[ARROW-3764] [C++] Port Python "ParquetDataset" business logic to C++

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.0.0
• Component/s: C++
• Labels: dataset dataset-parquet-read parquet
• External issue URL: https://github.com/apache/arrow/issues/20089
• Assignee: Unassigned
• Reporter: Wes McKinney
• Created: 11/Nov/18 22:22
• Updated: 11/Jan/23 07:29
• Resolved: 13/Aug/20 18:54

Description

Along with defining appropriate abstractions for dealing with generic filesystems in C+, we should implement the machinery for reading multiple Parquet files in C+ so that it can reused in GLib, R, and Ruby. Otherwise these languages will have to reimplement things, and this would surely result in inconsistent features, bugs in some implementations but not others

Issue Links

• is related to: ARROW-8208 [PYTHON] Row Group Filtering With ParquetDataset