Duplicate sub-keys in the partial key search #3
Description
In cases where there are some duplicate sub-keys in the partial key search, e.g. (a, b, a), and the search non-positional, both algorithms would effectively look for only one instance of the duplicate sub-key, e.g. (a, b).
The only remedy, for now, is to use the positional search for all the duplicate sub-keys, e.g. ((0, a), (-1, b), (2, a)). Or, one can get the results and then further filter out the duplicates based on the type of search (positional or not).
I am planning to implement post-processing like described above, or in-processing constraints at some point.