[ARROW-3404] [C++] Make CSV chunker faster

• Type: Improvement
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.11.0
• Fix Version/s: 0.11.0
• Component/s: C++
• Labels: pull-request-available
• Assignee: Antoine Pitrou
• Reporter: Antoine Pitrou
• Created: 02/Oct/18 15:13
• Updated: 02/Oct/18 19:34
• Resolved: 02/Oct/18 19:33

Description

Currently the CSV chunker can be the bottleneck in multi-threaded reads (starting from 6 threads, according to my experiments). One way to make it faster is to consider by default that CSV values cannot contain newline characters (overridable via a setting), and then simply search for the last newline character in each block of data.

Issue Links

• links to: GitHub Pull Request #2684