pitrou commented on PR #51384: URL: https://github.com/apache/arrow/pull/51384#issuecomment-5792669015
Caching some expensive compute kernel precomputations (such as compiling a regex) was the original intent but I then noticed two problems: 1. The current LRU cache implementation is not extremely fast (partly because of the hash table I assume) 2. How to cache such precomputed kernel state is not obvious in the current architecture (it's not clear whether `KernelState` is meant to be invocation-specific or should be reusable, for example) -- 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. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
