Support S3-backed Zeek log inputs

[keilogic]

Summary

• parse Zeek log headers through local files or optional fsspec URI paths
• let LogToDataFrame and JSONLogToDataFrame pass remote paths such as s3://... through to Pandas once the matching fsspec backend is installed
• add zat[s3] packaging/docs and memory-backed regression coverage for URI input paths

Fixes #141.

Testing

• PYTHONPATH=. python -m pytest zat\utils\field_info.py zat\log_to_dataframe.py zat\json_log_to_dataframe.py -q -> 3 passed
• python -m black --check zat\utils\field_info.py zat\log_to_dataframe.py zat\json_log_to_dataframe.py
• python -m isort --check zat\utils\field_info.py zat\log_to_dataframe.py zat\json_log_to_dataframe.py
• python -m flake8 zat\utils\field_info.py zat\log_to_dataframe.py zat\json_log_to_dataframe.py
• git diff --check
• Manual memory:// LogToDataFrame load returned (360, 3)
• PYTHONPATH=. python -m pytest zat -q -> 19 passed, 3 skipped, 2 unrelated local Windows failures:
  • zat/utils/cache.py::test hit the existing /tmp/zat_file_storage\my_test_cache rename collision
  • zat/utils/signal_utils.py::test uses signal.SIGQUIT, which is not available on Windows