ADD: Support reading NEXRAD Level 2 chunks from real-time S3 bucket

[aladinor]

Summary

• Add support for list/tuple of chunk sources (bytes, file-like, or paths) as input to open_nexradlevel2_datatree for streaming NEXRAD Level 2 data from the unidata-nexrad-level2-chunks S3 bucket
• Add incomplete_sweep parameter ("drop"/"pad") to handle partial volumes where not all chunks have arrived yet
• Add notebook example demonstrating real-time chunk streaming from S3
• Add comprehensive offline test suite for chunk list-input and incomplete sweep handling

Closes #330

Changes

Core implementation (xradar/io/backends/nexrad_level2.py)

• _concatenate_chunks(): Assembles a list of chunk byte objects, file-like objects, or file paths into a single byte stream, with automatic volume header detection and skipping for non-start chunks
• incomplete_sweeps property on NEXRADLevel2File: Tracks which sweeps were force-closed (incomplete) during parsing
• incomplete_sweep parameter in open_nexradlevel2_datatree:
  • "drop" (default): Excludes incomplete sweeps and emits a warning
  • "pad": Keeps incomplete sweeps and reindexes to a full azimuth grid (360 or 720 azimuths), filling missing rays with NaN

Fix (xradar/util.py)

• Use .values.copy() instead of .values to prevent modifying read-only arrays during angle reindexing

Notebook (examples/notebooks/nexrad_read_chunks.ipynb)

• Full walkthrough: connecting to S3, downloading chunks, opening full/partial volumes, drop vs pad mode, early streaming with few chunks

Documentation (docs/importers.md, docs/usage.md, docs/history.md)

• User-facing docs for chunk input and incomplete_sweep parameter with code examples
• Notebook registered in docs toctree
• Changelog entries

Tests (tests/io/test_nexrad_level2.py)

• 21 new tests across 5 test classes covering:
  • _concatenate_chunks edge cases (empty list, single element, volume header, short chunks, file-like objects, nonexistent paths, tuple input)
  • Sweep completeness detection (multiple incomplete, all incomplete, missing key defaults)
  • incomplete_sweep parameter behavior (drop warns, pad passes incomplete set, late-detection path)
  • Pad-mode reindexing with real data (full azimuth grid, data preservation, NaN percentage)
  • List/tuple input integration with drop/pad modes

Test plan

• All 75 tests pass (pytest tests/io/test_nexrad_level2.py)
• Notebook executes successfully against live S3 data
• Docs build successfully (cd docs && make html)

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 93.69%. Comparing base (b00e01f) to head (a2dc470).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #332      +/-   ##
==========================================
+ Coverage   93.59%   93.69%   +0.10%     
==========================================
  Files          27       27              
  Lines        5654     5730      +76     
==========================================
+ Hits         5292     5369      +77     
+ Misses        362      361       -1     

Flag	Coverage Δ
notebooktests	0.00% <0.00%> (ø)
unittests	93.69% <100.00%> (+0.10%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ 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.

[kmuehlbauer]

Two quick comments after skimming through this neat PR. I'll have a closer look over the weekend. Thanks @aladinor!