fix: weird typing issue with Postgres.

[jimsynz]

This partially reverts the changes in 3cad640.

I don't actually know what any of this code does, but it seems that the missing Ash.Query.Function.Type at the root of the expression causes weird typing issues. The customer app does have a lot of their own types added to Postgrex using Postgrex.Types.define/3.

Here's what the customer was seeing:

     ** (Ash.Error.Unknown)
     Bread Crumbs:
       > Error returned from: App.Domain.Resource.action

     Unknown Error

     * ** (DBConnection.EncodeError) Postgrex expected an integer in -9223372036854775808..9223372036854775807, got "update_failed". Please make sure the value you are passing matches the definition in your table or in your query or convert the value accordingly.
       (app 0.1.0) deps/postgrex/lib/postgrex/type_module.ex:1084: App.Core.PostgrexTypes.encode_list/3
       (postgrex 0.21.1) lib/postgrex/extensions/array.ex:88: Postgrex.Extensions.Array.encode/4
       (postgrex 0.21.1) lib/postgrex/extensions/array.ex:58: Postgrex.Extensions.Array.encode/3
       (app 0.1.0) deps/postgrex/lib/postgrex/type_module.ex:1084: App.Core.PostgrexTypes.encode_params/3
       (postgrex 0.21.1) lib/postgrex/query.ex:75: DBConnection.Query.Postgrex.Query.encode/3
       (db_connection 2.8.1) lib/db_connection.ex:1446: DBConnection.encode/5
       (db_connection 2.8.1) lib/db_connection.ex:1546: DBConnection.run_prepare_execute/5
       (db_connection 2.8.1) lib/db_connection.ex:769: DBConnection.parsed_prepare_execute/5
       (db_connection 2.8.1) lib/db_connection.ex:761: DBConnection.prepare_execute/4
       (ecto_sql 3.13.2) lib/ecto/adapters/postgres/connection.ex:102: Ecto.Adapters.Postgres.Connection.prepare_execute/5
       (ecto_sql 3.13.2) lib/ecto/adapters/sql.ex:1004: Ecto.Adapters.SQL.execute!/5
       (ecto_sql 3.13.2) lib/ecto/adapters/sql.ex:996: Ecto.Adapters.SQL.execute/6
       (ecto 3.13.3) lib/ecto/repo/queryable.ex:241: Ecto.Repo.Queryable.execute/4
       (ash_postgres 2.6.20) lib/data_layer.ex:1581: AshPostgres.DataLayer.update_query/4
       (ash 3.5.43) lib/ash/actions/update/bulk.ex:608: Ash.Actions.Update.Bulk.do_atomic_update/5
       (ash 3.5.43) lib/ash/data_layer/data_layer.ex:449: Ash.DataLayer.transaction/4
       (ash 3.5.43) lib/ash/actions/update/bulk.ex:276: Ash.Actions.Update.Bulk.run/6
       (app 0.1.0) deps/ash/lib/ash/code_interface.ex:1131: App.Domain.SecondResource.second_action/3
       (app 0.1.0) lib/app/domain/resource/changes/third_action_change.ex:19: anonymous fn/3 in App.Domain.Resource.Changes.ThirdActionChange.change/3
       (ash 3.5.43) lib/ash/changeset/changeset.ex:4706: anonymous fn/2 in Ash.Changeset.run_after_actions/3

[zachdaniel]

I'll need to see a failing test in ash_postgres or something to explain how we're getting to this issue.

[stevebrambilla]

@zachdaniel I've got a repro case for this here: stevebrambilla/ash_postgres@ce6a83f

That new test fails with ash_sql pointed to main, and works on this branch.

[zachdaniel]

@stevebrambilla mind PR-ing that failing test?

[zachdaniel]

@jimsynz feel free to merge once the credo issue is fixed 🙇

[stevebrambilla]

Opened the PR here: ash-project/ash_postgres#629