On Nov 11, 2014, at 10:51 AM, Vladimir Sitnikov <[email protected]> 
wrote:

>> How about creating your set of core rules as a data structure in Programs?
> 
> Not sure I get this. What do you mean by Program?

The class org.apache.calcite.tools.Programs contains utilities for building 
instances of org.apache.calcite.tools.Program, and it has some pre-built 
programs and rule-sets, such as Programs.CALC_RULES. You could add 
Programs.VLADIMIR_RULES. :)

> I do not mind having some rules that I enable on-demand. I just do not
> see the road ahead.
> 
>> Also, there isn’t a version of EnumerableScan that takes correlating 
>> variables (in sargs) and without that, there’s no point in doing a re-start 
>> — you’d get the same results every time
> 
> Is that intentional or just because no one asked that before?

No one asked before. I had thought of it — not just for indexes, but also for 
lateral join into nested collections (e.g. JdbcTest.Department.employees), 
tables based on parameterized web-service calls, and now potentially tables 
based ProjectableFilterableTable — but it never rose to the top of my list. I’m 
delighted that you are going there...

Julian

Reply via email to