Mapping-tester doesn't respect full cross product of rank choices

[fsimonis]

When using the mapping-tester based on branch precice-v3 with

{
  "general": {
    "ranks": {
      "A": [1,2],
      "B": [2,4]
    }
  }
}

Then the expected runs are 1-2, 1-4, 2-2, 2-4.
The actual runs are 1-2, 2-4

[kursatyurt]

Do we expect a cross-product or pairing? From the implementation does not seem like a bug but rather a choice

aste/tools/mapping-tester/generate.py

Lines 36 to 39 in a518b4a

	for ranksA, ranksB in zip(
	as_iter(setup["general"]["ranks"].get("A", 1)),
	as_iter(setup["general"]["ranks"].get("B", 1)),
	):

[davidscn]

Related question might also be: how do you want to process all runs when evaluating all combinations. Seems related to #163