Reconstruct dtypes correctly for list aggs of struct columns

[wence-]

Description

As usual when returning from libcudf, we need to reconstruct a struct
dtype with appropriate labels. For groupby.agg(list) this can be done
by matching on the element_type of the result column and
reconstructing with a new list dtype with a leaf from the original
column.

Closes #11765
Closes #11907

Checklist

• I am familiar with the Contributing Guidelines.
• New or existing tests cover these changes.
• The documentation is up to date with these changes.

[codecov]

Codecov Report

Base: 86.58% // Head: 85.68% // Decreases project coverage by -0.90% ⚠️

Coverage data is based on head (0111854) compared to base (b6dccb3).
Patch has no changes to coverable lines.

Additional details and impacted files

@@               Coverage Diff                @@
##           branch-23.02   #12290      +/-   ##
================================================
- Coverage         86.58%   85.68%   -0.90%     
================================================
  Files               155      155              
  Lines             24368    24868     +500     
================================================
+ Hits              21098    21309     +211     
- Misses             3270     3559     +289     

Impacted Files	Coverage Δ
python/cudf/cudf/_version.py	1.41% <0.00%> (-98.59%)	⬇️
python/cudf/cudf/core/buffer/spill_manager.py	72.50% <0.00%> (-7.50%)	⬇️
python/cudf/cudf/core/buffer/spillable_buffer.py	91.07% <0.00%> (-1.78%)	⬇️
python/cudf/cudf/utils/dtypes.py	77.85% <0.00%> (-1.61%)	⬇️
python/cudf/cudf/options.py	86.11% <0.00%> (-1.59%)	⬇️
python/cudf/cudf/core/single_column_frame.py	94.30% <0.00%> (-1.27%)	⬇️
...ython/custreamz/custreamz/tests/test_dataframes.py	98.38% <0.00%> (-1.01%)	⬇️
python/dask_cudf/dask_cudf/io/csv.py	96.34% <0.00%> (-1.00%)	⬇️
python/dask_cudf/dask_cudf/io/parquet.py	91.81% <0.00%> (-0.59%)	⬇️
python/cudf/cudf/core/multiindex.py	91.66% <0.00%> (-0.51%)	⬇️
... and 43 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[wence-]

One cibuildwheel run failed (I guess due to network timeouts, it looks like). Is there a way to restart this, or do I just push a(nother) merge commit and cross my fingers?

[vyasr]

One cibuildwheel run failed (I guess due to network timeouts, it looks like). Is there a way to restart this, or do I just push a(nother) merge commit and cross my fingers?

If you click on the "Details" link next to a failed check it will take you into the relevant "Actions" section. There you'll see a "Re-run jobs" in the top right, which you can use to only rerun failed tests.

[vyasr]

FWIW the latest failure appears to be the same one that @madsbk observed in #12554 (comment) and merging again seems to have fixed it (I didn't look into the underlying cause upstream, perhaps something related to the new recent pandas version causing the xfailed test to succeed, or maybe some transient inconsistency with some upstream being pulled? It's very weird that it's an xfail-related failure though, so not sure without digging further.).

[wence-]

/merge