API: consistent NaN treatment for pyarrow dtypes #24093

Job	Run time
python-dev	0s
Linux-Musl	15m 32s
Linux-32-bit	18m 27s
macos-13 actions-310.yaml	52m 28s
macos-13 actions-311.yaml	36m 36s
macos-13 actions-312.yaml	35m 52s
macos-14 actions-312.yaml	22m 50s
macos-14 actions-313.yaml	20m 2s
windows-latest actions-310.yaml	44m 0s
windows-latest actions-311.yaml	44m 6s
windows-latest actions-312.yaml	45m 30s
macos-13 actions-313.yaml	33m 6s
macos-14 actions-310.yaml	41m 21s
Pyodide build	23m 15s
windows-latest actions-313.yaml	49m 29s
macos-14 actions-311.yaml	24m 22s
ubuntu-24.04-arm actions-310.yaml	40m 58s
ubuntu-24.04-arm actions-312.yaml	42m 22s
Downstream Compat	6m 18s
Freethreading	21m 22s
Numpy Dev	17m 1s
ubuntu-24.04 actions-310.yaml	54m 57s
ubuntu-24.04-arm actions-313.yaml	39m 45s
ubuntu-24.04 actions-312.yaml	51m 26s
Minimum Versions	29m 37s
Pyarrow Nightly	34m 48s
ubuntu-24.04 actions-311.yaml	49m 10s
Locale: it_IT	30m 13s
ubuntu-24.04 actions-313.yaml	51m 31s
ubuntu-24.04-arm actions-311.yaml	41m 59s
Locale: zh_CN	29m 47s
Future infer strings	52m 37s
Future infer strings (without pyarrow)	39m 48s
Pypy	1m 56s
	19h 2m 31s

Provide feedback