API: consistent NaN treatment for pyarrow dtypes #24269

Job	Run time
Linux-32-bit	18m 29s
Linux-Musl	15m 18s
Pyodide build	23m 11s
windows-latest actions-311.yaml	37m 20s
python-dev	0s
macos-13 actions-310.yaml	37m 44s
macos-14 actions-311.yaml	20m 40s
macos-13 actions-312.yaml	37m 23s
windows-latest actions-312.yaml	38m 9s
macos-13 actions-311.yaml	41m 58s
windows-latest actions-313.yaml	48m 34s
macos-14 actions-313.yaml	17m 41s
macos-13 actions-313.yaml	38m 32s
macos-14 actions-312.yaml	17m 22s
windows-latest actions-310.yaml	38m 20s
ubuntu-24.04 actions-311.yaml	43m 26s
ubuntu-24.04-arm actions-312.yaml	35m 54s
Future infer strings	45m 15s
Downstream Compat	6m 35s
ubuntu-24.04 actions-310.yaml	47m 40s
ubuntu-24.04-arm actions-313.yaml	37m 36s
macos-14 actions-310.yaml	20m 16s
Freethreading	21m 2s
Locale: it_IT	24m 33s
Minimum Versions	24m 25s
ubuntu-24.04-arm actions-310.yaml	36m 21s
Numpy Dev	16m 49s
ubuntu-24.04-arm actions-311.yaml	37m 53s
Pyarrow Nightly	20m 52s
ubuntu-24.04 actions-313.yaml	48m 22s
ubuntu-24.04 actions-312.yaml	44m 19s
Pypy	1m 47s
Locale: zh_CN	23m 59s
Future infer strings (without pyarrow)	39m 15s
	16h 47m 0s

Provide feedback