[ARROW-10403] [C++] Implement unique kernel for dictionary type

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 4.0.0
• Component/s: C++
• Labels: pull-request-available
• External issue URL: https://github.com/apache/arrow/issues/26386
• Assignee: Rok Mihevc
• Reporter: Neal Richardson
• Created: 27/Oct/20 18:07
• Updated: 11/Jan/23 08:13
• Resolved: 08/Apr/21 12:38

Description

Calling the "unique" compute function on a ChunkedArray of dictionary type (as read by the CSV reader) errors with "Only hashing for data with equal dictionaries currently supported". But is it necessary to hash to get unique values from a dictionary type? The dictionary values are the unique values (for each chunk), they're already there.

Issue Links

• links to: GitHub Pull Request #9683