Github user amansinha100 commented on a diff in the pull request:
https://github.com/apache/drill/pull/517#discussion_r70487824
--- Diff:
exec/java-exec/src/main/codegen/templates/Decimal/CastIntDecimal.java ---
@@ -64,7 +64,15 @@ public void setup() {
public void eval() {
out.scale = (int) scale.value;
- out.precision = (int) precision.value;
--- End diff --
The eval() method is called for every row, so there will be a performance
overhead. I agree about the storage cost if we assume worst case precision;
however the per-row evaluation should be as lightweight as possible and likely
outweighs the storage consideration. One option is to treat the IntDecimal and
BigIntDecimal separately. For IntDecimal the max int value is 2^31 - 1 which
has a max precision of 10. Similarly, there is an upper bound on the
BigIntDecimal. You could initialize the precision in the setup() method which
is called *once* per new schema (in the normal case the schema of that column
does not change, this would avoid unnecessarily computing the precision).
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
