Consider adding non-primitive API operations

My understanding is that the [current API](https://kernelci-api.eastus.cloudapp.azure.com/latest/docs) contains a set of primitive operations that more or less match 1-to-1 to database queries.

Some typical use cases that could be rather frequent, such as the [kci show results](https://github.com/kernelci/kernelci-core/blob/fb81830485aa2ddd4b83cd786bbe628209fffeb6/kernelci/cli/show.py#L66) tool, perform a potentially large number of individual API queries to extract a whole set of data. Accessing this data directly from the database must be almost instantaneous but doing so through primitive API calls takes a non-practical amount of time.

Suggestion: implement an extended set of non-primitive API operations that can do this kind of complex data retrieval in a single query. Any internal query complexities, data post-processing and formatting can be abstracted in the API.
This has two important benefits: 1) it'll make possible to do certain operations that currently are not practical to do due to the large number of queries required, 2) while these operations will be more computationally expensive, they should greatly reduce the query requests processing overhead. The net result should be favorable to the "single-complex-query" approach.

Start with the current API documentation and kernelci/cli/show.py to understand the primitive calls and the expensive retrieval pattern. The issue needs a concrete operation and scope before implementation; done would mean providing a defined non-primitive API operation that retrieves the required data in a single request.

索引モデルが issue の本文から書いたものです。