Brig Polysemy: Port UserPendingActivationStore to polysemy

[isovector]

This PR moves Brig.Data.UserPendingActivation into a polysemy effect Brig.Sem.UserPendingActivationStore.

Checklist

• The PR Title explains the impact of the change.
• The PR description provides context as to why the change should occur and what the code contributes to that effect. This could also be a link to a JIRA ticket or a Github issue, if there is one.
• changelog.d contains the following bits of information (details):
  • A file with the changelog entry in one or more suitable sub-sections. The sub-sections are marked by directories inside changelog.d.

[isovector]

@mdimjasevic @pcapriotti please take a look!

[pcapriotti]

There is a similar abstraction for paging in galley. See Galley.Effects.Paging, and Galley.Cassandra.Paging. These are not using dynamic typing, but instead require a type variable for the paging implementation. I'm not sure I like the dynamic typing trick with unsafeCoerce. Honestly, keeping the type argument around is not that much trouble. I think we should choose one of the two approaches and use it in all cases.

[isovector]

Thanks for the pointers to Galley.Effects.Paging --- I like that approach more too. I'll change this one.

[isovector]

Depends on #2648.

[mdimjasevic]

Looks good!

[mdimjasevic]

@isovector , the branch fails to build Galley integration tests:

#10 299.1 [33 of 33] Compiling Main             ( test/integration/Main.hs, /wire-server/dist-newstyle/build/x86_64-linux/ghc-8.10.7/brig-2.0/x/brig-integration/noopt/build/brig-integration/brig-integration-tmp/Main.o, /wire-server/dist-newstyle/build/x86_64-linux/ghc-8.10.7/brig-2.0/x/brig-integration/noopt/build/brig-integration/brig-integration-tmp/Main.dyn_o )

#10 299.2 

#10 299.2 test/integration/Main.hs:160:64: error:

#10 299.2     • Overlapping instances for Polysemy.Internal.Union.Member

#10 299.2                                   (Brig.Sem.UserPendingActivationStore.UserPendingActivationStore

#10 299.2                                      p0)

#10 299.2                                   '[Brig.Sem.UserPendingActivationStore.UserPendingActivationStore

#10 299.2                                       polysemy-wire-zoo-0.1.0:Wire.Sem.Paging.Cassandra.InternalPaging,

#10 299.2                                     polysemy-wire-zoo-0.1.0:Wire.Sem.Now.Now,

#10 299.2                                     Brig.Sem.CodeStore.CodeStore,

#10 299.2                                     Polysemy.Embed.Type.Embed

#10 299.2                                       cql-io-1.1.1:Database.CQL.IO.Client.Client,

#10 299.2                                     Polysemy.Embed.Type.Embed IO, Polysemy.Final.Final IO]

#10 299.2         arising from a use of ‘sitemap’

#10 299.2       Matching instances:

#10 299.2         two instances involving out-of-scope types

#10 299.2         (use -fprint-potential-instances to see them all)

#10 299.2       (The choice depends on the instantiation of ‘p0’

#10 299.2        To pick the first instance above, use IncoherentInstances

#10 299.2        when compiling the other instance declarations)

#10 299.2     • In the second argument of ‘($)’, namely

#10 299.2         ‘sitemap @BrigCanonicalEffects’

#10 299.2       In the third argument of ‘assertEqual’, namely

#10 299.2         ‘(pathsConsistencyCheck . treeToPaths . compile

#10 299.2             $ sitemap @BrigCanonicalEffects)’

#10 299.2       In the second argument of ‘($)’, namely

#10 299.2         ‘assertEqual

#10 299.2            "inconcistent sitemap" mempty

#10 299.2            (pathsConsistencyCheck . treeToPaths . compile

#10 299.2               $ sitemap @BrigCanonicalEffects)’

#10 299.2     |

#10 299.2 160 |               (pathsConsistencyCheck . treeToPaths . compile $ Brig.API.sitemap @BrigCanonicalEffects),

#10 299.2     |                                                                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

[isovector]

@mdimjasevic should be fixed now!