gianm commented on pull request #11327: URL: https://github.com/apache/druid/pull/11327#issuecomment-861557741
> As far as the docs go, I think it would be worth trying to fill out the null handling of all functions, but I would maybe rather do this as a follow-up since there are a lot of them if that is ok. I imagine something similar to what #11188 added for aggregator functions, but I'm not yet sure if that format makes the most sense for the other functions, or if there is a more concise way to describe general behavior unless otherwise specified, need to think about it a bit. I think the way PostgreSQL does it is reasonable for a single-page doc style: https://www.postgresql.org/docs/13/functions-string.html. It describes null handling behavior only for functions where it's non-obvious (like concat) or where there's something special going on (like quote_literal and quote_nullable). It keeps the noise down. One day, we might want to consider a multi-page doc style, though, with a whole page for each function. Then we can go into all the details, give usage examples, etc. Microsoft docs are like that: https://docs.microsoft.com/en-us/sql/t-sql/functions/concat-transact-sql?view=sql-server-ver15. But I wouldn't worry about that as part of the null handling documentation stuff. -- 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: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
