Missing cloud data

[ndrwrbgs]

Cloud data is omitted from the parsed data, but returned in the response cloudLayers

[ndrwrbgs]

Ex:

{ 
  'id': 'https://api.weather.gov/stations/KPHL/observations/2025-12-23T04:15:00+00:00',
  'type': 'Feature',
  'geometry': {
	  'type': 'Point',
	  'coordinates': [-75.25, 39.87]
  },
  'properties': {
	  '@id': 'https://api.weather.gov/stations/KPHL/observations/2025-12-23T04:15:00+00:00',
	  '@type': 'wx:ObservationStation',
	  'elevation': {
		  'unitCode': 'wmoUnit:m',
		  'value': 6
	  },
	  'station': 'https://api.weather.gov/stations/KPHL',
	  'stationId': 'KPHL',
	  'stationName': 'Philadelphia, Philadelphia International Airport',
	  'timestamp': '2025-12-23T04:15:00+00:00',
	  'rawMessage': '',
	  'textDescription': 'Cloudy',
	  'icon': 'https://api.weather.gov/icons/land/night/ovc?size=medium',
	  'presentWeather': [],
	  'temperature': {
		  'unitCode': 'wmoUnit:degC',
		  'value': 3,
		  'qualityControl': 'V'
	  },
	  'dewpoint': {
		  'unitCode': 'wmoUnit:degC',
		  'value': -6,
		  'qualityControl': 'V'
	  },
	  'windDirection': {
		  'unitCode': 'wmoUnit:degree_(angle)',
		  'value': 200,
		  'qualityControl': 'V'
	  },
	  'windSpeed': {
		  'unitCode': 'wmoUnit:km_h-1',
		  'value': 7.416,
		  'qualityControl': 'V'
	  },
	  'windGust': {
		  'unitCode': 'wmoUnit:km_h-1',
		  'value': None,
		  'qualityControl': 'Z'
	  },
	  'barometricPressure': {
		  'unitCode': 'wmoUnit:Pa',
		  'value': 103047.8,
		  'qualityControl': 'V'
	  },
	  'seaLevelPressure': {
		  'unitCode': 'wmoUnit:Pa',
		  'value': None,
		  'qualityControl': 'Z'
	  },
	  'visibility': {
		  'unitCode': 'wmoUnit:m',
		  'value': 16093.44,
		  'qualityControl': 'C'
	  },
	  'maxTemperatureLast24Hours': {
		  'unitCode': 'wmoUnit:degC',
		  'value': None
	  },
	  'minTemperatureLast24Hours': {
		  'unitCode': 'wmoUnit:degC',
		  'value': None
	  },
	  'precipitationLast3Hours': {
		  'unitCode': 'wmoUnit:mm',
		  'value': None,
		  'qualityControl': 'Z'
	  },
	  'relativeHumidity': {
		  'unitCode': 'wmoUnit:percent',
		  'value': 51.631857192882,
		  'qualityControl': 'V'
	  },
	  'windChill': {
		  'unitCode': 'wmoUnit:degC',
		  'value': 0.9723700416888889,
		  'qualityControl': 'V'
	  },
	  'heatIndex': {
		  'unitCode': 'wmoUnit:degC',
		  'value': None,
		  'qualityControl': 'V'
	  },
	  'cloudLayers': [{
			  'base': {
				  'unitCode': 'wmoUnit:m',
				  'value': 2895.6
			  },
			  'amount': 'OVC'
		  }
	  ]
  }
}

[MatthewFlamm]

It needs to go here

pynws/src/pynws/simple_nws.py

Lines 137 to 158 in a48320c

	OBSERVATIONS: Final[Dict[str, Optional[MetarParam]]] = {
	"temperature": MetarParam("temp", "C"),
	"barometricPressure": None,
	"seaLevelPressure": MetarParam("press", "HPA", 100.0),
	"relativeHumidity": None,
	"windSpeed": MetarParam("wind_speed", "MPS", 3.6),
	"windDirection": MetarParam("wind_dir"),
	"visibility": MetarParam("vis", "M"),
	"elevation": None,
	"textDescription": None,
	"dewpoint": None,
	"windGust": None,
	"station": None,
	"timestamp": None,
	"icon": None,
	"maxTemperatureLast24Hours": None,
	"minTemperatureLast24Hours": None,
	"precipitationLastHour": None,
	"precipitationLast3Hours": None,
	"precipitationLast6Hours": None,
	"windChill": None,
	"heatIndex": None,

However the data structure is unique, so it will also need special handling.

[ndrwrbgs]

I wasn’t sure what would be best for this. I couldn’t find prior art of handling lists in home assistant, and the data returned is relatively unstructured, perhaps you have a suggestion?

What I actually want to get is to use the weather station near my home and get the observed cloud cover (or better yet- forecast cloud cover) which I am using to estimate human perceived solar radiation and adjust climate settings in response to.

For this, most sources said to use the densest returned value when density at multiple heights is returned. However, doing that kind of processing in an api library feels wrong as maybe someone else wouldn’t want to use it the same way? (Or is this concern YAGNI until someone else describes their use case?)

[MatthewFlamm]

This library has has some opinionated choices, but I agree we shouldn't make those choices based solely on HA usage. It should be based on wide usage.

pynws should should have a data structure that presents the fullness of the data. If most sources say that the densest layer should be picked out, I think it makes sense to add as an additional key/value in the dictionary.