Dandandan opened a new pull request #8781: URL: https://github.com/apache/arrow/pull/8781
Still WIP, but makes CSV reading (quite a bit) faster by reusing a bit more allocations, doing things a bit more manually. The nytaxi benchmark speeds up from ~4500ms to ~3500ms. I think a further optimization would be to stop using the `StringRecords` altogether (e.g. by using the underlying https://docs.rs/csv-core/0.1.10/csv_core/ library instead) but this could be a next step. FYI @alamb @nevi-me @jorgecarleitao ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org