kgyrtkirk commented on PR #16579: URL: https://github.com/apache/druid/pull/16579#issuecomment-2160740976
I wonder if the following is true: * suppose there is a table which has column which contains a string with a `\0` * based on my interpretation of the PR; normalization happens at `write` time * the 1st stage will see the field containing the `\0` - so if it computes some function say: `char_length` ; it will be counted in * and further stages or if the data is persisted the `\0` will not anymore be there that's why I thinked that normalizing at read time might be a better way to do this...as that will provide consistent behaviour even for the 1st usage as well. Now that I've thinked about it a bit more: I guess in that case it will be harder to identify which columns should be normalized at read time (and I guess a `\0` could possibly be added by a function as well). As this might be more complicated to do...maybe it doesn't worth the effort -- 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]
