On Tuesday, June 4, 2013 10:31:37 AM UTC-7, Patrick Carlisle wrote: > Check spec/integration/network/authconfig_spec.rb, >
Thanks Patrick, clearly a case of RTFspec. I have done that now and understand a bit more. Here are some observations so far: * :deny is important. We should fix the doc that says "A deny directive is syntactically permitted, but has no effect." here: http://docs.puppetlabs.com/guides/rest_auth_conf.html * Functionally the relative ordering of (:domain, :exact) vs (:opaque, :exact) looks unimportant since these two sets are disjoint. Anything that can be (:domain, :exact) will be. I believe we could remove the test that fails when "www.example.com" !< "a_name". OTOH it might be nice to explicitly collate :domain before :opaque by modifying <=>. It might simplify maintainability/testing. Something to think about... * :regex is:inexact with a nil length. This contraticts the pattern of other inexact declarations and leads to an undefined method '<' / ArgumentError for "*.example.com" <=> "/example/". When fixing this should :regex sort before or after :domain? Intuitively (:domain, :exact) < (:domain, :inexact) < (:regex, :inexact) seems "better" than (:domain, :exact) < (:regex, :inexact) < (:domain, :inexact) right? I think there are a few more issues lurking in there, but that's it for tonight... -- You received this message because you are subscribed to the Google Groups "Puppet Developers" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/puppet-dev?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
