perf: avoid per-chunk dtype checks in BytesCodec decoding

[victorwon2001]

Summary

Eliminate per-chunk HasEndianness and Struct type checks from BytesCodec._decode_sync. The stored-byte view now uses the native NumPy dtype with newbyteorder(self.endian) whenever an endian is set, and uses the unchanged dtype otherwise. Byte-swapping, shape restoration and prototype-selected buffers remain intact; the public API and CPU/CUDA abstraction are unchanged.

For reviewers

The change is intended to preserve existing data and byte-order semantics. Direct and store-backed tests now cover integers and floats of multiple widths, mixed-endian structured fields, one-byte data, raw bytes, zero-copy matching-endian decoding and copying only for byte-order conversion. An alternating CPU benchmark found lower per-chunk scalar overhead and similar structured performance; the focused tests and full lint passed, while CUDA could not be checked locally.

Author attestation

• I am a human, these are my changes, and I have reviewed and understood every change and can explain why each is correct.

TODO

• Add unit tests and/or doctests in docstrings
• Add docstrings and API docs for any new/modified user-facing classes and functions
• New/modified features documented in docs/user-guide/*.md
• Changes documented as a new file in changes/
• GitHub Actions have all passed
• Test coverage is 100% (Codecov passes)

Fixes #3703

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 94.72%. Comparing base (bd0dc84) to head (0a0f416).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #4517   +/-   ##
=======================================
  Coverage   94.71%   94.72%           
=======================================
  Files          94       94           
  Lines       13663    13659    -4     
=======================================
- Hits        12941    12938    -3     
+ Misses        722      721    -1     

Files with missing lines	Coverage Δ
src/zarr/codecs/bytes.py	100.00% <100.00%> (+1.21%)	⬆️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.