Allow for curbing logs while testing

**Is your feature request related to a problem? Please describe.**
Yes, I am writing tests for my graphene app. I need to test whether certain queries produce errors. When `client.execute(...)`is run, this logs an annoying `graphql.error.located_error.GraphQLLocatedError`. I cannot even use `pytest.raises` or `self.assertRaises` in unittest because it only logs without actually raising an error. It even logs the entire traceback which is even more annoying.

**Describe the solution you'd like**
It'd be much better if we had a solution to have a loglevel flag in `client.execute(...)` or some other way to disable logging of those specific errors.

**Describe alternatives you've considered**
I have considered disabling all error logs from the test runner (with `logging.disable(logging.ERROR)`). I am going with this for now but I don't want to miss out on other important logs which may help in surfacing faulty tests.

**Additional context**
Add any other context or screenshots about the feature request here.

Start at the client.execute(...) entry point and trace how GraphQLLocatedError instances are logged during query execution. Check the existing pytest and unittest behavior described in the issue, then make the logging controllable without suppressing unrelated test-runner logs; done means callers can curb these specific errors while still receiving other important logs.

索引モデルが issue の本文から書いたものです。