*SparkRules* has many rules, 2 of which does the opposite of the other. I'm referring to *EnumerableToSparkConverterRule* and *SparkToEnumerableConverterRule*.
I don't see how can these rules coexist together. I imagine they can force the planner into an infinite loop as one rule converts enumerable nodes to spark ones, while the other converts spark nodes to enumerable ones. 1) Would someone please tell me how are these two rules put to good use together ? And regarding the method *SparkHandlerImpl.compile*, I found that it's used in this one *EnumerableInterpretable.toBindable*, but I couldn't find a rule that converts nodes to *EnumerableInterpretable* to begin with ! 2) So my second question is, when is *EnumerableInterpretable* created ? Thanks, Gelbana
