[DRILL-1391] Some math functions on decimal type may cause compilation error in run-time generated code.

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.6.0
• Assignee: DrillCommitter
• Reporter: Jinfeng Ni
• Created: 08/Sep/14 22:35
• Updated: 09/Dec/14 17:49
• Resolved: 25/Sep/14 21:42

Description

If a query has an expression like the following :

log((cast(DEC18 as decimal18(18, 9))))

Drill will hit an compilation error in the run-time generated code.

org.apache.drill.exec.rpc.RpcException: Screen received stop request sent. Line 189, Column 35: Cannot determine simple type name "BigDecimal" [32dbdc99-a688-45c6-aa8f-50506be4730d].

The complained code is :

188: ExtendedMathFunctions$LogDecimal18_eval:

{ 189: double dblval = new BigDecimal(in.value).setScale(in.scale).doubleValue(); 190: 191: out.value = java.lang.Math.log(dblval); 192: }

The reason of this compilation error is that the function template for those math functions miss fully qualifying the type of "BigDecimal", causing this compilation error.