add full reference to CIDR_OR_ID_MATCHER

[jntullo]

Without the full reference to CIDR_OR_ID_MATCHER, tagging API is returning uninitialized constant Api::Subcollections::Tags::CID_OR_ID_MATCHER errors (as was referenced in #13581).

cc: @imtayadeway
@miq-bot add_label bug, api
@miq-bot assign @abellotti

[miq-bot]

Checked commit jntullo@90406ba with ruby 2.2.6, rubocop 0.46.0, and haml-lint 0.19.0
1 file checked, 0 offenses detected
Everything looks good. 🍰

[abellotti]

Can you add one test in this PR that exercises the fix ? Thanks.

[imtayadeway]

@jntullo I'm curious if the other places that reference this constant in the API are broken? Also, did you find out where/why this regressed?

[isimluk]

CIDR_OR_ID_MATCHER is also present in CompressedIds. The compressedIds seem to be still included in the same final class.

Perhaps it is now included in different order than it used to?

[Fryguy]

@imtayadeway Has the PR for the move away from ApplicationRecord been merged yet? If so, could that be responsible?

[imtayadeway]

@Fryguy it hasn't....probably need to go over that carefully before that gets merged. I suspect this is either a) ordering of includes in Api::BaseController, which may have been changed a while back when doing some housekeeping in that file, or b) some weirdness to do with constant lookup inside mixed-in modules, the structure of which changed when we broke out into separate controllers. I'm curious why the references in app/controllers/api/base_controller/parser.rb and
app/controllers/api/subcollections/policies.rb don't appear to have broken (or don't have coverage?). @jntullo LMK if you want me to look into this....I am putting myself as the leading suspect 😉

[jntullo]

@imtayadeway investigating now 😄

[jntullo]

@imtayadeway I believe it has to do with the reordering, and I am worried that some things don't have coverage (going to work on that). When include CompressedIds is added to Api::Subcollections::Tags, it also solves the problem

[imtayadeway]

@jntullo could you add a failing test for this, and also incorporate the changes to the order of includes as discussed above? Thanks!

EDIT: I believe there are several places where we are unnecessarily including that module into different parts of the API controller because of reasons above. I believe they could also be removed as part of this PR.

[jntullo]

@imtayadeway can do, but was also waiting for feedback on #13581 where I have refactored this line out.

[imtayadeway]

@jntullo I spent some time digging into this today and made a PR based on my findings in #13671. LMK what you think!

[abellotti]

Closing in preference of merged #13671

[miq-bot]

This pull request is not mergeable. Please rebase and repush.