Make query randomization more flexible

[peteralfonsi]

Is your feature request related to a problem? Please describe

Query randomization (#455) lets OSB automatically randomize the bounds of a range query in a controllable way, which is useful for testing cache performance. But it's currently hardcoded to only work on range queries with parameters gte/gt, lte/lt, and optionally format. This isn't ideal. For example, range queries like this don't usually enter the query cache, so testing the query cache requires more flexibility.

Describe the solution you'd like

For each operation, the user should be able to define the query name they are looking to randomize, and also specify the names of the parameters that should be randomized. For example, on a GeoPoint field named "pickup_location", they could randomize the following query:

"query": {
          "geo_bounding_box": {
            "pickup_location": {
              "top_left": [-74.27, 40.92],
              "bottom_right": [-73.68, 40.49]
            }
          }
        }

by specifying the query name geo_bounding_box and the parameter names top_left and bottom_right. This can be done by registering these values on an operation level in workload.py, like how standard value sources for the randomness are registered.

Describe alternatives you've considered

No response

Additional context

No response

[Divyaasm]

Hey @rishabh6788, @IanHoang Can you please take a look! Thanks

[IanHoang]

@peteralfonsi Thanks for contributing the PR. Will take a look when we free up