FIX-#4580: Fix access by row label in query and eval

[mvashishtha]

What do these changes do?

query and eval raise exceptions as in #4580 on partitioned execution engines when given expressions that access elements by row label. I don't see a way to do distributed execution for query and eval other than doing what pandas does when using the python engine, which is to evaluate the expression as python (and then pass it into df.loc[result, :] for query). That's what I do in this PR.

So far I have been able to use pandas.DataFrame.query and pandas.DataFrame.eval directly without copying code. I prefer that for now because we would have to copy many methods.

Supporting the numexpr engine this way will take more work.

• first commit message and PR title follow format outlined here

  NOTE: If you edit the PR title to match this format, you need to add another commit (even if it's empty) or amend your last commit for the CI job that checks the PR title to pick up the new PR title.

• passes flake8 modin/ asv_bench/benchmarks scripts/doc_checker.py
• passes black --check modin/ asv_bench/benchmarks scripts/doc_checker.py
• signed commit with git commit -s
• Resolves query and eval: accessing column data by position raises errors #4580
• tests added and passing
• module layout described at docs/development/architecture.rst is up-to-date

[devin-petersohn]

LGTM, thanks @mvashishtha!

[dchigarev]

After this PR we see about ~1.6x perf degradation for one of our internal workloads that uses simple queries a lot.

@mvashishtha can you please check #6499?