Details
-
Task
-
Status: Closed
-
Major
-
Resolution: Fixed
-
None
-
None
Description
Currently, the Druid Calcite adapter assumes that the row returned by druid has only one timestamp typed column, this is not true, in fact, we can have multiple projections of the time column with extraction functions. Thus code like this need to be revisited.
int posTimestampField = -1; for (int i = 0; i < fieldTypes.size(); i++) { if (fieldTypes.get(i) == ColumnMetaData.Rep.JAVA_SQL_TIMESTAMP) { posTimestampField = i; break; } }