation of data against schemas defined in the
    individual record objects

    * This is a permanent interface, implemented in a temp fashion *

    * Current we convert all the Record() objects to objects of the
      right type, and then validate. If validation is successful we restore
      the generic objects and send them to central.

    * We override the RecordSet validate() command, and add extra logic
      This allows for the v2 API to function, but recursive validation from
      a Domain object will not work, until another solution is found.

    * The way schemas are build up has also changed
    ** Each schema is built on .validate()
    ** There is no embedded "obj://<object_name>" references anymore
    ** The entire schema is added as one blob

    Closes-Bug: #1338256
    Implements: blueprint validation-cleanup
    APIImpact

    Change-Id: I8d1d614a9a9c0c1d3faeb0f98778231278f37bc4