mgaido91 commented on issue #23468: [SPARK-26535][SQL] Parse literals as DOUBLE instead of DECIMALS URL: https://github.com/apache/spark/pull/23468#issuecomment-451696246 @gatorsmile thanks for your comment. Let me cc also @cloud-fan, @dilipbiswal and @rxin who partecipated in the discussion related to this too. I agree on improving the doc on this. If you are fine with that, I'll create a JIRA and a PR for that ASAP. The main argument for this is: we are basing our decimal implementation on what Hive and MSSQL do. But we have mainly 2 differences with them at the moment: 1. We allow negative decimal scales (and at the moment we don't handle them properly); 2. We use decimals for literals, while Hive/MSSQL use DOUBLE (so we "enlarge" the issue). Using DOUBLE, when possible, seems safer and it is what is done by our references for these operations.
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
