Details
-
New Feature
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Won't Do
-
None
-
None
Description
Our hash tables are currently unordered_multimaps, where the key is redundant with one (or more) of the columns stored as the value. We could use reduce mem usage by using an unordered_multiset, not storing the keys seperately, and using functors that know what the key columns are to generate the hash() and equal_to() values to make it function like the multimaps we're using now.