Update groupby::hash to use new row operators for keys

[PointKernel]

Related to #8039 and #10181

Contributes to #10186

This PR updates groupby::hash to use new row operators. It gets rid of the current "flattened nested column" logic and allows groupby::hash to handle LIST and STRUCT keys. The work also involves small cleanups like getting rid of unnecessary template parameters and removing unused arguments.

It becomes a breaking PR since the updated groupby::hash will treat inner nulls as equal when top-level nulls are excluded
while the current behavior treats inner nulls as unequal.

[codecov]

Codecov Report

Merging #10770 (efd497e) into branch-22.06 (e0d94f3) will increase coverage by 0.03%.
The diff coverage is n/a.

@@               Coverage Diff                @@
##           branch-22.06   #10770      +/-   ##
================================================
+ Coverage         86.28%   86.32%   +0.03%     
================================================
  Files               144      144              
  Lines             22654    22668      +14     
================================================
+ Hits              19548    19569      +21     
+ Misses             3106     3099       -7     

Impacted Files	Coverage Δ
python/cudf/cudf/utils/ioutils.py	79.47% <0.00%> (-0.13%)	⬇️
python/cudf/cudf/io/json.py	97.56% <0.00%> (ø)
python/dask_cudf/dask_cudf/tests/test_groupby.py	100.00% <0.00%> (ø)
python/dask_cudf/dask_cudf/groupby.py	97.42% <0.00%> (+0.02%)	⬆️
python/cudf/cudf/core/dataframe.py	93.78% <0.00%> (+0.04%)	⬆️
python/cudf/cudf/core/column/string.py	88.78% <0.00%> (+0.12%)	⬆️
python/cudf/cudf/core/groupby/groupby.py	91.79% <0.00%> (+0.22%)	⬆️
python/dask_cudf/dask_cudf/core.py	73.62% <0.00%> (+0.26%)	⬆️
python/cudf/cudf/core/column/numerical.py	96.17% <0.00%> (+0.29%)	⬆️
python/cudf/cudf/core/tools/datetimes.py	84.49% <0.00%> (+0.30%)	⬆️
... and 1 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update fe9aaeb...efd497e. Read the comment docs.

[PointKernel]

@mythrocks requesting your review as well since you authored the flattened nested column work.

[PointKernel]

Setting the current work as breaking since the behavior is changed when nulls are excluded. See #10770 (comment)

[PointKernel]

rerun tests

[jrhemstad]

It becomes a breaking PR due to two reasons:

• The updated groupby::hash matches pandas' behavior, i.e. when nulls are excluded, the output will drop top-level nulls for struct/list but include struct/list containing null elements.
• Nulls are always treated as equal.

Wait, what? Why do we need to make these changes? I don't see these reflected in the top-level groupby API either.

[PointKernel]

PR description updated to provide clearer breaking information.

[PointKernel]

@gpucibot merge