DataFrame from ADBC Client of Cube

[borodark]

PR: Default Pre-Aggregations (Opt-In)

Overview

• Adds an opt-in default pre-aggregation for auto-generated cubes when updated_at exists.
• Prints the pre-aggregation block in the auto-generated Elixir snippet.
• Enforces a consistent pre-aggregation name suffix: _automatic_for_the_people.
• Adds Cube HTTP integration coverage across date granularities/ranges.
• Documents the new flag in README and quick reference.

What’s New

• default_pre_aggregation: true generates a single rollup pre-aggregation.
• Rollup defaults:
  • external: true
  • time_dimension: :updated_at
  • granularity: :hour
  • refresh_key: SELECT MAX(id) FROM <sql_table>
  • build_range_start/end based on NOW()
  • excludes updated_at and inserted_at from dimensions
• Printed cube snippet now shows the pre-aggregation block (suppressed when sql_table is unknown).

Testing

mix test test/power_of_three/default_cube_test.exs
mix test test/power_of_three/preagg_default_integration_test.exs --include live_cube

Notes

• Fully backward compatible.
• Pre-aggregation remains editable after generation.
• ADBC Client to support Explorer.DataFrame

Pending integration of two forks:

• Feature/ADBC Server cube#2
• ADBC Driver For Cube ADBC Server adbc#2