[ARROW-10043] [Rust] [DataFusion] Introduce support for DISTINCT by partially implementing COUNT(DISTINCT)

• Type: Wish
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Fix Version/s: 2.0.0
• Component/s: Rust, Rust - DataFusion
• Labels: pull-request-available
• External issue URL: https://github.com/apache/arrow/issues/26065
• Assignee: Daniel Russo
• Reporter: Daniel Russo
• Created: 19/Sep/20 18:41
• Updated: 11/Jan/23 08:10
• Resolved: 08/Oct/20 23:43

Description

I am unsure where support for DISTINCT may be on the DataFusion roadmap, so I've filed this with the "Wish" type and "Minor" priority to reflect that this is a proposal:

Introduce DISTINCT into DataFusion by partially implementing COUNT(DISTINCT). The ultimate goal is to fully support the DISTINCT keyword, but to get implementation started, limit the scope of this work to:

• the COUNT() aggregate function
• a single expression in COUNT(), i.e., COUNT(DISTINCT c1), but not COUNT(DISTINCT c1, c2)
• only queries with a GROUP BY clause
• integer types

Issue Links

• relates to: ARROW-10510 [Rust] [DataFusion] Add benchmarks for COUNT(DISTINCT)
• links to: GitHub Pull Request #8222