Fix a failure for nest["field"] and empty chunks

[hombit]

This morning I was running a pipeline with LSDB and got an exception complaining about empty input for pa.chunked_array.

I don't really know how to reproduce it (empty dataframe/series didn't work), but this PR should fix it. I also changed a similar code in another place for consistency.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 97.73%. Comparing base (e18390e) to head (cc40fb2).
Report is 2 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #292      +/-   ##
==========================================
+ Coverage   97.67%   97.73%   +0.05%     
==========================================
  Files          18       18              
  Lines        1632     1630       -2     
==========================================
- Hits         1594     1593       -1     
+ Misses         38       37       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

Before [e18390e] <v0.4.5>	After [4b5b976]	Ratio	Benchmark (Parameter)
46.4±0.01ms	41.3±1ms	~0.89	benchmarks.ReassignHalfOfNestedSeries.time_run
28.1±0.5ms	29.0±0.6ms	1.03	benchmarks.AssignSingleDfToNestedSeries.time_run
1.43±0.02ms	1.44±0.03ms	1.01	benchmarks.NestedFrameReduce.time_run
246M	246M	1.00	benchmarks.AssignSingleDfToNestedSeries.peakmem_run
134M	134M	1.00	benchmarks.CountNestedBy.peakmem_run
70.4±1ms	70.6±1ms	1.00	benchmarks.CountNestedBy.time_run
101M	101M	1.00	benchmarks.NestedFrameAddNested.peakmem_run
104M	104M	1.00	benchmarks.NestedFrameQuery.peakmem_run
104M	104M	1.00	benchmarks.NestedFrameReduce.peakmem_run
267M	267M	1.00	benchmarks.ReassignHalfOfNestedSeries.peakmem_run

Click here to view all benchmarks.

[dougbrn]

LGTM!