andygrove commented on issue #2893: URL: https://github.com/apache/datafusion-sqlparser-rs/issues/2893#issuecomment-6827378773
> Any comments [on my proposed approach](https://github.com/apache/datafusion-sqlparser-rs/issues/2893#issuecomment-6752884942)? I reckon it solves the largest performance issue, which is the size of the enum (currently I believe an entry thereof is north of 3kb see [#2592](https://github.com/apache/datafusion-sqlparser-rs/pull/2592)), and has user pay exactly for what they need. > > Any approach that does not allow for fractioning of the enum, without introducing duplications in doing so, and makes the code structured and modular, is doomed to fail in my opinion. > > Expecially if the intention is to rely more on LLMs, we need much more structure and tests (and less tautological ones than currently done) than it is currently present. I started working through your proposal yesterday. I'll try and get a response to you today. It does address issues that I wasn't thinking about. My primary motication was having vendor-specific parsing logic in each dialect rather than one huge parser full of conditionals. I wasn't looking at enum sizes. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
