Refactor of caching & data storage part 9 #118

providers/solarconditions/giroionosonde.py

@@ -0,0 +1,173 @@
+import csv
+import logging
+from datetime import datetime, timezone, timedelta
+from threading import Thread, Event
+
+import pytz
+import requests
+from requests.exceptions import ConnectionError, ReadTimeout, ConnectTimeout
+
+from core.constants import HTTP_HEADERS
+from providers.solarconditions.ionosonde_utils import compute_band_states
+from providers.solarconditions.solar_conditions_provider import SolarConditionsProvider
+
+# Each station gets polled roughly once every hour (3600 seconds). Note that to avoid a burst of requests to the server
+# every hour, the requests for data from each station are spaced out throughout the hour, leading to one request being
+# sent every 1-2 minutes.
+POLL_INTERVAL = 3600
+# To avoid looking up all stations in the GIRO system and working out which ones are providing live data, this has been
+# manually determined and a CSV provided of all the stations that we can query for live data.
+STATIONS_INDEX = "datafiles/didbase-stations.csv"
+LGDC_URL = "https://lgdc.uml.edu/common/DIDBGetValues"
+HISTORY_HOURS = 24
+
+
+class GIROIonosonde(SolarConditionsProvider):
+    """Solar conditions provider using ionosonde data from the GIRO Data Center.
+    Queries foF2, MUF, and LUF measurements for all stations in datafiles/didbase-stations.csv.
+
+    Designed to run alongside KC2GProp even though they produce similar data. GIRO has more stations and includes LUF
+    data, but is less reliable and often offline."""
+
+    def __init__(self, provider_config):
+        super().__init__("GIRO Ionosonde Data", provider_config)
+        self._stations = self._load_stations()
+        self._thread = None
+        self._stop_event = Event()
+
+        # Pre-populate ionosonde_data with known station names for stations not already present,
+        # so the station dropdown is available before the first poll. Does not overwrite existing
+        # entries so KC2G cache data is preserved.
+        existing = self._solar_conditions.ionosonde_data or {}
+        new_entries = {
+            s["ursi"]: {"ursi": s["ursi"], "name": s["name"], "fof2": None, "muf": None,
+                        "luf": None, "band_states": None}
+            for s in self._stations if s["ursi"] not in existing
+        }
+        if new_entries:
+            self.update_data({"ionosonde_data": {**existing, **new_entries}})
+
+    @staticmethod
+    def _load_stations():
+        stations = []
+        with open(STATIONS_INDEX, newline='') as f:
+            for row in csv.reader(f):
+                if len(row) >= 2:
+                    stations.append({"ursi": row[0].strip(), "name": row[1].strip()})
+        return stations
+
+    def start(self):
+        logging.info(f"Set up query of GIRO ionosonde data API every {POLL_INTERVAL} seconds.")
+        self._thread = Thread(target=self._run, daemon=True)
+        self._thread.start()
+
+    def stop(self):
+        self._stop_event.set()
+
+    def _run(self):
+        # Real interval at which we poll is the "once per hour" divided by the number of stations, so each one gets
+        # polled once per hour, just not all at once
+        interval = POLL_INTERVAL / len(self._stations)
+        station_index = 0
+        while True:
+            self._poll_station(self._stations[station_index])
+            station_index = (station_index + 1) % len(self._stations)
+            if self._stop_event.wait(timeout=interval):
+                break
+
+    def _poll_station(self, station):
+        ursi = station["ursi"]
+        name = station["name"]
+        try:
+            logging.debug(f"Polling GIRO ionosonde data for {ursi} ({name})...")
+            now = datetime.now(timezone.utc)
+            from_time = now - timedelta(hours=HISTORY_HOURS)
+            cutoff_ts = from_time.timestamp()
+
+            fof2, muf, luf = self._fetch_station_data(ursi, from_time, now)
+            if not fof2 or not muf:
+                return
+
+            # Start from the existing ionosonde_data so stations provided by other providers
+            # (e.g. KC2GProp) are preserved for stations GIRO does not cover.
+            ionosonde_data = dict(self._solar_conditions.ionosonde_data or {})
+
+            # Merge GIRO's readings into any existing data for this station.
+            existing = ionosonde_data.get(ursi, {})
+            merged_fof2 = {**{float(t): v for t, v in (existing.get("fof2") or {}).items()}, **fof2}
+            merged_muf = {**{float(t): v for t, v in (existing.get("muf") or {}).items()}, **muf}
+            merged_luf = dict(luf) if luf else {}
+
+            merged_fof2 = {t: v for t, v in merged_fof2.items() if t >= cutoff_ts}
+            merged_muf = {t: v for t, v in merged_muf.items() if t >= cutoff_ts}
+            merged_luf = {t: v for t, v in merged_luf.items() if t >= cutoff_ts}
+
+            band_states = compute_band_states(merged_fof2, merged_muf, merged_luf)
+            ionosonde_data[ursi] = {
+                "ursi": ursi, "name": name,
+                "fof2": merged_fof2 or None,
+                "muf": merged_muf or None,
+                "luf": merged_luf or None,
+                "band_states": band_states,
+            }
+            self.update_data({"ionosonde_data": ionosonde_data})
+            self.status = "OK"
+            self.last_update_time = datetime.now(pytz.UTC)
+            logging.debug(f"Updated ionosonde data for {ursi} ({name}).")
+
+        except Exception:
+            self.status = "Error"
+            logging.exception(f"Exception fetching GIRO ionosonde data for {ursi} ({name})")
+
+    def _fetch_station_data(self, ursi, from_time, to_time):
+        """Fetch foF2, MUF and LUF readings for a station. Returns (fof2_dict, muf_dict, luf_dict) keyed by UNIX timestamp."""
+
+        from_str = from_time.strftime("%Y.%m.%d+%H:%M:%S")
+        to_str = to_time.strftime("%Y.%m.%d+%H:%M:%S")
+        url = f"{LGDC_URL}?ursiCode={ursi}&charName=foF2,MUFD,fmin&DMUF=3000&fromDate={from_str}&toDate={to_str}"
+        try:
+            http_response = requests.get(url, headers=HTTP_HEADERS, timeout=(5, 15))
+            if not http_response.ok:
+                logging.warning(f"HTTP {http_response.status_code} when calling Giro ionosonde API.")
+                return None, None, None
+            return self._parse_all(http_response.text)
+        except (ConnectTimeout, ReadTimeout):
+            logging.warning(f"Timeout when accessing Giro ionosonde API.")
+            return None, None, None
+        except ConnectionError:
+            logging.warning("Connection error when accessing Giro ionosonde API.")
+            return None, None, None
+
+    @staticmethod
+    def _parse_all(text):
+        """Parse web server response and return (fof2_dict, muf_dict, luf_dict) keyed by UNIX timestamp."""
+
+        fof2_data = {}
+        muf_data = {}
+        luf_data = {}
+        for line in text.splitlines():
+            line = line.strip()
+            if not line or line.startswith('#'):
+                continue
+            # Data rows have the following format: timestamp CS foF2 QD MUFD QD fmin QD
+            parts = line.split()
+            if len(parts) >= 5:
+                try:
+                    # Python 3.8 TZ parsing fudge
+                    ts = datetime.fromisoformat(parts[0].replace('Z', '+00:00')).timestamp()
+                except ValueError:
+                    continue
+                try:
+                    fof2_data[ts] = float(parts[2])
+                except ValueError:
+                    pass
+                try:
+                    muf_data[ts] = float(parts[4])
+                except ValueError:
+                    pass
+                if len(parts) >= 7:
+                    try:
+                        luf_data[ts] = float(parts[6])
+                    except ValueError:
+                        pass
+        return fof2_data, muf_data, luf_data

providers/solarconditions/hamqsl.py

@@ -0,0 +1,112 @@
+import logging
+from xml.etree import ElementTree
+
+import pytz
+from dateutil import parser as dateutil_parser, tz as dateutil_tz
+
+from providers.solarconditions.http_solar_conditions_provider import HTTPSolarConditionsProvider
+
+POLL_INTERVAL = 3600  # 1 hour
+URL = "https://www.hamqsl.com/solarxml.php"
+
+
+class HamQSL(HTTPSolarConditionsProvider):
+    """Solar conditions provider using the HamQSL.com XML API (https://www.hamqsl.com/solarxml.php).
+    Provides solar flux index, geomagnetic indices, and HF/VHF propagation condition summaries."""
+
+    def __init__(self, provider_config):
+        super().__init__("HamQSL", provider_config, URL, POLL_INTERVAL)
+
+    def _http_response_to_solar_conditions(self, http_response):
+        root = ElementTree.fromstring(http_response.text)
+        sd = root.find("solardata")
+        if sd is None:
+            logging.warning("HamQSL solar conditions API returned unexpected XML structure")
+            return None
+
+        # Some error checking functions in case the data is janky.
+
+        def text(tag, default=None):
+            if sd is None:
+                logging.warning("HamQSL solar conditions API returned unexpected XML structure")
+                return default
+            el = sd.find(tag)
+            return el.text.strip() if el is not None and el.text else default
+
+        def float_val(tag, default=None):
+            try:
+                return float(text(tag))
+            except (ValueError, TypeError):
+                return default
+
+        def int_val(tag, default=None):
+            try:
+                return int(text(tag))
+            except (ValueError, TypeError):
+                return default
+
+        # Process HF band conditions
+        hf_conditions = {}
+        calc = sd.find("calculatedconditions")
+        if calc is not None:
+            for band_el in calc.findall("band"):
+                name = band_el.get("name")
+                time = band_el.get("time")
+                condition = band_el.text.strip() if band_el.text else None
+                if name and time and condition:
+                    hf_conditions[f"{name}-{time}"] = condition
+
+        # Process VHF propagation conditions
+        vhf_map = {}
+        vhf = sd.find("calculatedvhfconditions")
+        if vhf is not None:
+            for ph_el in vhf.findall("phenomenon"):
+                key = (ph_el.get("name"), ph_el.get("location"))
+                vhf_map[key] = ph_el.text.strip() if ph_el.text else None
+
+        # Parse the "updated" timestamp string (format: "28 Mar 2026 0949 GMT") to UTC epoch seconds.
+        updated = None
+        updated_str = text("updated")
+        if updated_str:
+            try:
+                tz_abbr = updated_str.split()[-1]
+                timezone = dateutil_tz.gettz(tz_abbr)
+                if timezone is None:
+                    raise ValueError("Unknown timezone abbreviation: " + tz_abbr)
+                dt = dateutil_parser.parse(updated_str, tzinfos={tz_abbr: timezone})
+                updated = dt.astimezone(pytz.UTC).timestamp()
+            except (ValueError, IndexError):
+                logging.warning("HamQSL solar conditions API returned unrecognised timestamp format: " + updated_str)
+
+        # Return the data ready to be put into the solar conditions object.
+        return {
+            "updated": updated,
+            "sfi": int_val("solarflux"),
+            "a_index": int_val("aindex"),
+            "k_index": int_val("kindex"),
+            "xray": text("xray"),
+            "sunspots": int_val("sunspots"),
+            "proton_flux": int_val("protonflux"),
+            "electron_flux": int_val("electonflux"),
+            "aurora": int_val("aurora"),
+            "aurora_latitude": float_val("latdegree"),
+            "solar_wind": float_val("solarwind"),
+            "magnetic_field": float_val("magneticfield"),
+            "geomag_field": text("geomagfield").title()
+            .replace("Vr Quiet", "Very Quiet")
+            .replace("Unsettld", "Unsettled")
+            .replace("Min Strm", "Minor Storm")
+            .replace("Maj Strm", "Major Storm")
+            .replace("Sev Strm", "Severe Storm")
+            .replace("Ext Strm", "Extreme Storm"),
+            "geomag_noise": text("signalnoise"),
+            "hf_conditions": hf_conditions,
+            "vhf_conditions": {
+                "vhf_aurora_northern_hemi": (vhf_map.get(("vhf-aurora", "northern_hemi")) or "").title().replace(
+                    "Lat Aur", "Latitude") or None,
+                "es_2m_europe": vhf_map.get(("E-Skip", "europe")),
+                "es_4m_europe": vhf_map.get(("E-Skip", "europe_4m")),
+                "es_6m_europe": vhf_map.get(("E-Skip", "europe_6m")),
+                "es_2m_na": vhf_map.get(("E-Skip", "north_america")),
+            },
+        }

providers/solarconditions/http_solar_conditions_provider.py

@@ -0,0 +1,69 @@
+import logging
+from datetime import datetime
+from threading import Thread, Event
+
+import pytz
+import requests
+from requests.exceptions import ConnectionError, ReadTimeout, ConnectTimeout
+
+from core.constants import HTTP_HEADERS
+from providers.solarconditions.solar_conditions_provider import SolarConditionsProvider
+
+
+class HTTPSolarConditionsProvider(SolarConditionsProvider):
+    """Generic solar conditions provider for providers that request data via HTTP(S). Subclasses implement
+    _http_response_to_solar_conditions() to parse the specific API response format."""
+
+    def __init__(self, name, provider_config, url, poll_interval):
+        super().__init__(name, provider_config)
+        self._url = url
+        self._poll_interval = poll_interval
+        self._thread = None
+        self._stop_event = Event()
+
+    def start(self):
+        logging.info(
+            "Set up query of " + self.name + " solar conditions API every " + str(self._poll_interval) + " seconds.")
+        self._thread = Thread(target=self._run, daemon=True)
+        self._thread.start()
+
+    def stop(self):
+        self._stop_event.set()
+
+    def _run(self):
+        while True:
+            self._poll()
+            if self._stop_event.wait(timeout=self._poll_interval):
+                break
+
+    def _poll(self):
+        try:
+            logging.debug("Polling " + self.name + " solar conditions API...")
+            http_response = requests.get(self._url, headers=HTTP_HEADERS, timeout=(5, 30))
+            # Check response code was good
+            if http_response.ok:
+                new_data = self._http_response_to_solar_conditions(http_response)
+                self.update_data(new_data)
+
+                self.status = "OK"
+                self.last_update_time = datetime.now(pytz.UTC)
+                logging.debug("Received data from " + self.name + " solar conditions API.")
+            else:
+                self.status = "Error"
+                logging.warning(f"HTTP {http_response.status_code} when calling {self.name} solar conditions API.")
+
+        except ConnectionError:
+            logging.warning(f"Connection error when accessing {self.name} solar conditions API.")
+        except (ConnectTimeout, ReadTimeout):
+            logging.warning(f"Timeout when accessing {self.name} solar conditions API.")
+        except Exception:
+            self.status = "Error"
+            logging.exception("Exception in HTTP Solar Conditions Provider (" + self.name + ")")
+            self._stop_event.wait(timeout=1)
+
+    def _http_response_to_solar_conditions(self, http_response):
+        """Convert an HTTP response into solar conditions data. Returns a dict mapping SolarConditions field
+        names to their new values, or None if the response could not be parsed. Only the fields returned will
+        be updated on the shared SolarConditions object; any fields not included will be left unchanged."""
+
+        raise NotImplementedError("Subclasses must implement this method")

providers/solarconditions/ionosonde_utils.py

@@ -0,0 +1,37 @@
+from core.constants import BANDS
+
+HF_BANDS = [b for b in BANDS if b.is_ham_hf]
+
+
+def _latest(d) -> float | None:
+    """Given a map where the key is a timestamp and the value is a number represented as a string, find the latest
+    timestamp and return the corresponding value as a float."""
+
+    val = str(d[max(d.keys())]) if d else None
+    return float(val) if (val is not None and val != "None") else None
+
+
+def compute_band_states(fof2_dict, muf_dict, luf_dict):
+    """Compute HF band states from the latest foF2, MUF and LUF values.
+
+    Returns a map where the keys are HF bands and the values are as follows:
+      "Closed" if band frequency is above MUF or below LUF (if known)
+      "Short" if band frequency is >= LUF and < foF2 (good for NVIS)
+      "Long" if band frequency is >= foF2 and < MUF (good for DX)
+    """
+
+    fof2 = _latest(fof2_dict)
+    muf = _latest(muf_dict)
+    luf = _latest(luf_dict) if luf_dict else None
+    if fof2 is None or muf is None:
+        return {}
+    band_states = {}
+    for band in HF_BANDS:
+        freq = band.start_freq / 1_000_000
+        if freq > muf or (luf is not None and freq < luf):
+            band_states[band.name] = "Closed"
+        elif freq < fof2:
+            band_states[band.name] = "Short"
+        else:
+            band_states[band.name] = "Long"
+    return band_states

providers/solarconditions/kc2gprop.py

@@ -0,0 +1,126 @@
+import logging
+from datetime import datetime, timezone, timedelta
+from threading import Thread, Event
+
+import pytz
+import requests
+from requests.exceptions import ConnectionError, ReadTimeout, ConnectTimeout
+
+from core.constants import HTTP_HEADERS
+from providers.solarconditions.ionosonde_utils import compute_band_states
+from providers.solarconditions.solar_conditions_provider import SolarConditionsProvider
+
+POLL_INTERVAL = 900  # 15 minutes
+KC2G_URL = "https://prop.kc2g.com/api/stations.json"
+HISTORY_HOURS = 24
+
+
+class KC2GProp(SolarConditionsProvider):
+    """Solar conditions provider using ionosonde data from prop.kc2g.com. The API returns only the latest reading per
+    station, so this provider polls every 15 minutes and accumulates a 24-hour time series by merging each new reading
+    into the persisted ionosonde_data, producing the same data structure as GIROIonosonde.
+
+    Designed to run alongside GIROIonosonde even though they produce similar data. KC2G is more reliable and is always
+    online, but has fewer stations and does not provide LUF data."""
+
+    def __init__(self, provider_config):
+        super().__init__("KC2G Propagation Data", provider_config)
+        self._thread = None
+        self._stop_event = Event()
+
+    def start(self):
+        logging.info(f"Set up query of KC2G ionosonde data API every {POLL_INTERVAL} seconds.")
+        self._thread = Thread(target=self._run, daemon=True)
+        self._thread.start()
+
+    def stop(self):
+        self._stop_event.set()
+
+    def _run(self):
+        while True:
+            self._poll()
+            if self._stop_event.wait(timeout=POLL_INTERVAL):
+                break
+
+    def _poll(self):
+        try:
+            logging.debug("Polling KC2G ionosonde data...")
+            http_response = requests.get(KC2G_URL, headers=HTTP_HEADERS, timeout=(5, 30))
+            if not http_response.ok:
+                logging.warning(f"HTTP {http_response.status_code} when calling KG2G ionosonde API.")
+                return
+
+            now = datetime.now(timezone.utc)
+            cutoff_ts = (now - timedelta(hours=HISTORY_HOURS)).timestamp()
+
+            # Start from existing ionosonde_data so the accumulated time series survives across polls and restarts and
+            # stations provided only by GIROIonosonde are not discarded
+            ionosonde_data = dict(self._solar_conditions.ionosonde_data or {})
+            updated_count = 0
+
+            for reading in http_response.json():
+                station = reading.get("station", {})
+                ursi = station.get("code")
+                name = station.get("name")
+                if not ursi or not name:
+                    continue
+
+                time_str = reading.get("time")
+                if not time_str:
+                    continue
+                try:
+                    ts = datetime.fromisoformat(time_str)
+                    if ts.tzinfo is None:
+                        ts = ts.replace(tzinfo=timezone.utc)
+                    ts_float = ts.timestamp()
+                except ValueError:
+                    continue
+
+                # Skip readings outside our history window (some stations have months-old data)
+                if ts_float < cutoff_ts:
+                    continue
+
+                fof2_val = reading.get("fof2")
+                muf_val = reading.get("mufd")
+                if fof2_val is None and muf_val is None:
+                    continue
+
+                # Merge this reading into the existing time series for the station.
+                existing = ionosonde_data.get(ursi, {})
+                fof2_dict = dict(existing.get("fof2") or {})
+                muf_dict = dict(existing.get("muf") or {})
+                # LUF is not available from KC2G; carry forward whatever GIRO has written.
+                luf_dict = {float(t): v for t, v in (existing.get("luf") or {}).items()}
+
+                fof2_dict[ts_float] = fof2_val
+                muf_dict[ts_float] = muf_val
+
+                # Trim all series to the 24-hour window.
+                fof2_dict = {t: v for t, v in fof2_dict.items() if t >= cutoff_ts}
+                muf_dict = {t: v for t, v in muf_dict.items() if t >= cutoff_ts}
+                luf_dict = {t: v for t, v in luf_dict.items() if t >= cutoff_ts}
+
+                band_states = compute_band_states(fof2_dict, muf_dict, luf_dict)
+                ionosonde_data[ursi] = {
+                    "ursi": ursi,
+                    "name": name,
+                    "fof2": fof2_dict or None,
+                    "muf": muf_dict or None,
+                    "luf": luf_dict or None,
+                    "band_states": band_states,
+                }
+                updated_count += 1
+
+            self.update_data({"ionosonde_data": ionosonde_data})
+            self.status = "OK"
+            self.last_update_time = datetime.now(pytz.UTC)
+            logging.debug(f"Updated KC2G ionosonde data for {updated_count} stations.")
+
+        except ConnectionError:
+            logging.warning("Connection error when accessing KC2G ionosonde API.")
+        except (ConnectTimeout, ReadTimeout):
+            logging.warning(f"Timeout when accessing KC2G ionosonde API.")
+        except Exception:
+            self.status = "Error"
+            logging.exception("Exception in KC2G ionosonde data provider")
+            self._stop_event.wait(timeout=1)

providers/solarconditions/noaa3dayforecast.py

@@ -0,0 +1,174 @@
+import logging
+import re
+from datetime import datetime, timezone
+
+from providers.solarconditions.http_solar_conditions_provider import HTTPSolarConditionsProvider
+
+POLL_INTERVAL = 10800  # Every 3 hours
+URL = "https://services.swpc.noaa.gov/text/3-day-forecast.txt"
+
+
+class NOAA3dayForecast(HTTPSolarConditionsProvider):
+    """Solar conditions provider using the NOAA 3-day forecast text file. Parses the NOAA forecast and populates
+    corresponding fields in the solar conditions object.."""
+
+    def __init__(self, provider_config):
+        super().__init__("NOAA 3-day Forecast", provider_config, URL, POLL_INTERVAL)
+
+    @staticmethod
+    def _parse_percentage_table(lines, section_header, year):
+        """Find and parse a forecast table using percentages, identified by section_header. This is common to the lookup
+        of the solar storm and radio blackout forecast parsing."""
+
+        start_idx = None
+        for i, line in enumerate(lines):
+            if section_header in line:
+                start_idx = i
+                break
+        if start_idx is None:
+            logging.warning(f"NOAA 3-day forecast: could not find '{section_header}' section")
+            return None
+
+        # Find the date header line by scanning the next few lines for month & day patterns
+        date_header_idx = None
+        for j in range(start_idx + 1, min(start_idx + 6, len(lines))):
+            if re.search(r'[A-Za-z]{3}\s+\d{2}', lines[j]):
+                date_header_idx = j
+                break
+        if date_header_idx is None:
+            logging.warning(f"NOAA 3-day forecast: could not find date header after '{section_header}'")
+            return None
+        date_matches = re.findall(r'([A-Za-z]{3})\s+(\d{2})', lines[date_header_idx])
+        if not date_matches:
+            logging.warning(f"NOAA 3-day forecast: no dates in header: {lines[date_header_idx]}")
+            return None
+
+        # Figure out the date based on the line found
+        column_timestamps = []
+        for month_str, day_str in date_matches:
+            try:
+                dt = datetime.strptime(f"{day_str} {month_str} {year}", "%d %b %Y").replace(tzinfo=timezone.utc)
+                column_timestamps.append(dt.timestamp())
+            except ValueError:
+                logging.warning(f"NOAA 3-day forecast: could not parse date: {month_str} {day_str} {year}")
+                return None
+
+        # Parse data rows. Each non-empty line should have a text label followed by percentage values
+        result = {}
+        for line in lines[date_header_idx + 1:]:
+            line_stripped = line.strip()
+            if not line_stripped:
+                if result:
+                    break
+                continue
+            pct_matches = list(re.finditer(r'\b(\d+)%', line_stripped))
+            if not pct_matches:
+                if result:
+                    break
+                continue
+
+            # Row label is everything before the first percentage value
+            row_label = line_stripped[:line_stripped.index(pct_matches[0].group())].strip()
+            row_data = {}
+            for j, match in enumerate(pct_matches):
+                if j >= len(column_timestamps):
+                    break
+                row_data[column_timestamps[j]] = int(match.group(1))
+            if row_data:
+                result[row_label] = row_data
+
+        return result if result else None
+
+    def _http_response_to_solar_conditions(self, http_response):
+        lines = http_response.text.splitlines()
+
+        # Find the "NOAA Kp index breakdown" section header
+        start_idx = None
+        for i, line in enumerate(lines):
+            if "NOAA Kp index breakdown" in line:
+                start_idx = i
+                break
+        if start_idx is None:
+            logging.warning("NOAA K-index forecast: could not find 'NOAA Kp index breakdown' section")
+            return None
+
+        # Extract the year from the header line, e.g. "NOAA Kp index breakdown Apr 2-Apr 4, 2026"
+        header_line = lines[start_idx]
+        year_match = re.search(r'\b(\d{4})\b', header_line)
+        if not year_match:
+            logging.warning("NOAA K-index forecast: could not extract year from: " + header_line)
+            return None
+        year = int(year_match.group(1))
+
+        # Parse the column date headers on the next line, e.g. "              Apr 02     Apr 03     Apr 04"
+        if start_idx + 1 >= len(lines):
+            logging.warning("NOAA K-index forecast: missing date header line")
+            return None
+
+        date_header_line = lines[start_idx + 2]
+        date_matches = re.findall(r'([A-Za-z]{3})\s+(\d{2})', date_header_line)
+        if not date_matches:
+            logging.warning("NOAA K-index forecast: could not parse date headers from: " + date_header_line)
+            return None
+
+        column_dates = []
+        for month_str, day_str in date_matches:
+            try:
+                column_dates.append(datetime.strptime(f"{day_str} {month_str} {year}", "%d %b %Y").date())
+            except ValueError:
+                logging.warning(f"NOAA K-index forecast: could not parse date: {month_str} {day_str} {year}")
+                return None
+
+        # Parse each data row, e.g. "00-03UT          2.00          3.00          2.00"
+        k_index_forecast = {}
+        for line in lines[start_idx + 3:]:
+            time_match = re.match(r'^(\d{2})-(\d{2})UT\s+(.*)', line.strip())
+            if not time_match:
+                if k_index_forecast:
+                    break
+                continue
+
+            start_hour = int(time_match.group(1))
+            # Split on 2 or more spaces so that e.g. "5.67 (G2)" stays as one token per column
+            raw_values = re.split(r' {2,}', time_match.group(3).strip())
+
+            for i, val in enumerate(raw_values):
+                if i >= len(column_dates):
+                    break
+                # Take only the leading numeric part, discarding any bracketed section
+                try:
+                    kp = float(val.split()[0])
+                except (ValueError, IndexError):
+                    continue
+
+                date = column_dates[i]
+                start_dt = datetime(date.year, date.month, date.day, start_hour, 0, 0, tzinfo=timezone.utc)
+
+                # Key the data dict by start time
+                key = start_dt.timestamp()
+                k_index_forecast[key] = kp
+
+        if not k_index_forecast:
+            logging.warning("NOAA K-index forecast: no data rows parsed")
+            return None
+
+        # Parse Solar Radiation Storm Forecast (single row: "S1 or greater")
+        solar_storm_forecast = None
+        radiation_table = self._parse_percentage_table(lines, "Solar Radiation Storm Forecast", year)
+        if radiation_table:
+            solar_storm_forecast = radiation_table.get("S1 or greater")
+
+        # Parse Radio Blackout Forecast (two rows: "R1-R2" and "R3 or greater")
+        blackout_forecast_r1r2 = None
+        blackout_forecast_r3_or_greater = None
+        blackout_table = self._parse_percentage_table(lines, "Radio Blackout Forecast", year)
+        if blackout_table:
+            blackout_forecast_r1r2 = blackout_table.get("R1-R2")
+            blackout_forecast_r3_or_greater = blackout_table.get("R3 or greater")
+
+        return {
+            "k_index_forecast": k_index_forecast,
+            "solar_storm_forecast": solar_storm_forecast,
+            "blackout_forecast_r1r2": blackout_forecast_r1r2,
+            "blackout_forecast_r3_or_greater": blackout_forecast_r3_or_greater,
+        }

providers/solarconditions/solar_conditions_provider.py

@@ -0,0 +1,38 @@
+from datetime import datetime
+
+import pytz
+
+from core.data_store import DATA_STORE
+
+
+class SolarConditionsProvider:
+    """Generic solar conditions provider class. Subclasses of this query individual APIs for space weather and
+     propagation data."""
+
+    def __init__(self, name, provider_config):
+        """Constructor"""
+
+        self.name = name
+        self.enabled = provider_config["enabled"]
+        self.last_update_time = datetime.min.replace(tzinfo=pytz.UTC)
+        self.status = "Not Started" if self.enabled else "Disabled"
+        self._solar_conditions = DATA_STORE.solar_conditions
+
+    def start(self):
+        """Start the provider. This should return immediately after spawning threads to access the remote resources"""
+
+        raise NotImplementedError("Subclasses must implement this method")
+
+    def stop(self):
+        """Stop any threads and prepare for application shutdown"""
+
+        raise NotImplementedError("Subclasses must implement this method")
+
+    def update_data(self, new_data):
+        """Update the solar conditions object with new data"""
+
+        if new_data:
+            for key, value in new_data.items():
+                if hasattr(self._solar_conditions, key):
+                    setattr(self._solar_conditions, key, value)
+            self._solar_conditions.infer_descriptions()