Use logging filter for `--comfyui-inference-log-level` flag

[eliteprox]

A logging filter would be more appropriate than the current async with approach for --comfyui-inference-log-level

comfystream/src/comfystream/pipeline.py

Lines 292 to 293 in 4f48695

	async with temporary_log_level("comfy", self._comfyui_inference_log_level):
	out_tensor = await self.client.get_audio_output()

comfystream/server/app.py

Lines 622 to 629 in 4f48695

	parser.add_argument(
	"--comfyui-inference-log-level",
	default=None,
	choices=logging._nameToLevel.keys(),
	help="Set the logging level for ComfyUI inference",
	)
	args = parser.parse_args()

comfystream/server/app.py

Lines 693 to 696 in 4f48695

	# Allow overriding of ComyfUI log levels.
	if args.comfyui_log_level:
	log_level = logging._nameToLevel.get(args.comfyui_log_level.upper())
	logging.getLogger("comfy").setLevel(log_level)

[linear]

COMFY-189 Use logging filter for `--comfyui-inference-log-level` flag