Refactor of caching & data storage part 5 #118

alertproviders/alert_provider.py

@@ -2,27 +2,20 @@ from datetime import datetime
 
 import pytz
 
-from core.config import MAX_ALERT_AGE
+from core.data_store import DATA_STORE
 
 
 class AlertProvider:
     """Generic alert provider class. Subclasses of this query the individual APIs for alerts."""
 
-    def __init__(self, provider_config):
+    def __init__(self, name, provider_config):
         """Constructor"""
 
-        self.name = provider_config["name"]
+        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._alerts = None
-        self._web_server = None
-
-    def setup(self, alerts, web_server):
-        """Set up the provider, e.g. giving it the alert list to work from"""
-
-        self._alerts = alerts
-        self._web_server = web_server
+        self._alerts = DATA_STORE.alerts
 
     def start(self):
         """Start the provider. This should return immediately after spawning threads to access the remote resources"""
@@ -44,10 +37,7 @@ class AlertProvider:
 
     def _add_alert(self, alert):
         if not alert.expired():
-            self._alerts.add(alert.id, alert, expire=MAX_ALERT_AGE)
-            # Ping the web server in case we have any SSE connections that need to see this immediately
-            if self._web_server:
-                self._web_server.notify_new_alert(alert)
+            self._alerts.set(alert.id, alert)
 
     def stop(self):
         """Stop any threads and prepare for application shutdown"""

alertproviders/bota.py

@@ -15,12 +15,12 @@ class BOTA(HTTPAlertProvider):
     ALERTS_URL = "https://www.beachesontheair.com/"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("BOTA", provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_alerts(self, http_response):
         new_alerts = []
         # Find the table of upcoming alerts
-        bs = BeautifulSoup(http_response.content.decode(), features="lxml")
+        bs = BeautifulSoup(http_response.content.decode("utf-8-sig"), features="lxml")
         if not bs.body:
             return new_alerts
         div = bs.body.find('div', attrs={'class': 'view-activations-public'})

alertproviders/http_alert_provider.py

@@ -14,8 +14,8 @@ class HTTPAlertProvider(AlertProvider):
     """Generic alert provider class for providers that request data via HTTP(S). Just for convenience to avoid code
     duplication. Subclasses of this query the individual APIs for data."""
 
-    def __init__(self, provider_config, url, poll_interval):
-        super().__init__(provider_config)
+    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

alertproviders/ng3k.py

@@ -18,11 +18,11 @@ class NG3K(HTTPAlertProvider):
     AS_CALL_PATTERN = re.compile("as ([a-z0-9/]+)", re.IGNORECASE)
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("NG3K", provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_alerts(self, http_response):
         new_alerts = []
-        rss = cast(RSS, Parser.parse(http_response.content.decode()))
+        rss = cast(RSS, Parser.parse(http_response.content.decode("utf-8-sig")))
         # Iterate through source data
         for source_alert in rss.channel.items:
             # Deal with "the format"...

alertproviders/parksnpeaks.py

@@ -15,14 +15,14 @@ class ParksNPeaks(HTTPAlertProvider):
     ALERTS_URL = "https://parksnpeaks.org/api/ALERTS/"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("ParksNPeaks", provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_alerts(self, http_response):
         new_alerts = []
         # Iterate through source data
         for source_alert in http_response.json():
             # Calculate some things
-            sig = source_alert["Class"]
+            sig = source_alert["Class"].upper()
             if " - " in source_alert["Location"]:
                 split = source_alert["Location"].split(" - ")
                 sig_ref = split[0]
@@ -50,7 +50,7 @@ class ParksNPeaks(HTTPAlertProvider):
                           is_dxpedition=False)
 
             # Log a warning for the developer if PnP gives us an unknown programme we've never seen before
-            if sig and sig not in ["POTA", "SOTA", "WWFF", "SiOTA", "ZLOTA", "KRMNPA", "LLOTA", "QRP"]:
+            if sig and sig not in ["POTA", "SOTA", "WWFF", "SIOTA", "ZLOTA", "KRMNPA", "LLOTA", "QRP"]:
                 logging.warning("PNP alert found with sig " + sig + ", developer needs to add support for this!")
 
             # If this is POTA, SOTA or WWFF data we already have it through other means, so ignore. Otherwise, add to

alertproviders/pota.py

@@ -14,7 +14,7 @@ class POTA(HTTPAlertProvider):
     ALERTS_URL = "https://api.pota.app/activation"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("POTA", provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_alerts(self, http_response):
         new_alerts = []

alertproviders/sota.py

@@ -14,7 +14,7 @@ class SOTA(HTTPAlertProvider):
     ALERTS_URL = "https://api-db2.sota.org.uk/api/alerts/365/all/all"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("SOTA", provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_alerts(self, http_response):
         new_alerts = []

alertproviders/wota.py

@@ -18,11 +18,11 @@ class WOTA(HTTPAlertProvider):
     RSS_DATE_TIME_FORMAT = "%a, %d %b %Y %H:%M:%S %z"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("WOTA", provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_alerts(self, http_response):
         new_alerts = []
-        rss = cast(RSS, RSSParser.parse(http_response.content.decode()))
+        rss = cast(RSS, RSSParser.parse(http_response.content.decode("utf-8-sig")))
         # Iterate through source data
         for source_alert in rss.channel.items:
 

alertproviders/wwff.py

@@ -14,7 +14,7 @@ class WWFF(HTTPAlertProvider):
     ALERTS_URL = "https://spots.wwff.co/static/agendas.json"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("WWFF", provider_config, self.ALERTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_alerts(self, http_response):
         new_alerts = []

config-example.yml

@@ -19,66 +19,55 @@ base-url: "http://localhost:8080"
 
 # Spot providers to use. This is an example set, tailor it to your liking by commenting and uncommenting.
 # RBN and APRS-IS are supported but have such a high data rate, you probably don't want them enabled.
-# Each provider needs a class, a name, and an enabled/disabled state. Some require more config such as hostnames/IP
+# Each provider needs a class and an enabled/disabled state. Some require more config such as hostnames/IP
 # addresses and ports. You can duplicate them if you like, e.g. to support several DX clusters. RBN uses two ports, 7000
 # for CW/RTTY and 7001 for FT8, so if you want both, you need two entries, as shown below.
 # Feel free to write your own provider classes! There are details in the README.
 spot-providers:
   - class: "POTA"
-    name: "POTA"
     enabled: true
 
   - class: "SOTA"
-    name: "SOTA"
     enabled: true
 
   - class: "WWFF"
-    name: "WWFF"
     enabled: true
 
   - class: "WWBOTA"
-    name: "WWBOTA"
     enabled: true
 
   - class: "GMA"
-    name: "GMA"
     enabled: true
     # GMA requires an API key to fetch spots. After creating an account on cqgma.org, email support and request one.
     api-key: ""
 
   - class: "HEMA"
-    name: "HEMA"
     enabled: true
 
   - class: "ParksNPeaks"
-    name: "ParksNPeaks"
     enabled: true
 
   - class: "ZLOTA"
-    name: "ZLOTA"
     enabled: true
 
   - class: "WOTA"
-    name: "WOTA"
     enabled: true
 
   - class: "LLOTA"
-    name: "LLOTA"
     enabled: true
 
   - class: "Towers"
-    name: "Towers"
     enabled: true
 
   - class: "Tiles"
-    name: "Tiles"
     enabled: true
 
   - class: "APRSIS"
-    name: "APRS-IS"
     enabled: false
 
   - class: "DXCluster"
+    # Clusters have a "name" property in case you want to say which cluster node the data is from, or connect to
+    # several clusters and name them separately.
     name: "HRD Cluster"
     enabled: true
     host: "hrd.wa9pie.net"
@@ -99,18 +88,12 @@ spot-providers:
     enabled: false
     host: "w3lpl.net"
     port: 7373
-    # Prompt the cluster node gives when asking for a callsign to log in. Varies between cluster node software.
     login_prompt: "Please enter your call:"
-    # Callsign Spothole will use to log into this cluster. Ensure the SSID (e.g. -99) is different to any personal
-    # connection you might make to this cluster node.
     login_callsign: "N0CALL-99"
-    # Whether to allow RBN spots that come via this cluster. If you don't want RBN spots or you are making a separate
-    # connection to RBN directly, leave this as False. If you want RBN spots from this cluster, set this to True. (Make
-    # sure you aren't also separately connecting to RBN directly, otherwise you may get duplicate spots.) Note that not
-    # all clusters sent RBN spots anyway.
     allow_rbn_spots: false
 
   - class: "RBN"
+  - # RBN sources have a "name" property in case you want to differentiate between the CW/RTTY and FT8 sources.
     name: "RBN CW/RTTY"
     enabled: false
     port: 7000
@@ -127,76 +110,129 @@ spot-providers:
     enabled-by-default-in-web-ui: false
 
   - class: "UKPacketNet"
-    name: "UK Packet Radio Net"
     enabled: false
     enabled-by-default-in-web-ui: false
 
   - class: "XOTA"
-    name: "39C3 TOTA"
+    # xOTA sources have a "name" property so you can define what programme it is for, since xOTA is generic software
+    # for running "something-on-the-air" programmes.
+    name: "C3 TOTA"
     enabled: false
     url: "wss://39c3.totawatch.de/api/spot/live"
-    # Fixed SIG for all spots from a provider & location CSV are currently only a feature for the "XOTA" provider,
-    # the software found at https://github.com/nischu/xOTA/. This is because this is a generic backend for xOTA
-    # programmes and so different URLs provide different programmes.
-    sig: "TOTA"
-    locations-csv: "datafiles/39c3-tota.csv"
+    # For the "XOTA" provider, a SIG must be set menually here because xOTA is a generic backend for xOTA
+    # programmes and so different URLs potentially provide different programmes.
+    sig: "Toilets"
+    # For Toilets on the Air, we prefix the SIG references (T-01 etc) with some characters that define the conference:
+    # C3, EH or HOPE - so we can look up the correct locations in our database, because each conference starts from T-01
+    # but refers to a toilet in a different building (or continent!)
+    sig-ref-prefix: "C3"
 
   - class: "XOTA"
     name: "EH23 TOTA"
     enabled: false
     url: "wss://eh23.totawatch.de/api/spot/live"
-    sig: "TOTA"
-    locations-csv: "datafiles/eh23-tota.csv"
+    sig: "Toilets"
+    sig-ref-prefix: "EH"
+
+  - class: "XOTA"
+    name: "HOPE26 TOTA"
+    enabled: false
+    url: "wss://hope-26.totawatch.de/api/spot/live"
+    sig: "Toilets"
+    sig-ref-prefix: "HOPE"
 
 
 # Alert providers to use. Same setup as the spot providers list above.
 alert-providers:
   - class: "POTA"
-    name: "POTA"
     enabled: true
 
   - class: "SOTA"
-    name: "SOTA"
     enabled: true
 
   - class: "WWFF"
-    name: "WWFF"
     enabled: true
 
   - class: "ParksNPeaks"
-    name: "ParksNPeaks"
     enabled: true
 
   - class: "WOTA"
-    name: "WOTA"
     enabled: true
 
   - class: "BOTA"
-    name: "BOTA"
     enabled: true
 
   - class: "NG3K"
-    name: "NG3K"
     enabled: true
 
 
+# SIG reference data providers to use. This allows Spothole to download, for example, the WWFF directory that maps WWFF
+# park IDs to their name and location.
+sig-ref-data-providers:
+  - class: "POTA"
+    enabled: true
+
+  - class: "SOTA"
+    enabled: true
+
+  - class: "WWFF"
+    enabled: true
+
+  - class: "WWBOTA"
+    enabled: true
+
+  - class: "GMA"
+    enabled: true
+
+  - class: "MOTA"
+    enabled: true
+
+  - class: "ILLW"
+    enabled: true
+
+  - class: "ARLHS"
+    enabled: true
+
+  - class: "WCA"
+    enabled: true
+
+  - class: "IOTA"
+    enabled: true
+
+  - class: "WOTA"
+    enabled: true
+
+  - class: "SIOTA"
+    enabled: true
+
+  - class: "ZLOTA"
+    enabled: true
+
+  - class: "LLOTA"
+    enabled: true
+
+  - class: "Towers"
+    enabled: true
+
+  - class: "DME"
+    enabled: true
+
+  - class: "Toilets"
+    enabled: true
+
 # Solar condition providers to use. These poll external APIs for solar propagation data (SFI, A/K indices, band
 # conditions, etc.) and make it available via the /api/v1/solar endpoint.
 solar-condition-providers:
   - class: "HamQSL"
-    name: "HamQSL"
     enabled: true
 
   - class: "NOAA3dayForecast"
-    name: "NOAA 3-day Forecast"
     enabled: true
 
   - class: "GIROIonosonde"
-    name: "GIRO Ionosonde Data"
     enabled: true
 
   - class: "KC2GProp"
-    name: "KC2G Propagation Data"
     enabled: true
 
 # Maximum time to keep spots and alerts in the system before deleting them. By default, one hour for spots and one week

core/cache_utils.py

@@ -1,27 +0,0 @@
-import threading
-from datetime import timedelta
-
-from requests_cache import CachedSession
-
-# Cache for "semi-static" data such as the locations of parks, CSVs of reference lists, etc.
-# This has an expiry time of 30 days, so will re-request from the source after that amount
-# of time has passed. This is used throughout Spothole to cache data that does not change
-# rapidly. The ThreadSafeSession construct here protects it against some multithreading
-# contention weirdness we sometimes used to see on startup where the cache was hammered
-# pretty hard. The expanded list of allowable_codes ensures we also cache and return 400-type
-# responses, e.g "this SOTA summit ref doesn't actually exist", to avoid hammering remote
-# servers for data they've told us they can't provide.
-_session = CachedSession("cache/semi_static_url_data_cache", expire_after=timedelta(days=30),
-                         allowable_codes=(200, 400, 401, 403, 404))
-_lock = threading.Lock()
-
-
-class _ThreadSafeSession:
-    """Wraps CachedSession with a lock to prevent concurrent SQLite access across threads."""
-
-    def get(self, *args, **kwargs):
-        with _lock:
-            return _session.get(*args, **kwargs)
-
-
-SEMI_STATIC_URL_DATA_CACHE = _ThreadSafeSession()

core/cleanup.py

@@ -1,73 +0,0 @@
-import logging
-from datetime import datetime
-from threading import Event, Thread
-
-import pytz
-
-
-class CleanupTimer:
-    """Provides a timed cleanup of the spot list."""
-
-    def __init__(self, spots, alerts, web_server, cleanup_interval):
-        """Constructor"""
-
-        self._spots = spots
-        self._alerts = alerts
-        self._web_server = web_server
-        self._cleanup_interval = cleanup_interval
-        self.last_cleanup_time = datetime.min.replace(tzinfo=pytz.UTC)
-        self.status = "Starting"
-        self._thread = None
-        self._stop_event = Event()
-
-    def start(self):
-        """Start the cleanup timer"""
-
-        self._thread = Thread(target=self._run, daemon=True)
-        self._thread.start()
-
-    def stop(self):
-        """Stop any threads and prepare for application shutdown"""
-
-        self._stop_event.set()
-
-    def _run(self):
-        while not self._stop_event.wait(timeout=self._cleanup_interval):
-            self._cleanup()
-
-    def _cleanup(self):
-        """Perform cleanup and reschedule next timer"""
-
-        try:
-            # Perform cleanup via letting the data expire
-            self._spots.expire()
-            self._alerts.expire()
-
-            # Explicitly clean up any spots and alerts that have expired
-            for i in list(self._spots.iterkeys()):
-                try:
-                    spot = self._spots[i]
-                    if spot.expired():
-                        self._spots.delete(i)
-                except KeyError:
-                    # Must have already been deleted, OK with that
-                    pass
-            for i in list(self._alerts.iterkeys()):
-                try:
-                    alert = self._alerts[i]
-                    if alert.expired():
-                        self._alerts.delete(i)
-                except KeyError:
-                    # Must have already been deleted, OK with that
-                    pass
-
-            # Clean up web server SSE spot/alert queues
-            self._web_server.clean_up_sse_queues()
-
-            self.status = "OK"
-            self.last_cleanup_time = datetime.now(pytz.UTC)
-
-        except Exception:
-            self.status = "Error"
-            logging.exception("Exception in Cleanup thread")
-            self._stop_event.wait(timeout=1)

core/config.py

@@ -1,3 +1,4 @@
+import importlib
 import logging
 import os
 
@@ -33,3 +34,35 @@ WEB_UI_OPTIONS["spot-providers-enabled-by-default"] = [p["name"] for p in config
 # one of our proviers. We set that to also be enabled by default.
 if ALLOW_SPOTTING:
     WEB_UI_OPTIONS["spot-providers-enabled-by-default"].append("API")
+
+
+def get_spot_provider_from_config(config_providers_entry):
+    """Utility method to get a spot provider based on the class specified in its config entry."""
+
+    module = importlib.import_module('spotproviders.' + config_providers_entry["class"].lower())
+    provider_class = getattr(module, config_providers_entry["class"])
+    return provider_class(config_providers_entry)
+
+
+def get_alert_provider_from_config(config_providers_entry):
+    """Utility method to get an alert provider based on the class specified in its config entry."""
+
+    module = importlib.import_module('alertproviders.' + config_providers_entry["class"].lower())
+    provider_class = getattr(module, config_providers_entry["class"])
+    return provider_class(config_providers_entry)
+
+
+def get_solar_conditions_provider_from_config(config_providers_entry):
+    """Utility method to get a solar conditions provider based on the class specified in its config entry."""
+
+    module = importlib.import_module('solarconditionsproviders.' + config_providers_entry["class"].lower())
+    provider_class = getattr(module, config_providers_entry["class"])
+    return provider_class(config_providers_entry)
+
+
+def get_sig_ref_data_provider_from_config(config_providers_entry):
+    """Utility method to get a SIG reference data provider based on the class specified in its config entry."""
+
+    module = importlib.import_module('sigrefdataproviders.' + config_providers_entry["class"].lower())
+    provider_class = getattr(module, config_providers_entry["class"])
+    return provider_class(config_providers_entry)

core/constants.py

@@ -19,9 +19,9 @@ SIGS = [
     SIG(name="HEMA", comment_names=["HEMA"], description="HuMPs Excluding Marilyns Award", ref_regex=r"[A-Z0-9]{1,3}\/[A-Z]{3}\-\d{3}"),
     SIG(name="IOTA", comment_names=["IOTA"], description="Islands on the Air", ref_regex=r"[A-Z]{2}\-\d{3}"),
     SIG(name="MOTA", comment_names=["MOTA"], description="Mills on the Air", ref_regex=r"X\d{4,6}"),
-    SIG(name="ARLHS", comment_names=["ARLHS"], description="Amateur Radio Lighthouse Society", ref_regex=r"[A-Z]{3}\-\d{3,4}"),
+    SIG(name="ARLHS", comment_names=["ARLHS"], description="Amateur Radio Lighthouse Society", ref_regex=r"[A-Z]{3}[\- ]\d{3,4}"),
     SIG(name="ILLW", comment_names=["ILLW"], description="International Lighthouse & Lightship Weekend", ref_regex=r"[A-Z]{2}\d{4}"),
-    SIG(name="SiOTA", comment_names=["SIOTA"], description="Silos on the Air", ref_regex=r"[A-Z]{2}\-[A-Z]{3}\d"),
+    SIG(name="SIOTA", comment_names=["SIOTA"], description="Silos on the Air", ref_regex=r"[A-Z]{2}\-[A-Z]{3}\d"),
     SIG(name="WCA", comment_names=["WCA"], description="World Castles Award", ref_regex=r"[A-Z0-9]{1,3}\-\d{5}"),
     SIG(name="ZLOTA", comment_names=["ZLOTA"], description="New Zealand on the Air", ref_regex=r"ZL[A-Z]/[A-Z]{2}\-\d{3,4}"),
     SIG(name="WOTA", comment_names=["WOTA"], description="Wainwrights on the Air", ref_regex=r"[A-Z]{3}-[0-9]{2}"),

core/data_store.py

@@ -0,0 +1,80 @@
+import logging
+from pathlib import Path
+
+import diskcache
+
+from core.config import MAX_SPOT_AGE, MAX_ALERT_AGE
+from core.constants import SIGS
+from core.live_data_cache import LiveDataCache
+from data.solar_conditions import SolarConditions
+
+
+class DataStore:
+    """Data caching/storage object. Handles storage of spots, alerts, solar conditions, SIG reference data, and callsign
+    lookup data using different caching strategies for each."""
+
+    def __init__(self):
+        self._CACHE_DIR = "./cache"
+        self._MAX_SPOT_COUNT = 100000
+        self._MAX_ALERT_COUNT = 100000
+        self._SPOT_ALERT_SNAPSHOT_INTERVAL_SEC = 300
+        self._CALLSIGN_DATA_TTL_SEC = 30 * 24 * 60 * 60
+        self.alerts = None
+        self.spots = None
+        self.callsigns = None
+        self.sigrefs = None
+        self.status_data = None
+        self._status = None
+        self.solar_conditions = None
+        self._solar = None
+
+    def setup(self):
+        Path(self._CACHE_DIR).mkdir(parents=True, exist_ok=True)
+
+        # Standard disk cache for solar data and status data, but each cache contains only a single object which we
+        # expose to the wider application
+        self._solar = diskcache.Cache(self._CACHE_DIR + "/solar")
+        if "solar_conditions" not in self._solar:
+            self._solar.add("solar_conditions", SolarConditions())
+        self.solar_conditions = self._solar.get("solar_conditions")
+        self._status = diskcache.Cache(self._CACHE_DIR + "/status")
+        if "status_data" not in self._status:
+            self._status.add("status_data", {})
+        self.status_data = self._status.get("status_data")
+
+        # Standard disk cache for SIG ref data. Separate provider threads will repopulate theis on a regular basis
+        # but there's no need for a TTL since old data is better than no data. We need to key on both SIG and reference,
+        # and trying to do two layers of dict in diskcache absolutely destroys performance with unpickling huge dicts,
+        # so we have an ugly "SIG:ref" syntax for keys to keep it a single level.
+        self.sigrefs = diskcache.Cache(self._CACHE_DIR + "/sigrefs")
+        logging.info(f"Loaded data for %d SIG references.", len(self.sigrefs))
+
+        # Standard disk cache for callsign data. This data does have a TTL to trigger an occasional re-lookup.
+        # Old data *is* better than no data, but we can't have a background thread re-looking-up every callsign
+        # we've seen, so we rely on them timing out and this triggering another lookup.
+        self.callsigns = diskcache.Cache(self._CACHE_DIR + "/callsigns")
+        logging.info(f"Loaded data for %d callsigns.", len(self.callsigns))
+
+        # Special caches for spots and alerts, which have TTL and write snapshots to disk at an interval. We
+        # specifically load these caches *last* so that any sigref and callsign data is already loaded from disk cache
+        # before the spots and alerts are live in the system.
+        self.spots = LiveDataCache(maxsize=self._MAX_SPOT_COUNT, ttl=MAX_SPOT_AGE,
+                                   snapshot_dir=self._CACHE_DIR + "/spots",
+                                   snapshot_interval_sec=self._SPOT_ALERT_SNAPSHOT_INTERVAL_SEC)
+        logging.info(f"Loaded %d spots from a previous run.", len(self.spots.keys()))
+
+        self.alerts = LiveDataCache(maxsize=self._MAX_ALERT_COUNT, ttl=MAX_ALERT_AGE,
+                                    snapshot_dir=self._CACHE_DIR + "/alerts",
+                                    snapshot_interval_sec=self._SPOT_ALERT_SNAPSHOT_INTERVAL_SEC)
+        logging.info(f"Loaded %d alerts from a previous run.", len(self.alerts.keys()))
+
+    def close(self):
+        self.spots.close()
+        self.alerts.close()
+        self._solar.close()
+        self._status.close()
+        self.sigrefs.close()
+        self.callsigns.close()
+
+# Global object
+DATA_STORE = DataStore()

core/live_data_cache.py

@@ -0,0 +1,100 @@
+import logging
+import threading
+import time
+
+import diskcache
+from cachetools import TTLCache
+
+
+class LiveDataCache:
+    """Cache for spots and alerts. Uses the fast in-memory TTLCache for normal data I/O, including the TTL to enforce
+    maximum lifetime, and adds a separate diskcache to which we can save and load the TTLCache to provide persistence.
+    Also adds thread safety so spots and alerts can come from any thread, and a listener mechanism so the web server
+    can get a callback when new spots/alerts are added, and send them to any SSE clients."""
+
+    def __init__(self, maxsize, ttl, snapshot_dir, snapshot_interval_sec):
+        self._cache = TTLCache(maxsize=maxsize, ttl=ttl)
+        self._lock = threading.Lock()
+        self._ttl = ttl
+        self._listeners = []
+        self._listeners_lock = threading.Lock()
+        self._snapshot_dir = snapshot_dir
+        self._disk_cache = diskcache.Cache(str(snapshot_dir))
+        self._load_snapshot()
+        self._start_periodic_snapshot(snapshot_interval_sec)
+
+    def set(self, key, value):
+        with self._lock:
+            self._cache[key] = value
+
+        # Notify listeners
+        with self._listeners_lock:
+            listeners = list(self._listeners)
+        for callback in listeners:
+            try:
+                callback(value)
+            except Exception:
+                logging.error("Listener raised an exception for key %s", key, exc_info=True)
+
+
+    def get(self, key, default=None):
+        with self._lock:
+            return self._cache.get(key, default)
+
+    def delete(self, key):
+        with self._lock:
+            self._cache.pop(key, None)
+
+    def keys(self):
+        with self._lock:
+            return list(self._cache.keys())
+
+    def values(self):
+        with self._lock:
+            return list(self._cache.values())
+
+    def add_listener(self, callback):
+        """Register callback(value) which will be called whenever a new spot/alert item is added via set(). Used by the
+        web server (via SSEBroadcaster) to send SSE clients an update on every new spot."""
+
+        with self._listeners_lock:
+            self._listeners.append(callback)
+
+    def remove_listener(self, callback):
+        with self._listeners_lock:
+            self._listeners.remove(callback)
+
+    def save_snapshot(self):
+        with self._lock:
+            # Store the time with the data so we can avoid loading anything nxt time that's older than TTL
+            data = [(k, v, time.time()) for k, v in self._cache.items()]
+        try:
+            self._disk_cache.set("snapshot", data)
+        except Exception as e:
+            logging.error("Failed to write snapshot to %s", self._snapshot_dir, e, exc_info=True)
+
+    def _load_snapshot(self):
+        data = self._disk_cache.get("snapshot")
+        if not data:
+            return
+
+        now = time.time()
+        with self._lock:
+            for key, value, saved_at in data:
+                # Only restore entries that would still be within TTL
+                if now - saved_at < self._ttl:
+                    self._cache[key] = value
+        logging.info("Loaded snapshot from %s", self._snapshot_dir)
+
+    def _start_periodic_snapshot(self, interval):
+        def loop():
+            while True:
+                time.sleep(interval)
+                self.save_snapshot()
+
+        t = threading.Thread(target=loop, daemon=True, name=f"snapshot-{self._snapshot_dir}")
+        t.start()
+
+    def close(self):
+        self.save_snapshot()
+        self._disk_cache.close()

core/lookup_helper.py

@@ -14,10 +14,10 @@ from pyhamtools.locator import latlong_to_locator
 from requests.exceptions import ConnectionError, ReadTimeout, ConnectTimeout
 from requests_cache import CachedSession
 
-from core.cache_utils import SEMI_STATIC_URL_DATA_CACHE
 from core.config import config
 from core.constants import BANDS, UNKNOWN_BAND, CW_MODES, PHONE_MODES, DATA_MODES, ALL_MODES, \
     HTTP_HEADERS, HAMQTH_PRG, MODE_ALIASES
+from core.url_data_cache import URL_DATA_CACHE
 
 # QRZ XML field names differ from pyhamtools' normalised names; map them here.
 _QRZ_FIELD_MAP = {
@@ -142,8 +142,8 @@ class LookupHelper:
 
         try:
             logging.info("Downloading Country-files.com cty.plist...")
-            response = SEMI_STATIC_URL_DATA_CACHE.get("https://www.country-files.com/cty/cty.plist",
-                                                      headers=HTTP_HEADERS)
+            response = URL_DATA_CACHE.get("https://www.country-files.com/cty/cty.plist",
+                                          headers=HTTP_HEADERS)
 
             if response.ok:
                 with open(self._country_files_cty_plist_download_location, "w") as f:
@@ -167,7 +167,7 @@ class LookupHelper:
 
         try:
             logging.info("Downloading dxcc.json...")
-            response = SEMI_STATIC_URL_DATA_CACHE.get(
+            response = URL_DATA_CACHE.get(
                 "https://raw.githubusercontent.com/k0swe/dxcc-json/refs/heads/main/dxcc.json",
                 headers=HTTP_HEADERS)
 
@@ -515,7 +515,7 @@ class LookupHelper:
 
         for lookup_call in calls_to_try:
             try:
-                response = SEMI_STATIC_URL_DATA_CACHE.get(
+                response = URL_DATA_CACHE.get(
                     self._qrz_base_url + "?s=" + session_key + "&callsign=" + urllib.parse.quote_plus(lookup_call),
                     headers=HTTP_HEADERS, timeout=10)
                 if response.ok:
@@ -593,7 +593,7 @@ class LookupHelper:
 
         for lookup_call in calls_to_try:
             try:
-                response = SEMI_STATIC_URL_DATA_CACHE.get(
+                response = URL_DATA_CACHE.get(
                     self._hamqth_base_url + "?id=" + session_id + "&callsign=" + urllib.parse.quote_plus(
                         lookup_call) + "&prg=" + HAMQTH_PRG, headers=HTTP_HEADERS)
                 if response.ok:

core/sig_utils.py

@@ -1,26 +1,11 @@
-import csv
 import logging
 
 from pyhamtools.locator import latlong_to_locator, locator_to_latlong
-from requests.exceptions import ConnectionError, ReadTimeout, ConnectTimeout
 
-from core.cache_utils import SEMI_STATIC_URL_DATA_CACHE
-from core.constants import SIGS, HTTP_HEADERS
+from core.constants import SIGS
+from core.data_store import DATA_STORE
 from core.geo_utils import wab_wai_square_to_lat_lon
 
-# Load Spanish municipality data for the DME programme. There's no convenient lookup API for this, so we embed the data
-# file in Spothole and load it on startup.
-with open("datafiles/MUNICIPIOS.csv", encoding="latin-1") as _f:
-    _DME_INDEX = {row["COD_INE"][:5]: row for row in csv.DictReader(_f, delimiter=";")}
-# Caches for data for the SIGs where we have to download a whole global reference list, rather than looking up a single
-# reference. These get populated from the SEMI_STATIC_URL_DATA_CACHE only if the data actually came
-# live from the internet, to avoid repopulating them every time we pull the same data from the cache.
-_WWFF_INDEX_CACHE = {}
-_SIOTA_INDEX_CACHE = {}
-_WOTA_INDEX_CACHE = {}
-_ZLOTA_INDEX_CACHE = {}
-_LLOTA_INDEX_CACHE = {}
-
 
 def get_ref_regex_for_sig(sig):
     """Utility function to get the regex string for a SIG reference for a named SIG. If no match is found, None will be returned."""
@@ -42,220 +27,31 @@ def get_sig_name_from_comment_name(sig):
 
 
 def populate_sig_ref_info(sig_ref):
-    """Look up details of a SIG reference (e.g. POTA park) such as name, lat/lon, and grid. Takes in a sig_ref object which
-    must at minimum have a "sig" and an "id". The rest of the object will be populated and returned.
+    """Look up details of a SIG reference (e.g. POTA park) such as name, lat/lon, and grid. Takes in a sig_ref object
+    which must at minimum have a "sig" and an "id". The rest of the object will be populated and returned. This makes
+    use of SIG ref data in the data store, live lookups from the web, or just automatic calculation depending on which
+    SIG we are getting data for.
     Note there is currently no support for KRMNPA location lookup, see issue #61."""
 
-    if sig_ref.sig is None or sig_ref.id is None:
-        logging.warning("Failed to look up sig_ref info, sig or id were not set.")
+    if sig_ref.sig is None or sig_ref.sig == "" or sig_ref.id is None or sig_ref.id == "":
+        logging.debug("Failed to look up sig_ref info, sig or id were not set.")
+        return sig_ref
 
-    sig = sig_ref.sig or ""
+    sig = sig_ref.sig
     ref_id = sig_ref.id
+
+    # DME fudge. Our database has leading zeros padding to 5 digits which is the expected format, but not all activators
+    # add leading zeros.
+    if sig.upper() == "DME":
+        ref_id = ref_id.zfill(5)
+
     try:
-        if sig.upper() == "POTA":
-            response = SEMI_STATIC_URL_DATA_CACHE.get("https://api.pota.app/park/" + ref_id, headers=HTTP_HEADERS)
-            if response.ok:
-                data = response.json()
-                if data:
-                    fullname = str(data["name"]) if "name" in data else None
-                    if fullname and "parktypeDesc" in data and data["parktypeDesc"] != "":
-                        fullname = fullname + " " + data["parktypeDesc"]
-                    sig_ref.name = fullname
-                    sig_ref.url = "https://pota.app/#/park/" + ref_id
-                    sig_ref.grid = data["grid6"] if "grid6" in data else None
-                    sig_ref.latitude = data["latitude"] if "latitude" in data else None
-                    sig_ref.longitude = data["longitude"] if "longitude" in data else None
-                elif not response.from_cache:
-                    logging.warning("Malformed response looking up %s ref %s", sig, ref_id)
-            elif not response.from_cache:
-                logging.warning("HTTP %d looking up %s ref %s", response.status_code, sig, ref_id)
-
-        elif sig.upper() == "SOTA":
-            response = SEMI_STATIC_URL_DATA_CACHE.get("https://api-db2.sota.org.uk/api/summits/" + ref_id,
-                                                      headers=HTTP_HEADERS)
-            if response.ok:
-                data = response.json()
-                if data:
-                    sig_ref.name = data["name"] if "name" in data else None
-                    sig_ref.url = "https://www.sotadata.org.uk/en/summit/" + ref_id
-                    sig_ref.grid = data["locator"] if "locator" in data else None
-                    sig_ref.latitude = data["latitude"] if "latitude" in data else None
-                    sig_ref.longitude = data["longitude"] if "longitude" in data else None
-                    sig_ref.activation_score = data["points"] if "points" in data else None
-                elif not response.from_cache:
-                    logging.warning("Malformed response looking up %s ref %s", sig, ref_id)
-            elif not response.from_cache:
-                logging.warning("HTTP %d looking up %s ref %s", response.status_code, sig, ref_id)
-
-        elif sig.upper() == "WWBOTA":
-            response = SEMI_STATIC_URL_DATA_CACHE.get("https://api.wwbota.org/bunkers/" + ref_id,
-                                                      headers=HTTP_HEADERS)
-            if response.ok:
-                data = response.json()
-                if data:
-                    sig_ref.name = data["name"] if "name" in data else None
-                    sig_ref.url = "https://bunkerwiki.org/?s=" + ref_id if ref_id.startswith("B/G") else None
-                    sig_ref.grid = data["locator"] if "locator" in data else None
-                    sig_ref.latitude = data["lat"] if "lat" in data else None
-                    sig_ref.longitude = data["long"] if "long" in data else None
-                elif not response.from_cache:
-                    logging.warning("Malformed response looking up %s ref %s", sig, ref_id)
-            elif not response.from_cache:
-                logging.warning("HTTP %d looking up %s ref %s", response.status_code, sig, ref_id)
-
-        elif sig.upper() == "GMA" or sig.upper() == "ARLHS" or sig.upper() == "ILLW" or sig.upper() == "WCA" or sig.upper() == "MOTA" or sig.upper() == "IOTA":
-            response = SEMI_STATIC_URL_DATA_CACHE.get("https://www.cqgma.org/api/ref/?" + ref_id,
-                                                      headers=HTTP_HEADERS)
-            if response.ok:
-                data = response.json()
-                if data:
-                    sig_ref.name = data["name"] if "name" in data else None
-                    sig_ref.url = "https://www.cqgma.org/zinfo.php?ref=" + ref_id
-                    sig_ref.grid = data["locator"] if "locator" in data else None
-
-                    # For some things (just IOTA?) the GMA actually returns a box where "latitude" and "longitude" are
-                    # the zeroest corner of the box, then "lat2" and "lng2" provide the other corner. We detect this
-                    # and provide a single lat/lon for the centre. Otherwise if we don't have these extra parameters,
-                    # just use the single point we have.
-                    if data.get("latitude") is not None and data.get("longitude") is not None and data.get(
-                            "lat2") is not None and data.get("lng2") is not None:
-                        sig_ref.latitude = (float(data["latitude"]) + float(data["lat2"])) / 2.0
-                        sig_ref.longitude = (float(data["longitude"]) + float(data["lng2"])) / 2.0
-                    else:
-                        sig_ref.latitude = float(data["latitude"]) if data.get("latitude") is not None else None
-                        sig_ref.longitude = float(data["longitude"]) if data.get("longitude") is not None else None
-                elif not response.from_cache:
-                    logging.warning("Malformed response looking up %s ref %s via GMA", sig, ref_id)
-            elif not response.from_cache:
-                logging.warning("HTTP %d looking up %s ref %s", response.status_code, sig, ref_id)
-
-        elif sig.upper() == "WWFF":
-            response = SEMI_STATIC_URL_DATA_CACHE.get("https://wwff.co/wwff-data/wwff_directory.csv",
-                                                      headers=HTTP_HEADERS)
-            if response.ok:
-                global _WWFF_INDEX_CACHE
-                if not bool(_WWFF_INDEX_CACHE) or not response.from_cache:
-                    # New data from WWFF, update our internal map
-                    _WWFF_INDEX_CACHE = {row["reference"]: row for row in
-                                         csv.DictReader(response.content.decode().splitlines())}
-                row = _WWFF_INDEX_CACHE.get(ref_id)
-                if row:
-                    sig_ref.name = row["name"] if "name" in row else None
-                    sig_ref.url = "https://wwff.co/directory/?showRef=" + ref_id
-                    sig_ref.grid = row["iaruLocator"] if "iaruLocator" in row and row["iaruLocator"] != "-" else None
-                    sig_ref.latitude = float(row["latitude"]) if "latitude" in row and row["latitude"] != "-" else None
-                    sig_ref.longitude = float(row["longitude"]) if "longitude" in row and row[
-                        "longitude"] != "-" else None
-                elif not response.from_cache:
-                    logging.warning("WWFF database did not contain data for ref %s", ref_id)
-            elif not response.from_cache:
-                logging.warning("HTTP %d looking up %s ref %s", response.status_code, sig, ref_id)
-
-        elif sig.upper() == "SIOTA":
-            response = SEMI_STATIC_URL_DATA_CACHE.get("https://www.silosontheair.com/data/silos.csv",
-                                                      headers=HTTP_HEADERS)
-            if response.ok:
-                global _SIOTA_INDEX_CACHE
-                if not bool(_SIOTA_INDEX_CACHE) or not response.from_cache:
-                    # New data from SIOTA, update our internal map
-                    _SIOTA_INDEX_CACHE = {row["SILO_CODE"]: row for row in
-                                          csv.DictReader(response.content.decode().splitlines())}
-                row = _SIOTA_INDEX_CACHE.get(ref_id)
-                if row:
-                    sig_ref.name = row["NAME"] if "NAME" in row else None
-                    sig_ref.grid = row["LOCATOR"] if "LOCATOR" in row else None
-                    sig_ref.latitude = float(row["LAT"]) if "LAT" in row else None
-                    sig_ref.longitude = float(row["LNG"]) if "LNG" in row else None
-                elif not response.from_cache:
-                    logging.warning("SIOTA database did not contain data for ref %s", ref_id)
-            elif not response.from_cache:
-                logging.warning("HTTP %d looking up %s ref %s", response.status_code, sig, ref_id)
-
-        elif sig.upper() == "WOTA":
-            response = SEMI_STATIC_URL_DATA_CACHE.get("https://www.wota.org.uk/mapping/data/summits.json",
-                                                      headers=HTTP_HEADERS)
-            if response.ok:
-                data = response.json()
-                if data:
-                    global _WOTA_INDEX_CACHE
-                    if not bool(_WOTA_INDEX_CACHE) or not response.from_cache:
-                        # New data from WOTA, update our internal map
-                        _WOTA_INDEX_CACHE = {feature["properties"]["wotaId"]: feature for feature in
-                                             data.get("features", [])}
-                    feature = _WOTA_INDEX_CACHE.get(ref_id)
-                    if feature:
-                        sig_ref.name = feature["properties"]["title"]
-                        # Fudge WOTA URLs. Outlying fell (LDO) URLs don't match their ID numbers but require 214 to be
-                        # added to them
-                        sig_ref.url = "https://www.wota.org.uk/MM_" + ref_id
-                        if ref_id.upper().startswith("LDO-"):
-                            number = int(ref_id.upper().replace("LDO-", ""))
-                            sig_ref.url = "https://www.wota.org.uk/MM_LDO-" + str(number + 214)
-                        sig_ref.grid = feature["properties"]["qthLocator"]
-                        sig_ref.latitude = feature["geometry"]["coordinates"][1]
-                        sig_ref.longitude = feature["geometry"]["coordinates"][0]
-                elif not response.from_cache:
-                    logging.warning("Malformed response looking up %s ref %s", sig, ref_id)
-            elif not response.from_cache:
-                logging.warning("HTTP %d looking up %s ref %s", response.status_code, sig, ref_id)
-
-        elif sig.upper() == "ZLOTA":
-            response = SEMI_STATIC_URL_DATA_CACHE.get("https://ontheair.nz/assets/assets.json", headers=HTTP_HEADERS)
-            if response.ok:
-                data = response.json()
-                if isinstance(data, list):
-                    global _ZLOTA_INDEX_CACHE
-                    if not bool(_ZLOTA_INDEX_CACHE) or not response.from_cache:
-                        # New data from ZLOTA, update our internal map
-                        _ZLOTA_INDEX_CACHE = {asset["code"]: asset for asset in data}
-                    asset = _ZLOTA_INDEX_CACHE.get(ref_id)
-                    if asset:
-                        sig_ref.name = asset["name"]
-                        sig_ref.url = "https://ontheair.nz/assets/" + ref_id.replace("/", "_")
-                        try:
-                            sig_ref.grid = latlong_to_locator(asset["y"], asset["x"], 6)
-                        except:
-                            logging.debug("Invalid lat/lon received for reference")
-                            sig_ref.latitude = asset["y"]
-                            sig_ref.longitude = asset["x"]
-                elif not response.from_cache:
-                    logging.warning("Malformed response looking up %s ref %s", sig, ref_id)
-            elif not response.from_cache:
-                logging.warning("HTTP %d looking up %s ref %s", response.status_code, sig, ref_id)
-
-        elif sig.upper() == "BOTA":
-            if not sig_ref.name:
-                sig_ref.name = sig_ref.id
-            sig_ref.url = "https://www.beachesontheair.com/beaches/" + sig_ref.name.lower().replace(" ", "-")
-
-        elif sig.upper() == "LLOTA":
-            response = SEMI_STATIC_URL_DATA_CACHE.get("https://llota.app/api/public/references",
-                                                      headers=HTTP_HEADERS)
-            if response.ok:
-                data = response.json()
-                if isinstance(data, list):
-                    global _LLOTA_INDEX_CACHE
-                    if not bool(_LLOTA_INDEX_CACHE) or not response.from_cache:
-                        # New data from LLOTA, update our internal map
-                        _LLOTA_INDEX_CACHE = {ref["reference_code"]: ref for ref in data}
-                    ref = _LLOTA_INDEX_CACHE.get(ref_id)
-                    if ref:
-                        sig_ref.name = str(ref["name"])
-                        sig_ref.url = "https://llota.app/list/ref/" + ref_id
-                        sig_ref.grid = str(ref["grid_locator"])
-                        ll = locator_to_latlong(sig_ref.grid)
-                        sig_ref.latitude = ll[0]
-                        sig_ref.longitude = ll[1]
-                elif not response.from_cache:
-                    logging.warning("Malformed response looking up %s ref %s", sig, ref_id)
-            elif not response.from_cache:
-                logging.warning("HTTP %d looking up %s ref %s", response.status_code, sig, ref_id)
-
-        elif sig.upper() == "WWTOTA":
-            if not sig_ref.name:
-                sig_ref.name = sig_ref.id
-            sig_ref.url = "https://wwtota.com/seznam/karta_rozhledny.php?ref=" + str(sig_ref.name)
+        # If the SIG is HEMA or KRMNPA, we have no current lookup for this so just skip it.
+        if sig.upper() == "HEMA" or sig.upper() == "KRMNPA":
+            return sig_ref
 
+        # If the SIG is Tiles, WAB, WAI or BOTA (Beaches), we don't have anything to look up from the data store, we can
+        # calculate all the information we are going to get directly. So handle those cases first
         elif sig.upper() == "TILES":
             # Tiles on the Air just uses Maidenhead 6-digit squares, so ID, Name and Grid are all the same
             if not sig_ref.name:
@@ -278,27 +74,26 @@ def populate_sig_ref_info(sig_ref):
                 except:
                     logging.warning("Invalid lat/lon received for WAB/WAI reference")
 
-        elif sig.upper() == "DME":
-            # Zero-pad to 5 digits to match our source data
-            row = _DME_INDEX.get(ref_id.zfill(5))
-            if row:
-                sig_ref.name = row["NOMBRE_ACTUAL"] + ", " + row["PROVINCIA"]
-                sig_ref.latitude = float(row["LATITUD_ETRS89_REGCAN95"].replace(",", ".")) if row.get(
-                    "LATITUD_ETRS89_REGCAN95") else None
-                sig_ref.longitude = float(row["LONGITUD_ETRS89_REGCAN95"].replace(",", ".")) if row.get(
-                    "LONGITUD_ETRS89_REGCAN95") else None
-                if sig_ref.latitude and sig_ref.longitude:
-                    try:
-                        sig_ref.grid = latlong_to_locator(sig_ref.latitude, sig_ref.longitude, 6)
-                    except Exception:
-                        logging.warning("Invalid lat/lon received for DME reference")
-            else:
-                logging.warning("DME database did not contain data for ref %s", ref_id)
+        elif sig.upper() == "BOTA":
+            # For BOTA all we can ever generate is the URL, there is no data file or lookup for lat/longs
+            if not sig_ref.name:
+                sig_ref.name = sig_ref.id
+            sig_ref.url = "https://www.beachesontheair.com/beaches/" + sig_ref.name.lower().replace(" ", "-")
+
+        # OK, this is something we have to look up. Now check to see if our data store contains SIG ref information for
+        # this SIG. If so, check for the reference data and use that.
+        elif sig in DATA_STORE.sigrefs:
+            key = sig + ":" + ref_id
+            lookup_data = DATA_STORE.sigrefs.get(key) if key in DATA_STORE.sigrefs else None
+            if lookup_data:
+                # Copy new sig ref data into existing object
+                sig_ref.__dict__.update(lookup_data.__dict__)
+            else:
+                logging.warning("%s database did not contain data for ref %s", sig, ref_id)
+
+        else:
+            logging.warning(f"Tried to look up a SIG called %s but Spothole does not know what that is.", sig)
 
-    except ConnectionError:
-        logging.warning("Connection error when looking up sig_ref info for " + sig + " ref " + ref_id)
-    except (ConnectTimeout, ReadTimeout):
-        logging.warning(f"Timeout when looking up sig_ref info for " + sig + " ref " + ref_id)
     except Exception:
         logging.error("Exception when looking up sig_ref info for " + sig + " ref " + ref_id, exc_info=True)
     return sig_ref

core/status_reporter.py

@@ -7,31 +7,29 @@ import pytz
 
 from core.config import SERVER_OWNER_CALLSIGN
 from core.constants import SOFTWARE_VERSION
+from core.data_store import DATA_STORE
 from core.prometheus_metrics_handler import memory_use_gauge, spots_gauge, alerts_gauge
 
 
 class StatusReporter:
     """Provides a timed update of the application's status data."""
 
-    def __init__(self, status_data, run_interval, web_server, cleanup_timer, spots, spot_providers, alerts,
-                 alert_providers, solar_condition_providers):
+    def __init__(self, run_interval, web_server, spot_providers, alert_providers, solar_condition_providers,
+                 sig_ref_data_providers):
         """Constructor"""
 
-        self._status_data = status_data
         self._run_interval = run_interval
         self._web_server = web_server
-        self._cleanup_timer = cleanup_timer
-        self._spots = spots
         self._spot_providers = spot_providers
-        self._alerts = alerts
         self._alert_providers = alert_providers
         self._solar_condition_providers = solar_condition_providers
+        self._sig_ref_data_providers = sig_ref_data_providers
         self._thread = None
         self._stop_event = Event()
         self._startup_time = datetime.now(pytz.UTC)
 
-        self._status_data["software-version"] = SOFTWARE_VERSION
-        self._status_data["server-owner-callsign"] = SERVER_OWNER_CALLSIGN
+        DATA_STORE.status_data["software-version"] = SOFTWARE_VERSION
+        DATA_STORE.status_data["server-owner-callsign"] = SERVER_OWNER_CALLSIGN
 
     def start(self):
         """Start the reporter thread"""
@@ -55,44 +53,47 @@ class StatusReporter:
     def _report(self):
         """Write status information"""
 
-        self._status_data["uptime"] = (datetime.now(pytz.UTC) - self._startup_time).total_seconds()
-        self._status_data["mem_use_mb"] = round(psutil.Process(os.getpid()).memory_info().rss / (1024 * 1024), 3)
-        self._status_data["num_spots"] = len(self._spots)
-        self._status_data["num_alerts"] = len(self._alerts)
-        self._status_data["spot_providers"] = list(
-            map(lambda p: {"name": p.name, "enabled": p.enabled, "status": p.status,
+        DATA_STORE.status_data["uptime"] = (datetime.now(pytz.UTC) - self._startup_time).total_seconds()
+        DATA_STORE.status_data["mem_use_mb"] = round(psutil.Process(os.getpid()).memory_info().rss / (1024 * 1024), 3)
+        DATA_STORE.status_data["num_spots"] = len(DATA_STORE.spots.values())
+        DATA_STORE.status_data["num_alerts"] = len(DATA_STORE.alerts.values())
+        DATA_STORE.status_data["spot_providers"] = list(
+        map(lambda p: {"name": p.name, "enabled": p.enabled, "status": p.status,
+                       "last_updated": p.last_update_time.replace(
+                           tzinfo=pytz.UTC).timestamp() if p.last_update_time.year > 2000 else 0,
+                       "last_spot": p.last_spot_time.replace(
+                           tzinfo=pytz.UTC).timestamp() if p.last_spot_time.year > 2000 else 0},
+            self._spot_providers))
+        DATA_STORE.status_data["alert_providers"] = list(
+        map(lambda p: {"name": p.name, "enabled": p.enabled, "status": p.status,
+                       "last_updated": p.last_update_time.replace(
+                           tzinfo=pytz.UTC).timestamp() if p.last_update_time.year > 2000 else 0},
+            self._alert_providers))
+        DATA_STORE.status_data["solar_condition_providers"] = list(
+        map(lambda p: {"name": p.name, "enabled": p.enabled, "status": p.status,
+                       "last_updated": p.last_update_time.replace(
+                           tzinfo=pytz.UTC).timestamp() if p.last_update_time.year > 2000 else 0},
+            self._solar_condition_providers))
+        DATA_STORE.status_data["sig_ref_data_providers"] = list(
+            map(lambda p: {"sig_name": p.sig_name, "enabled": p.enabled, "status": p.status,
                            "last_updated": p.last_update_time.replace(
                                tzinfo=pytz.UTC).timestamp() if p.last_update_time.year > 2000 else 0,
-                           "last_spot": p.last_spot_time.replace(
-                               tzinfo=pytz.UTC).timestamp() if p.last_spot_time.year > 2000 else 0},
-                self._spot_providers))
-        self._status_data["alert_providers"] = list(
-            map(lambda p: {"name": p.name, "enabled": p.enabled, "status": p.status,
-                           "last_updated": p.last_update_time.replace(
-                               tzinfo=pytz.UTC).timestamp() if p.last_update_time.year > 2000 else 0},
-                self._alert_providers))
-        self._status_data["solar_condition_providers"] = list(
-            map(lambda p: {"name": p.name, "enabled": p.enabled, "status": p.status,
-                           "last_updated": p.last_update_time.replace(
-                               tzinfo=pytz.UTC).timestamp() if p.last_update_time.year > 2000 else 0},
-                self._solar_condition_providers))
-        self._status_data["cleanup"] = {"status": self._cleanup_timer.status,
-                                        "last_ran": self._cleanup_timer.last_cleanup_time.replace(
-                                            tzinfo=pytz.UTC).timestamp() if self._cleanup_timer.last_cleanup_time else 0}
-        self._status_data["webserver"] = {"status": self._web_server.web_server_metrics["status"],
-                                          "last_api_access": self._web_server.web_server_metrics[
-                                              "last_api_access_time"].replace(
-                                              tzinfo=pytz.UTC).timestamp() if self._web_server.web_server_metrics[
-                                              "last_api_access_time"] else 0,
-                                          "api_access_count": self._web_server.web_server_metrics["api_access_counter"],
-                                          "last_page_access": self._web_server.web_server_metrics[
-                                              "last_page_access_time"].replace(
-                                              tzinfo=pytz.UTC).timestamp() if self._web_server.web_server_metrics[
-                                              "last_page_access_time"] else 0,
-                                          "page_access_count": self._web_server.web_server_metrics[
-                                              "page_access_counter"]}
+                           "reference_count": p.reference_count},
+                self._sig_ref_data_providers))
+        DATA_STORE.status_data["webserver"] = {"status": self._web_server.web_server_metrics["status"],
+                                      "last_api_access": self._web_server.web_server_metrics[
+                                          "last_api_access_time"].replace(
+                                          tzinfo=pytz.UTC).timestamp() if self._web_server.web_server_metrics[
+                                          "last_api_access_time"] else 0,
+                                      "api_access_count": self._web_server.web_server_metrics["api_access_counter"],
+                                      "last_page_access": self._web_server.web_server_metrics[
+                                          "last_page_access_time"].replace(
+                                          tzinfo=pytz.UTC).timestamp() if self._web_server.web_server_metrics[
+                                          "last_page_access_time"] else 0,
+                                      "page_access_count": self._web_server.web_server_metrics[
+                                          "page_access_counter"]}
 
         # Update Prometheus metrics
         memory_use_gauge.set(psutil.Process(os.getpid()).memory_info().rss)
-        spots_gauge.set(len(self._spots))
-        alerts_gauge.set(len(self._alerts))
+        spots_gauge.set(len(DATA_STORE.spots.values()))
+        alerts_gauge.set(len(DATA_STORE.alerts.values()))

core/url_data_cache.py

@@ -0,0 +1,23 @@
+import threading
+from datetime import timedelta
+
+from requests_cache import CachedSession
+
+# Cache for "semi-static" data retrieved from a URL. This is a layet of caching in addition to the normal caching of
+# spots, alerts and other data in the DataStore class. Its purpose is to avoid hitting remote endpoints frequently when
+# e.g. restarting Spothole many times during testing.
+_session = CachedSession("cache/semi_static_urls", expire_after=timedelta(days=1),
+                         allowable_codes=(200, 400, 401, 403, 404))
+_lock = threading.Lock()
+
+
+class _ThreadSafeSession:
+    """Wraps CachedSession with a lock to prevent concurrent SQLite access across threads. This allows a single object
+    to be used freely across the application."""
+
+    def get(self, *args, **kwargs):
+        with _lock:
+            return _session.get(*args, **kwargs)
+
+# Global object
+URL_DATA_CACHE = _ThreadSafeSession()

core/utils.py

@@ -5,14 +5,4 @@ def safe_json_dumps(obj):
     """Safe version of json.dumps that also converts objects to dicts so they can be output, and ignores NaN floats
     which are invalid in JSON."""
 
-    return simplejson.dumps(obj, ensure_ascii=False, ignore_nan=True, default=lambda o: o.__dict__)
-
-
-def empty_queue(q):
-    """Empty a queue"""
-
-    while not q.empty():
-        try:
-            q.get_nowait()
-        except:
-            break
+    return simplejson.dumps(obj, ensure_ascii=False, ignore_nan=True, default=lambda o: o.__dict__)

data/alert.py

@@ -1,6 +1,7 @@
 import copy
 import hashlib
 import json
+import logging
 from dataclasses import dataclass
 from datetime import datetime, timedelta
 
@@ -64,69 +65,73 @@ class Alert:
     def infer_missing(self, credentials=None):
         """Infer missing parameters where possible"""
 
-        # If we somehow don't have a start time, set it to zero so it sorts off the bottom of any list but
-        # clients can still reliably parse it as a number.
-        if not self.start_time:
-            self.start_time = 0
+        try:
+            # If we somehow don't have a start time, set it to zero so it sorts off the bottom of any list but
+            # clients can still reliably parse it as a number.
+            if not self.start_time:
+                self.start_time = 0
 
-        # If we don't have a received time, this has just been received so set that to "now"
-        if not self.received_time:
-            self.received_time = datetime.now(pytz.UTC).timestamp()
+            # If we don't have a received time, this has just been received so set that to "now"
+            if not self.received_time:
+                self.received_time = datetime.now(pytz.UTC).timestamp()
 
-        # Fill in ISO versions of times, in case the client prefers that
-        if self.start_time and not self.start_time_iso:
-            self.start_time_iso = datetime.fromtimestamp(self.start_time, pytz.UTC).isoformat()
-        if self.end_time and not self.end_time_iso:
-            self.end_time_iso = datetime.fromtimestamp(self.end_time, pytz.UTC).isoformat()
-        if self.received_time and not self.received_time_iso:
-            self.received_time_iso = datetime.fromtimestamp(self.received_time, pytz.UTC).isoformat()
+            # Fill in ISO versions of times, in case the client prefers that
+            if self.start_time and not self.start_time_iso:
+                self.start_time_iso = datetime.fromtimestamp(self.start_time, pytz.UTC).isoformat()
+            if self.end_time and not self.end_time_iso:
+                self.end_time_iso = datetime.fromtimestamp(self.end_time, pytz.UTC).isoformat()
+            if self.received_time and not self.received_time_iso:
+                self.received_time_iso = datetime.fromtimestamp(self.received_time, pytz.UTC).isoformat()
 
-        # DX country, continent, zones etc. from callsign. CQ/ITU zone are better looked up with a location but we don't
-        # have a real location for alerts.
-        if self.dx_calls and self.dx_calls[0] and not self.dx_country:
-            self.dx_country = lookup_helper.infer_country_from_callsign(self.dx_calls[0], credentials)
-        if self.dx_calls and self.dx_calls[0] and not self.dx_continent:
-            self.dx_continent = lookup_helper.infer_continent_from_callsign(self.dx_calls[0], credentials)
-        if self.dx_calls and self.dx_calls[0] and not self.dx_cq_zone:
-            self.dx_cq_zone = lookup_helper.infer_cq_zone_from_callsign(self.dx_calls[0], credentials)
-        if self.dx_calls and self.dx_calls[0] and not self.dx_itu_zone:
-            self.dx_itu_zone = lookup_helper.infer_itu_zone_from_callsign(self.dx_calls[0], credentials)
-        if self.dx_calls and self.dx_calls[0] and not self.dx_dxcc_id:
-            self.dx_dxcc_id = lookup_helper.infer_dxcc_id_from_callsign(self.dx_calls[0], credentials)
-        if self.dx_dxcc_id and not self.dx_flag:
-            self.dx_flag = lookup_helper.get_flag_for_dxcc(self.dx_dxcc_id)
+            # DX country, continent, zones etc. from callsign. CQ/ITU zone are better looked up with a location but we don't
+            # have a real location for alerts.
+            if self.dx_calls and self.dx_calls[0] and not self.dx_country:
+                self.dx_country = lookup_helper.infer_country_from_callsign(self.dx_calls[0], credentials)
+            if self.dx_calls and self.dx_calls[0] and not self.dx_continent:
+                self.dx_continent = lookup_helper.infer_continent_from_callsign(self.dx_calls[0], credentials)
+            if self.dx_calls and self.dx_calls[0] and not self.dx_cq_zone:
+                self.dx_cq_zone = lookup_helper.infer_cq_zone_from_callsign(self.dx_calls[0], credentials)
+            if self.dx_calls and self.dx_calls[0] and not self.dx_itu_zone:
+                self.dx_itu_zone = lookup_helper.infer_itu_zone_from_callsign(self.dx_calls[0], credentials)
+            if self.dx_calls and self.dx_calls[0] and not self.dx_dxcc_id:
+                self.dx_dxcc_id = lookup_helper.infer_dxcc_id_from_callsign(self.dx_calls[0], credentials)
+            if self.dx_dxcc_id and not self.dx_flag:
+                self.dx_flag = lookup_helper.get_flag_for_dxcc(self.dx_dxcc_id)
 
-        # Fetch SIG data. In case a particular API doesn't provide a full set of name, lat, lon & grid for a reference
-        # in its initial call, we use this code to populate the rest of the data. This includes working out grid refs
-        # from WAB and WAI, which count as a SIG even though there's no real lookup, just maths
-        if self.sig_refs and len(self.sig_refs) > 0:
-            for sig_ref in self.sig_refs:
-                populate_sig_ref_info(sig_ref)
+            # Fetch SIG data. In case a particular API doesn't provide a full set of name, lat, lon & grid for a reference
+            # in its initial call, we use this code to populate the rest of the data. This includes working out grid refs
+            # from WAB and WAI, which count as a SIG even though there's no real lookup, just maths
+            if self.sig_refs and len(self.sig_refs) > 0:
+                for sig_ref in self.sig_refs:
+                    populate_sig_ref_info(sig_ref)
 
-        # If the spot itself doesn't have a SIG yet, but we have at least one SIG reference, take that reference's SIG
-        # and apply it to the whole spot.
-        if self.sig_refs and len(self.sig_refs) > 0 and self.sig_refs[0] and not self.sig:
-            self.sig = self.sig_refs[0].sig
+            # If the spot itself doesn't have a SIG yet, but we have at least one SIG reference, take that reference's SIG
+            # and apply it to the whole spot.
+            if self.sig_refs and len(self.sig_refs) > 0 and self.sig_refs[0] and not self.sig:
+                self.sig = self.sig_refs[0].sig
 
-        # Always create an ID based on a hash of every parameter *except* received_time. This is used as the index
-        # to a map, which as a byproduct avoids us having multiple duplicate copies of the object that are identical
-        # apart from that they were retrieved from the API at different times. Note that the simple Python hash()
-        # function includes a seed randomly generated at runtime; this is therefore not consistent between runs. But we
-        # use diskcache to store our data between runs, so we use SHA256 which does not include this random element.
-        # The ID is computed before the online lookups below so that it is stable regardless of whether credentials
-        # are provided, allowing the enriched API response to be matched to the stored alert by ID.
-        if not self.id:
-            self_copy = copy.deepcopy(self)
-            self_copy.received_time = 0
-            self_copy.received_time_iso = ""
-            self.id = hashlib.sha256(str(self_copy).encode("utf-8")).hexdigest()
+            # Always create an ID based on a hash of every parameter *except* received_time. This is used as the index
+            # to a map, which as a byproduct avoids us having multiple duplicate copies of the object that are identical
+            # apart from that they were retrieved from the API at different times. Note that the simple Python hash()
+            # function includes a seed randomly generated at runtime; this is therefore not consistent between runs. But we
+            # use diskcache to store our data between runs, so we use SHA256 which does not include this random element.
+            # The ID is computed before the online lookups below so that it is stable regardless of whether credentials
+            # are provided, allowing the enriched API response to be matched to the stored alert by ID.
+            if not self.id:
+                self_copy = copy.deepcopy(self)
+                self_copy.received_time = 0
+                self_copy.received_time_iso = ""
+                self.id = hashlib.sha256(str(self_copy).encode("utf-8")).hexdigest()
 
-        # DX operator details lookup, using QRZ.com/HamQTH. This should be the last resort compared to taking the data
-        # from the actual alerting service, e.g. we don't want to accidentally use a user's QRZ.com home lat/lon
-        # instead of the one from the park reference they're at.
-        if self.dx_calls and not self.dx_names:
-            self.dx_names = list(
-                map(lambda c: lookup_helper.infer_name_from_callsign_online_lookup(c, credentials), self.dx_calls))
+            # DX operator details lookup, using QRZ.com/HamQTH. This should be the last resort compared to taking the data
+            # from the actual alerting service, e.g. we don't want to accidentally use a user's QRZ.com home lat/lon
+            # instead of the one from the park reference they're at.
+            if self.dx_calls and not self.dx_names:
+                self.dx_names = list(
+                    map(lambda c: lookup_helper.infer_name_from_callsign_online_lookup(c, credentials), self.dx_calls))
+
+        except Exception as e:
+            logging.error("Exception while inferring missing data from spot", e, exc_info=True)
 
     def to_json(self):
         """JSON serialise"""

data/spot.py

@@ -147,295 +147,299 @@ class Spot:
     def infer_missing(self, credentials=None):
         """Infer missing parameters where possible"""
 
-        # If we somehow don't have a spot time, set it to zero so it sorts off the bottom of any list but
-        # clients can still reliably parse it as a number.
-        if not self.time:
-            self.time = 0
+        try:
+            # If we somehow don't have a spot time, set it to zero so it sorts off the bottom of any list but
+            # clients can still reliably parse it as a number.
+            if not self.time:
+                self.time = 0
 
-        # If we don't have a received time, this has just been received so set that to "now"
-        if not self.received_time:
-            self.received_time = datetime.now(pytz.UTC).timestamp()
+            # If we don't have a received time, this has just been received so set that to "now"
+            if not self.received_time:
+                self.received_time = datetime.now(pytz.UTC).timestamp()
 
-        # Fill in ISO versions of times, in case the client prefers that
-        if self.time and not self.time_iso:
-            self.time_iso = datetime.fromtimestamp(self.time, pytz.UTC).isoformat()
-        if self.received_time and not self.received_time_iso:
-            self.received_time_iso = datetime.fromtimestamp(self.received_time, pytz.UTC).isoformat()
+            # Fill in ISO versions of times, in case the client prefers that
+            if self.time and not self.time_iso:
+                self.time_iso = datetime.fromtimestamp(self.time, pytz.UTC).isoformat()
+            if self.received_time and not self.received_time_iso:
+                self.received_time_iso = datetime.fromtimestamp(self.received_time, pytz.UTC).isoformat()
 
-        # Clean up DX call if it has an SSID or -# from RBN
-        if self.dx_call and "-" in self.dx_call:
-            split = self.dx_call.split("-")
-            self.dx_call = split[0]
-            if len(split) > 1 and split[1] != "#":
-                self.dx_ssid = split[1]
+            # Clean up DX call if it has an SSID or -# from RBN
+            if self.dx_call and "-" in self.dx_call:
+                split = self.dx_call.split("-")
+                self.dx_call = split[0]
+                if len(split) > 1 and split[1] != "#":
+                    self.dx_ssid = split[1]
 
-        # DX country, continent etc. from callsign
-        if self.dx_call and not self.dx_country:
-            self.dx_country = lookup_helper.infer_country_from_callsign(self.dx_call, credentials)
-        if self.dx_call and not self.dx_continent:
-            self.dx_continent = lookup_helper.infer_continent_from_callsign(self.dx_call, credentials)
-        if self.dx_call and not self.dx_dxcc_id:
-            self.dx_dxcc_id = lookup_helper.infer_dxcc_id_from_callsign(self.dx_call, credentials)
-        if self.dx_dxcc_id and not self.dx_flag:
-            self.dx_flag = lookup_helper.get_flag_for_dxcc(self.dx_dxcc_id)
+            # DX country, continent etc. from callsign
+            if self.dx_call and not self.dx_country:
+                self.dx_country = lookup_helper.infer_country_from_callsign(self.dx_call, credentials)
+            if self.dx_call and not self.dx_continent:
+                self.dx_continent = lookup_helper.infer_continent_from_callsign(self.dx_call, credentials)
+            if self.dx_call and not self.dx_dxcc_id:
+                self.dx_dxcc_id = lookup_helper.infer_dxcc_id_from_callsign(self.dx_call, credentials)
+            if self.dx_dxcc_id and not self.dx_flag:
+                self.dx_flag = lookup_helper.get_flag_for_dxcc(self.dx_dxcc_id)
 
-        # Clean up spotter call if it has an SSID or -# from RBN
-        if self.de_call and "-" in self.de_call:
-            split = self.de_call.split("-")
-            self.de_call = split[0]
-            if len(split) > 1 and split[1] != "#":
-                self.de_ssid = split[1]
+            # Clean up spotter call if it has an SSID or -# from RBN
+            if self.de_call and "-" in self.de_call:
+                split = self.de_call.split("-")
+                self.de_call = split[0]
+                if len(split) > 1 and split[1] != "#":
+                    self.de_ssid = split[1]
 
-        # If we have a spotter of "RBNHOLE", we should have the actual spotter callsign in the comment, so extract it.
-        # RBNHole posts come from a number of providers, so it's dealt with here in the generic spot handling code.
-        if self.de_call == "RBNHOLE" and self.comment:
-            rbnhole_call_match = re.search(r"\Wat ([a-z0-9/]+)\W", self.comment, re.IGNORECASE)
-            if rbnhole_call_match:
-                self.de_call = rbnhole_call_match.group(1).upper()
+            # If we have a spotter of "RBNHOLE", we should have the actual spotter callsign in the comment, so extract it.
+            # RBNHole posts come from a number of providers, so it's dealt with here in the generic spot handling code.
+            if self.de_call == "RBNHOLE" and self.comment:
+                rbnhole_call_match = re.search(r"\Wat ([a-z0-9/]+)\W", self.comment, re.IGNORECASE)
+                if rbnhole_call_match:
+                    self.de_call = rbnhole_call_match.group(1).upper()
 
-        # If we have a spotter of "SOTAMAT", we might have the actual spotter callsign in the comment, if so extract it.
-        # SOTAMAT can do POTA as well as SOTA, so it's dealt with here in the generic spot handling code.
-        if self.de_call == "SOTAMAT" and self.comment:
-            sotamat_call_match = re.search(r"\Wfrom ([a-z0-9/]+)]", self.comment, re.IGNORECASE)
-            if sotamat_call_match:
-                self.de_call = sotamat_call_match.group(1).upper()
+            # If we have a spotter of "SOTAMAT", we might have the actual spotter callsign in the comment, if so extract it.
+            # SOTAMAT can do POTA as well as SOTA, so it's dealt with here in the generic spot handling code.
+            if self.de_call == "SOTAMAT" and self.comment:
+                sotamat_call_match = re.search(r"\Wfrom ([a-z0-9/]+)]", self.comment, re.IGNORECASE)
+                if sotamat_call_match:
+                    self.de_call = sotamat_call_match.group(1).upper()
 
-        # Spotter country, continent, zones etc. from callsign.
-        # DE call with no digits, or APRS servers starting "T2" are not things we can look up location for
-        if self.de_call and any(char.isdigit() for char in self.de_call) and not (
-                self.de_call.startswith("T2") and self.source == "APRS-IS"):
-            if not self.de_country:
-                self.de_country = lookup_helper.infer_country_from_callsign(self.de_call, credentials)
-            if not self.de_continent:
-                self.de_continent = lookup_helper.infer_continent_from_callsign(self.de_call, credentials)
-            if not self.de_dxcc_id:
-                self.de_dxcc_id = lookup_helper.infer_dxcc_id_from_callsign(self.de_call, credentials)
-            if self.de_dxcc_id and not self.de_flag:
-                self.de_flag = lookup_helper.get_flag_for_dxcc(self.de_dxcc_id)
+            # Spotter country, continent, zones etc. from callsign.
+            # DE call with no digits, or APRS servers starting "T2" are not things we can look up location for
+            if self.de_call and any(char.isdigit() for char in self.de_call) and not (
+                    self.de_call.startswith("T2") and self.source == "APRS-IS"):
+                if not self.de_country:
+                    self.de_country = lookup_helper.infer_country_from_callsign(self.de_call, credentials)
+                if not self.de_continent:
+                    self.de_continent = lookup_helper.infer_continent_from_callsign(self.de_call, credentials)
+                if not self.de_dxcc_id:
+                    self.de_dxcc_id = lookup_helper.infer_dxcc_id_from_callsign(self.de_call, credentials)
+                if self.de_dxcc_id and not self.de_flag:
+                    self.de_flag = lookup_helper.get_flag_for_dxcc(self.de_dxcc_id)
 
-        # Remove NaNs in frequency
-        if self.freq and self.freq == float("nan"):
-            self.freq = None
+            # Remove NaNs in frequency
+            if self.freq and self.freq == float("nan"):
+                self.freq = None
 
-        # Band from frequency
-        if self.freq and not self.band:
-            band = infer_band_from_freq(self.freq)
-            self.band = band.name
+            # Band from frequency
+            if self.freq and not self.band:
+                band = infer_band_from_freq(self.freq)
+                self.band = band.name
 
-        # Mode from comments or bandplan
-        if self.mode:
-            self.mode_source = "SPOT"
-        if self.comment and not self.mode:
-            self.mode = infer_mode_from_comment(self.comment)
-            self.mode_source = "COMMENT"
-        if self.freq and not self.mode:
-            self.mode = infer_mode_from_frequency(self.freq)
-            self.mode_source = "BANDPLAN"
+            # Mode from comments or bandplan
+            if self.mode:
+                self.mode_source = "SPOT"
+            if self.comment and not self.mode:
+                self.mode = infer_mode_from_comment(self.comment)
+                self.mode_source = "COMMENT"
+            if self.freq and not self.mode:
+                self.mode = infer_mode_from_frequency(self.freq)
+                self.mode_source = "BANDPLAN"
 
-        # Normalise mode if necessary.
-        if self.mode in MODE_ALIASES:
-            self.mode = MODE_ALIASES[self.mode]
+            # Normalise mode if necessary.
+            if self.mode in MODE_ALIASES:
+                self.mode = MODE_ALIASES[self.mode]
 
-        # Mode type from mode
-        if self.mode and not self.mode_type:
-            self.mode_type = infer_mode_type_from_mode(self.mode)
+            # Mode type from mode
+            if self.mode and not self.mode_type:
+                self.mode_type = infer_mode_type_from_mode(self.mode)
 
-        # If we have a latitude or grid at this point, it can only have been provided by the spot itself
-        if self.dx_latitude or self.dx_grid:
-            self.dx_location_source = "SPOT"
+            # If we have a latitude or grid at this point, it can only have been provided by the spot itself
+            if self.dx_latitude or self.dx_grid:
+                self.dx_location_source = "SPOT"
 
-        # Set the top-level "SIG" if it is missing but we have at least one SIG ref.
-        if not self.sig and self.sig_refs and len(self.sig_refs) > 0:
-            self.sig = self.sig_refs[0].sig.upper()
+            # Set the top-level "SIG" if it is missing but we have at least one SIG ref.
+            if not self.sig and self.sig_refs and len(self.sig_refs) > 0:
+                self.sig = self.sig_refs[0].sig.upper()
 
-        # See if we already have a SIG reference, but the comment looks like it contains more for the same SIG. This
-        # should catch e.g. POTA comments like "2-fer: GB-0001 GB-0002".
-        if self.comment and self.sig_refs and len(self.sig_refs) > 0 and self.sig_refs[0].sig:
-            sig = self.sig_refs[0].sig.upper()
-            regex = get_ref_regex_for_sig(sig)
-            if regex:
-                all_comment_ref_matches = re.finditer(r"(^|\W)(" + regex + r")(^|\W)", self.comment, re.IGNORECASE)
-                for ref_match in all_comment_ref_matches:
-                    self._append_sig_ref_if_missing(SIGRef(id=ref_match.group(2).upper(), sig=sig))
+            # See if we already have a SIG reference, but the comment looks like it contains more for the same SIG. This
+            # should catch e.g. POTA comments like "2-fer: GB-0001 GB-0002".
+            if self.comment and self.sig_refs and len(self.sig_refs) > 0 and self.sig_refs[0].sig:
+                sig = self.sig_refs[0].sig.upper()
+                regex = get_ref_regex_for_sig(sig)
+                if regex:
+                    all_comment_ref_matches = re.finditer(r"(^|\W)(" + regex + r")(^|\W)", self.comment, re.IGNORECASE)
+                    for ref_match in all_comment_ref_matches:
+                        self._append_sig_ref_if_missing(SIGRef(id=ref_match.group(2).upper(), sig=sig))
 
-        # See if the comment looks like it contains any SIGs (and optionally SIG references) that we can
-        # add to the spot. This should catch cluster spot comments like "POTA GB-0001 WWFF GFF-0001" and e.g. POTA
-        # comments like "also WWFF GFF-0001".
-        if self.comment:
-            sig_matches = re.finditer(r"(^|\W)" + ANY_SIG_REGEX + r"($|\W)", self.comment, re.IGNORECASE)
-            for sig_match in sig_matches:
-                # First of all, if we haven't got a SIG for this spot set yet, now we have. This covers things like cluster
-                # spots where the comment is just "POTA".
-                found_sig = get_sig_name_from_comment_name(sig_match.group(2))
-                if not self.sig:
-                    self.sig = found_sig
+            # See if the comment looks like it contains any SIGs (and optionally SIG references) that we can
+            # add to the spot. This should catch cluster spot comments like "POTA GB-0001 WWFF GFF-0001" and e.g. POTA
+            # comments like "also WWFF GFF-0001".
+            if self.comment:
+                sig_matches = re.finditer(r"(^|\W)" + ANY_SIG_REGEX + r"($|\W)", self.comment, re.IGNORECASE)
+                for sig_match in sig_matches:
+                    # First of all, if we haven't got a SIG for this spot set yet, now we have. This covers things like cluster
+                    # spots where the comment is just "POTA".
+                    found_sig = get_sig_name_from_comment_name(sig_match.group(2))
+                    if not self.sig:
+                        self.sig = found_sig
 
-                # Now look to see if that SIG name was followed by something that looks like a reference ID for that SIG.
-                # If so, add that to the sig_refs list for this spot.
-                ref_regex = get_ref_regex_for_sig(found_sig)
-                if ref_regex:
-                    ref_matches = re.finditer(r"(^|\W)" + found_sig + r"([ -])(" + ref_regex + r")($|\W)", self.comment,
-                                              re.IGNORECASE)
-                    for ref_match in ref_matches:
-                        self._append_sig_ref_if_missing(SIGRef(id=ref_match.group(3).upper(), sig=found_sig))
+                    # Now look to see if that SIG name was followed by something that looks like a reference ID for that SIG.
+                    # If so, add that to the sig_refs list for this spot.
+                    ref_regex = get_ref_regex_for_sig(found_sig)
+                    if ref_regex:
+                        ref_matches = re.finditer(r"(^|\W)" + found_sig + r"([ -])(" + ref_regex + r")($|\W)", self.comment,
+                                                  re.IGNORECASE)
+                        for ref_match in ref_matches:
+                            self._append_sig_ref_if_missing(SIGRef(id=ref_match.group(3).upper(), sig=found_sig))
 
-        # Fetch SIG data. In case a particular API doesn't provide a full set of name, lat, lon & grid for a reference
-        # in its initial call, we use this code to populate the rest of the data. This includes working out grid refs
-        # from WAB and WAI, which count as a SIG even though there's no real lookup, just maths
-        if self.sig_refs and len(self.sig_refs) > 0:
-            for sig_ref in self.sig_refs:
-                sig_ref = populate_sig_ref_info(sig_ref)
-                # If the spot itself doesn't have location yet, but the SIG ref does, extract it
-                if sig_ref.grid and not self.dx_grid:
-                    self.dx_grid = sig_ref.grid
-                if sig_ref.latitude and not self.dx_latitude:
-                    self.dx_latitude = sig_ref.latitude
-                    self.dx_longitude = sig_ref.longitude
-                    if self.sig == "WAB" or self.sig == "WAI":
-                        self.dx_location_source = "WAB/WAI GRID"
-                    else:
-                        self.dx_location_source = "SIG REF LOOKUP"
+            # Fetch SIG data. In case a particular API doesn't provide a full set of name, lat, lon & grid for a reference
+            # in its initial call, we use this code to populate the rest of the data. This includes working out grid refs
+            # from WAB and WAI, which count as a SIG even though there's no real lookup, just maths
+            if self.sig_refs and len(self.sig_refs) > 0:
+                for sig_ref in self.sig_refs:
+                    sig_ref = populate_sig_ref_info(sig_ref)
+                    # If the spot itself doesn't have location yet, but the SIG ref does, extract it
+                    if sig_ref.grid and not self.dx_grid:
+                        self.dx_grid = sig_ref.grid
+                    if sig_ref.latitude and not self.dx_latitude:
+                        self.dx_latitude = sig_ref.latitude
+                        self.dx_longitude = sig_ref.longitude
+                        if self.sig == "WAB" or self.sig == "WAI":
+                            self.dx_location_source = "WAB/WAI GRID"
+                        else:
+                            self.dx_location_source = "SIG REF LOOKUP"
 
-        # If the spot itself doesn't have a SIG yet, but we have at least one SIG reference, take that reference's SIG
-        # and apply it to the whole spot.
-        if self.sig_refs and len(self.sig_refs) > 0 and not self.sig:
-            self.sig = self.sig_refs[0].sig
+            # If the spot itself doesn't have a SIG yet, but we have at least one SIG reference, take that reference's SIG
+            # and apply it to the whole spot.
+            if self.sig_refs and len(self.sig_refs) > 0 and not self.sig:
+                self.sig = self.sig_refs[0].sig
 
-        # Parse "de_grid<prop_mode>dx_grid" structures from the comment, e.g. "JN61ES(ES)JM56XT" or "JO02GQ<>KN17LG".
-        # These are common on cluster spots and can provide grid references in preference to e.g. QRZ lookup, as well as
-        # being the only source we have for propagation mode. Brace for nightmare regex from hell.
-        if self.comment:
-            grid_mode_grid_match = re.search(
-                r'\b([A-Ra-r]{2}\d{2}(?:[A-Xa-x]{2}(?:\d{2})?)?)(?:<([^>]*)>|\(([^)]*)\))([A-Ra-r]{2}\d{2}(?:[A-Xa-x]{2}(?:\d{2})?)?)\b',
-                self.comment)
-            if grid_mode_grid_match:
-                # regex matches, so extract grids:
-                if not self.de_grid:
-                    self.de_grid = grid_mode_grid_match.group(1).upper()
-                if not self.dx_grid:
-                    self.dx_grid = grid_mode_grid_match.group(4).upper()
-                    self.dx_location_source = "SPOT"
+            # Parse "de_grid<prop_mode>dx_grid" structures from the comment, e.g. "JN61ES(ES)JM56XT" or "JO02GQ<>KN17LG".
+            # These are common on cluster spots and can provide grid references in preference to e.g. QRZ lookup, as well as
+            # being the only source we have for propagation mode. Brace for nightmare regex from hell.
+            if self.comment:
+                grid_mode_grid_match = re.search(
+                    r'\b([A-Ra-r]{2}\d{2}(?:[A-Xa-x]{2}(?:\d{2})?)?)(?:<([^>]*)>|\(([^)]*)\))([A-Ra-r]{2}\d{2}(?:[A-Xa-x]{2}(?:\d{2})?)?)\b',
+                    self.comment)
+                if grid_mode_grid_match:
+                    # regex matches, so extract grids:
+                    if not self.de_grid:
+                        self.de_grid = grid_mode_grid_match.group(1).upper()
+                    if not self.dx_grid:
+                        self.dx_grid = grid_mode_grid_match.group(4).upper()
+                        self.dx_location_source = "SPOT"
 
-                # And extract propagation mode (group 2 for <...>, group 3 for (...)):
-                mode_tag = (grid_mode_grid_match.group(2) or grid_mode_grid_match.group(3) or "").upper()
-                if mode_tag and not self.propagation_mode:
-                    if mode_tag in PROPAGATION_MODES:
-                        self.propagation_mode = PROPAGATION_MODES[mode_tag]
-                    else:
-                        self.propagation_mode = mode_tag
-                        logging.info("Seen a new propagation mode tag not yet in the system: %s", mode_tag)
+                    # And extract propagation mode (group 2 for <...>, group 3 for (...)):
+                    mode_tag = (grid_mode_grid_match.group(2) or grid_mode_grid_match.group(3) or "").upper()
+                    if mode_tag and not self.propagation_mode:
+                        if mode_tag in PROPAGATION_MODES:
+                            self.propagation_mode = PROPAGATION_MODES[mode_tag]
+                        else:
+                            self.propagation_mode = mode_tag
+                            logging.info("Seen a new propagation mode tag not yet in the system: %s", mode_tag)
 
-        # DX Grid to lat/lon and vice versa in case one is missing
-        if self.dx_grid and not self.dx_latitude:
-            try:
-                ll = locator_to_latlong(self.dx_grid)
-                self.dx_latitude = ll[0]
-                self.dx_longitude = ll[1]
-            except:
-                logging.debug("Invalid grid received for spot")
-        if self.dx_latitude and self.dx_longitude and not self.dx_grid:
-            try:
-                self.dx_grid = latlong_to_locator(self.dx_latitude, self.dx_longitude, 8)
-            except:
-                logging.debug("Invalid lat/lon received for spot")
+            # DX Grid to lat/lon and vice versa in case one is missing
+            if self.dx_grid and not self.dx_latitude:
+                try:
+                    ll = locator_to_latlong(self.dx_grid)
+                    self.dx_latitude = ll[0]
+                    self.dx_longitude = ll[1]
+                except:
+                    logging.debug("Invalid grid received for spot")
+            if self.dx_latitude and self.dx_longitude and not self.dx_grid:
+                try:
+                    self.dx_grid = latlong_to_locator(self.dx_latitude, self.dx_longitude, 8)
+                except:
+                    logging.debug("Invalid lat/lon received for spot")
 
-        # QRT comment detection
-        if self.comment and not self.qrt:
-            self.qrt = "QRT" in self.comment.upper()
+            # QRT comment detection
+            if self.comment and not self.qrt:
+                self.qrt = "QRT" in self.comment.upper()
 
-        # Always create an ID based on a hash of every parameter *except* received_time. This is used as the index
-        # to a map, which as a byproduct avoids us having multiple duplicate copies of the object that are identical
-        # apart from that they were retrieved from the API at different times. Note that the simple Python hash()
-        # function includes a seed randomly generated at runtime; this is therefore not consistent between runs. But we
-        # use diskcache to store our data between runs, so we use SHA256 which does not include this random element.
-        # The ID is computed before the online lookups below so that it is stable regardless of whether credentials
-        # are provided, allowing the enriched API response to be matched to the stored spot by ID.
-        if not self.id:
-            self_copy = copy.deepcopy(self)
-            self_copy.received_time = 0
-            self_copy.received_time_iso = ""
-            self.id = hashlib.sha256(str(self_copy).encode("utf-8")).hexdigest()
+            # Always create an ID based on a hash of every parameter *except* received_time. This is used as the index
+            # to a map, which as a byproduct avoids us having multiple duplicate copies of the object that are identical
+            # apart from that they were retrieved from the API at different times. Note that the simple Python hash()
+            # function includes a seed randomly generated at runtime; this is therefore not consistent between runs. But we
+            # use diskcache to store our data between runs, so we use SHA256 which does not include this random element.
+            # The ID is computed before the online lookups below so that it is stable regardless of whether credentials
+            # are provided, allowing the enriched API response to be matched to the stored spot by ID.
+            if not self.id:
+                self_copy = copy.deepcopy(self)
+                self_copy.received_time = 0
+                self_copy.received_time_iso = ""
+                self.id = hashlib.sha256(str(self_copy).encode("utf-8")).hexdigest()
 
-        # DX operator details lookup, using QRZ.com/HamQTH. This should be the last resort compared to taking the data
-        # from the actual spotting service, e.g. we don't want to accidentally use a user's QRZ.com home lat/lon
-        # instead of the one from the park reference they're at.
-        if self.dx_call and not self.dx_name:
-            self.dx_name = lookup_helper.infer_name_from_callsign_online_lookup(self.dx_call, credentials)
-        if self.dx_call and not self.dx_latitude:
-            latlon = lookup_helper.infer_latlon_from_callsign_online_lookup(self.dx_call, credentials)
-            if latlon:
-                self.dx_latitude = latlon[0]
-                self.dx_longitude = latlon[1]
-            self.dx_grid = lookup_helper.infer_grid_from_callsign_online_lookup(self.dx_call, credentials)
-            self.dx_location_source = "HOME QTH"
-
-        # Determine a "QTH" string. If we have a SIG ref, pick the first one and turn it into a suitable string,
-        # otherwise see what they have set on an online lookup service.
-        if self.sig_refs and len(self.sig_refs) > 0:
-            qth = self.sig_refs[0].id
-            if self.sig_refs[0].name:
-                qth += " " + self.sig_refs[0].name
-            self.dx_qth = qth
-        else:
-            self.dx_qth = lookup_helper.infer_qth_from_callsign_online_lookup(self.dx_call, credentials)
-
-        # Last resort for getting a DX position, use the DXCC entity.
-        if self.dx_call and not self.dx_latitude:
-            latlon = lookup_helper.infer_latlon_from_callsign_dxcc(self.dx_call)
-            if latlon:
-                self.dx_latitude = latlon[0]
-                self.dx_longitude = latlon[1]
-                self.dx_grid = lookup_helper.infer_grid_from_callsign_dxcc(self.dx_call)
-                self.dx_location_source = "DXCC"
-
-        # It looks like we can sometimes get a string into lat/lon, so try to parse as float, reject if not valid
-        if isinstance(self.dx_latitude, str) or isinstance(self.dx_longitude, str):
-            try:
-                self.dx_latitude = float(str(self.dx_latitude))
-                self.dx_longitude = float(str(self.dx_longitude))
-            except (TypeError, ValueError):
-                logging.warning("Received non-numeric strings in lat/lon (" + str(self.dx_latitude) + ", " + str(
-                    self.dx_longitude) + ") for call " + str(self.dx_call) + ", rejecting it")
-                self.dx_latitude = None
-                self.dx_longitude = None
-
-        # CQ and ITU zone lookup, preferably from location but failing that, from callsign
-        if not self.dx_cq_zone:
-            if self.dx_latitude:
-                self.dx_cq_zone = lat_lon_to_cq_zone(self.dx_latitude, self.dx_longitude)
-            elif self.dx_call:
-                self.dx_cq_zone = lookup_helper.infer_cq_zone_from_callsign(self.dx_call, credentials)
-        if not self.dx_itu_zone:
-            if self.dx_latitude:
-                self.dx_itu_zone = lat_lon_to_itu_zone(self.dx_latitude, self.dx_longitude)
-            elif self.dx_call:
-                self.dx_itu_zone = lookup_helper.infer_itu_zone_from_callsign(self.dx_call, credentials)
-
-        # DX Location is "good" if it is from a spot, or from QRZ if the callsign doesn't contain a slash, so the operator
-        # is likely at home.
-        self.dx_location_good = bool(self.dx_latitude and self.dx_longitude and (
-                self.dx_location_source == "SPOT" or self.dx_location_source == "SIG REF LOOKUP"
-                or self.dx_location_source == "WAB/WAI GRID"
-                or (self.dx_location_source == "HOME QTH" and "/" not in (self.dx_call or ""))))
-
-        # DE with no digits and APRS servers starting "T2" are not things we can look up location for
-        if self.de_call and any(char.isdigit() for char in self.de_call) and not (
-                self.de_call.startswith("T2") and self.source == "APRS-IS"):
-            # DE operator position lookup, using QRZ.com/HamQTH.
-            if not self.de_latitude:
-                latlon = lookup_helper.infer_latlon_from_callsign_online_lookup(self.de_call, credentials)
+            # DX operator details lookup, using QRZ.com/HamQTH. This should be the last resort compared to taking the data
+            # from the actual spotting service, e.g. we don't want to accidentally use a user's QRZ.com home lat/lon
+            # instead of the one from the park reference they're at.
+            if self.dx_call and not self.dx_name:
+                self.dx_name = lookup_helper.infer_name_from_callsign_online_lookup(self.dx_call, credentials)
+            if self.dx_call and not self.dx_latitude:
+                latlon = lookup_helper.infer_latlon_from_callsign_online_lookup(self.dx_call, credentials)
                 if latlon:
-                    self.de_latitude = latlon[0]
-                    self.de_longitude = latlon[1]
-                self.de_grid = lookup_helper.infer_grid_from_callsign_online_lookup(self.de_call, credentials)
+                    self.dx_latitude = latlon[0]
+                    self.dx_longitude = latlon[1]
+                self.dx_grid = lookup_helper.infer_grid_from_callsign_online_lookup(self.dx_call, credentials)
+                self.dx_location_source = "HOME QTH"
 
-            # Last resort for getting a DE position, use the DXCC entity.
-            if not self.de_latitude:
-                latlon = lookup_helper.infer_latlon_from_callsign_dxcc(self.de_call)
+            # Determine a "QTH" string. If we have a SIG ref, pick the first one and turn it into a suitable string,
+            # otherwise see what they have set on an online lookup service.
+            if self.sig_refs and len(self.sig_refs) > 0:
+                qth = self.sig_refs[0].id
+                if self.sig_refs[0].name:
+                    qth += " " + self.sig_refs[0].name
+                self.dx_qth = qth
+            else:
+                self.dx_qth = lookup_helper.infer_qth_from_callsign_online_lookup(self.dx_call, credentials)
+
+            # Last resort for getting a DX position, use the DXCC entity.
+            if self.dx_call and not self.dx_latitude:
+                latlon = lookup_helper.infer_latlon_from_callsign_dxcc(self.dx_call)
                 if latlon:
-                    self.de_latitude = latlon[0]
-                    self.de_longitude = latlon[1]
-                    self.de_grid = lookup_helper.infer_grid_from_callsign_dxcc(self.de_call)
+                    self.dx_latitude = latlon[0]
+                    self.dx_longitude = latlon[1]
+                    self.dx_grid = lookup_helper.infer_grid_from_callsign_dxcc(self.dx_call)
+                    self.dx_location_source = "DXCC"
+
+            # It looks like we can sometimes get a string into lat/lon, so try to parse as float, reject if not valid
+            if isinstance(self.dx_latitude, str) or isinstance(self.dx_longitude, str):
+                try:
+                    self.dx_latitude = float(str(self.dx_latitude))
+                    self.dx_longitude = float(str(self.dx_longitude))
+                except (TypeError, ValueError):
+                    logging.warning("Received non-numeric strings in lat/lon (" + str(self.dx_latitude) + ", " + str(
+                        self.dx_longitude) + ") for call " + str(self.dx_call) + ", rejecting it")
+                    self.dx_latitude = None
+                    self.dx_longitude = None
+
+            # CQ and ITU zone lookup, preferably from location but failing that, from callsign
+            if not self.dx_cq_zone:
+                if self.dx_latitude:
+                    self.dx_cq_zone = lat_lon_to_cq_zone(self.dx_latitude, self.dx_longitude)
+                elif self.dx_call:
+                    self.dx_cq_zone = lookup_helper.infer_cq_zone_from_callsign(self.dx_call, credentials)
+            if not self.dx_itu_zone:
+                if self.dx_latitude:
+                    self.dx_itu_zone = lat_lon_to_itu_zone(self.dx_latitude, self.dx_longitude)
+                elif self.dx_call:
+                    self.dx_itu_zone = lookup_helper.infer_itu_zone_from_callsign(self.dx_call, credentials)
+
+            # DX Location is "good" if it is from a spot, or from QRZ if the callsign doesn't contain a slash, so the operator
+            # is likely at home.
+            self.dx_location_good = bool(self.dx_latitude and self.dx_longitude and (
+                    self.dx_location_source == "SPOT" or self.dx_location_source == "SIG REF LOOKUP"
+                    or self.dx_location_source == "WAB/WAI GRID"
+                    or (self.dx_location_source == "HOME QTH" and "/" not in (self.dx_call or ""))))
+
+            # DE with no digits and APRS servers starting "T2" are not things we can look up location for
+            if self.de_call and any(char.isdigit() for char in self.de_call) and not (
+                    self.de_call.startswith("T2") and self.source == "APRS-IS"):
+                # DE operator position lookup, using QRZ.com/HamQTH.
+                if not self.de_latitude:
+                    latlon = lookup_helper.infer_latlon_from_callsign_online_lookup(self.de_call, credentials)
+                    if latlon:
+                        self.de_latitude = latlon[0]
+                        self.de_longitude = latlon[1]
+                    self.de_grid = lookup_helper.infer_grid_from_callsign_online_lookup(self.de_call, credentials)
+
+                # Last resort for getting a DE position, use the DXCC entity.
+                if not self.de_latitude:
+                    latlon = lookup_helper.infer_latlon_from_callsign_dxcc(self.de_call)
+                    if latlon:
+                        self.de_latitude = latlon[0]
+                        self.de_longitude = latlon[1]
+                        self.de_grid = lookup_helper.infer_grid_from_callsign_dxcc(self.de_call)
+
+        except Exception as e:
+            logging.error("Exception while inferring missing data from spot", e, exc_info=True)
 
     def to_json(self):
         """JSON serialise"""

datafiles/39c3-tota.csv

@@ -1,18 +0,0 @@
-ref,lat,lon
-T-01,53.56278090617755,9.984341869295505
-T-02,53.562383404176416,9.98551893027115
-T-03,53.56170184391514,9.985416035619778
-T-04,53.562026534393176,9.986372919078974
-T-11,53.56284641242506,9.98475590239655
-T-12,53.562431705517035,9.98551675702443
-T-13,53.56223704898424,9.985774520335664
-T-14,53.5617893512591,9.986344302837976
-T-21,53.56284641242506,9.98475590239655
-T-22,53.56245816412497,9.985456089490567
-T-23,53.56199560857136,9.985636761412673
-T-24,53.5617893512591,9.986344302837976
-T-31,53.56247470064887,9.985611427551902
-T-32,53.5617893512591,9.986344302837976
-T-41,53.56245039134992,9.985486136112701
-T-91,53.56147934973529,9.984626806439744
-T-92,53.561396810300735,9.987553052152899

datafiles/eh23-tota.csv

@@ -1,13 +0,0 @@
-ref,lat,lon
-T-01,50.3636495,7.5584857
-T-02,50.3636495,7.5584857
-T-03,50.3636495,7.5584857
-T-11,50.3636495,7.5584857
-T-13,50.3636495,7.5584857
-T-14,50.3636495,7.5584857
-T-21,50.3636495,7.5584857
-T-31,50.3636495,7.5584857
-T-33,50.3636495,7.5584857
-T-34,50.3636495,7.5584857
-T-41,50.3636495,7.5584857
-T-51,50.3636495,7.5584857

datafiles/toilets.csv

@@ -0,0 +1,30 @@
+ref,lat,lon
+C3 T-01,53.56278090617755,9.984341869295505
+C3 T-02,53.562383404176416,9.98551893027115
+C3 T-03,53.56170184391514,9.985416035619778
+C3 T-04,53.562026534393176,9.986372919078974
+C3 T-11,53.56284641242506,9.98475590239655
+C3 T-12,53.562431705517035,9.98551675702443
+C3 T-13,53.56223704898424,9.985774520335664
+C3 T-14,53.5617893512591,9.986344302837976
+C3 T-21,53.56284641242506,9.98475590239655
+C3 T-22,53.56245816412497,9.985456089490567
+C3 T-23,53.56199560857136,9.985636761412673
+C3 T-24,53.5617893512591,9.986344302837976
+C3 T-31,53.56247470064887,9.985611427551902
+C3 T-32,53.5617893512591,9.986344302837976
+C3 T-41,53.56245039134992,9.985486136112701
+C3 T-91,53.56147934973529,9.984626806439744
+C3 T-92,53.561396810300735,9.987553052152899
+EH T-01,50.3636495,7.5584857
+EH T-02,50.3636495,7.5584857
+EH T-03,50.3636495,7.5584857
+EH T-11,50.3636495,7.5584857
+EH T-13,50.3636495,7.5584857
+EH T-14,50.3636495,7.5584857
+EH T-21,50.3636495,7.5584857
+EH T-31,50.3636495,7.5584857
+EH T-33,50.3636495,7.5584857
+EH T-34,50.3636495,7.5584857
+EH T-41,50.3636495,7.5584857
+EH T-51,50.3636495,7.5584857

requirements.txt

@@ -17,4 +17,5 @@ websocket-client~=1.8.0
 tornado~=6.4.2
 tornado_eventsource~=3.0.0
 geopandas~=0.13.2
-simplejson~=4.1.1
+simplejson~=4.1.1
+cachetools~=7.1.6

server/handlers/api/addspot.py

@@ -8,7 +8,7 @@ import tornado
 from tornado import httputil
 from tornado.web import Application
 
-from core.config import ALLOW_SPOTTING, MAX_SPOT_AGE
+from core.config import ALLOW_SPOTTING
 from core.constants import UNKNOWN_BAND
 from core.lookup_helper import infer_band_from_freq
 from core.prometheus_metrics_handler import api_requests_counter
@@ -119,7 +119,7 @@ class APISpotHandler(tornado.web.RequestHandler):
             # infer missing data, and add it to our database.
             spot.source = "API"
             spot.infer_missing()
-            self._spots.add(spot.id, spot, expire=MAX_SPOT_AGE)
+            self._spots.set(spot.id, spot)
 
             self.write(safe_json_dumps("OK"))
             self.set_status(201)

server/handlers/api/alerts.py

@@ -1,7 +1,6 @@
 import copy
 import logging
 from datetime import datetime
-from queue import Queue
 from typing import Any
 
 import pytz
@@ -11,12 +10,9 @@ from tornado import httputil
 from tornado.web import Application
 
 from core.prometheus_metrics_handler import api_requests_counter
-from core.utils import safe_json_dumps, empty_queue
+from core.utils import safe_json_dumps
 from data.lookup_credentials import extract_credentials
 
-SSE_HANDLER_MAX_QUEUE_SIZE = 100
-SSE_HANDLER_QUEUE_CHECK_INTERVAL = 5000
-
 
 class APIAlertsHandler(tornado.web.RequestHandler):
     """API request handler for /api/v1/alerts"""
@@ -73,16 +69,14 @@ class APIAlertsStreamHandler(tornado_eventsource.handler.EventSourceHandler):
     """API request handler for /api/v1/alerts/stream"""
 
     def __init__(self, application, request, **kwargs: Any):
-        self._sse_alert_queues = None
+        self._sse_alert_broadcaster = None
         self._web_server_metrics = None
         self._query_params = None
         self._credentials = None
-        self._alert_queue = None
-        self._heartbeat = None
         super().__init__(application, request, **kwargs)
 
-    def initialize(self, sse_alert_queues, web_server_metrics):
-        self._sse_alert_queues = sse_alert_queues
+    def initialize(self, _sse_alert_broadcaster, web_server_metrics):
+        self._sse_alert_broadcaster = _sse_alert_broadcaster
         self._web_server_metrics = web_server_metrics
 
     def custom_headers(self):
@@ -104,59 +98,32 @@ class APIAlertsStreamHandler(tornado_eventsource.handler.EventSourceHandler):
             self._query_params = {k: v[0].decode("utf-8") for k, v in self.request.arguments.items()}
             self._credentials = extract_credentials(self._query_params)
 
-            # Create a alert queue and add it to the web server's list. The web server will fill this when alerts arrive
-            self._alert_queue = Queue(maxsize=SSE_HANDLER_MAX_QUEUE_SIZE)
-            self._sse_alert_queues.append(self._alert_queue)
-
-            # Set up a timed callback to check if anything is in the queue
-            self._heartbeat = tornado.ioloop.PeriodicCallback(self._callback, SSE_HANDLER_QUEUE_CHECK_INTERVAL)
-            self._heartbeat.start()
-
             # Flush headers immediately so nginx doesn't time out waiting for a response
             self.write_message("keepalive", "")
 
+            # Register to handle new alerts arriving. The callback() method will get called with the new alert as an
+            # argument.
+            self._sse_alert_broadcaster.register(self)
+
         except Exception as e:
             logging.warning("Exception when serving SSE socket: %s", e, exc_info=True)
             self.close()
 
     def close(self):
-        """When the user closes the socket, empty our queue and remove it from the list so the server no longer fills it"""
+        """When the user closes the socket, deregister ourselves from the alert broadcaster"""
 
-        try:
-            if self._alert_queue in self._sse_alert_queues:
-                self._sse_alert_queues.remove(self._alert_queue)
-            empty_queue(self._alert_queue)
-        except:
-            pass
-        try:
-            self._heartbeat.stop()
-        except:
-            pass
-        self._alert_queue = None
+        self._sse_alert_broadcaster.unregister(self)
         super().close()
 
-    def _callback(self):
-        """Callback to check if anything has arrived in the queue, and if so send it to the client"""
+    def callback(self, alert):
+        """Callback when a new alert arrives"""
 
         try:
-            if self._alert_queue:
-                if not self._alert_queue.empty():
-                    while not self._alert_queue.empty():
-                        alert = self._alert_queue.get()
-                        # If the new alert matches our param filters, send it to the client. If not, ignore it.
-                        if alert_allowed_by_query(alert, self._query_params):
-                            if self._credentials:
-                                alert = copy.deepcopy(alert)
-                                alert.infer_missing(self._credentials)
-                            self.write_message(msg=safe_json_dumps(alert))
-
-                else:
-                    # Send a keepalive comment if the queue was empty
-                    self.write_message("keepalive", "")
-
-                if self._alert_queue not in self._sse_alert_queues:
-                    logging.error("Web server cleared up a queue of an active connection!")
-                    self.close()
+            if alert_allowed_by_query(alert, self._query_params):
+                if self._credentials:
+                    alert = copy.deepcopy(alert)
+                    alert.infer_missing(self._credentials)
+                self.write_message(msg=safe_json_dumps(alert))
         except Exception as e:
             logging.warning("Exception in SSE callback, connection will be closed: %s", e, exc_info=True)
             self.close()
@@ -169,7 +136,7 @@ def get_alert_list_with_filters(all_alerts, query):
     # Create a shallow copy of the alert list ordered by start time, then filter the list to reduce it only to alerts
     # that match the filter parameters in the query string. Finally, apply a limit to the number of alerts returned.
     # The list of query string filters is defined in the API docs.
-    alert_ids = list(all_alerts.iterkeys())
+    alert_ids = all_alerts.keys()
     alerts = []
     for k in alert_ids:
         a = all_alerts.get(k)

server/handlers/api/dxstats.py

@@ -40,7 +40,7 @@ class APIDxStatsHandler(tornado.web.RequestHandler):
             one_hour_ago = (datetime.now(pytz.UTC) - timedelta(hours=1)).timestamp()
             counts = Counter()
 
-            for key in self._spots.iterkeys():
+            for key in self._spots.keys():
                 spot = self._spots.get(key)
                 if spot is None:
                     continue

server/handlers/api/spots.py

@@ -1,7 +1,6 @@
 import copy
 import logging
 from datetime import datetime, timedelta
-from queue import Queue
 from typing import Any
 
 import pytz
@@ -11,12 +10,9 @@ from tornado import httputil
 from tornado.web import Application
 
 from core.prometheus_metrics_handler import api_requests_counter
-from core.utils import safe_json_dumps, empty_queue
+from core.utils import safe_json_dumps
 from data.lookup_credentials import extract_credentials
 
-SSE_HANDLER_MAX_QUEUE_SIZE = 1000
-SSE_HANDLER_QUEUE_CHECK_INTERVAL = 5000
-
 
 class APISpotsHandler(tornado.web.RequestHandler):
     """API request handler for /api/v1/spots"""
@@ -73,16 +69,14 @@ class APISpotsStreamHandler(tornado_eventsource.handler.EventSourceHandler):
     """API request handler for /api/v1/spots/stream"""
 
     def __init__(self, application, request, **kwargs: Any):
-        self._sse_spot_queues = None
+        self._sse_spot_broadcaster = None
         self._web_server_metrics = None
         self._query_params = None
         self._credentials = None
-        self._spot_queue = None
-        self._heartbeat = None
         super().__init__(application, request, **kwargs)
 
-    def initialize(self, sse_spot_queues, web_server_metrics):
-        self._sse_spot_queues = sse_spot_queues
+    def initialize(self, sse_spot_broadcaster, web_server_metrics):
+        self._sse_spot_broadcaster = sse_spot_broadcaster
         self._web_server_metrics = web_server_metrics
 
     def custom_headers(self):
@@ -106,59 +100,33 @@ class APISpotsStreamHandler(tornado_eventsource.handler.EventSourceHandler):
             self._query_params = {k: v[0].decode("utf-8") for k, v in self.request.arguments.items()}
             self._credentials = extract_credentials(self._query_params)
 
-            # Create a spot queue and add it to the web server's list. The web server will fill this when spots arrive
-            self._spot_queue = Queue(maxsize=SSE_HANDLER_MAX_QUEUE_SIZE)
-            self._sse_spot_queues.append(self._spot_queue)
-
-            # Set up a timed callback to check if anything is in the queue
-            self._heartbeat = tornado.ioloop.PeriodicCallback(self._callback, SSE_HANDLER_QUEUE_CHECK_INTERVAL)
-            self._heartbeat.start()
-
             # Flush headers immediately so nginx doesn't time out waiting for a response
             self.write_message("keepalive", "")
 
+            # Register to handle new spots arriving. The callback() method will get called with the new spot as an
+            # argument.
+            self._sse_spot_broadcaster.register(self)
+
         except Exception as e:
             logging.warning("Exception when serving SSE socket: %s", e, exc_info=True)
             self.close()
 
     def close(self):
-        """When the user closes the socket, empty our queue and remove it from the list so the server no longer fills it"""
+        """When the user closes the socket, deregister ourselves from the spot broadcaster"""
 
-        try:
-            if self._spot_queue in self._sse_spot_queues:
-                self._sse_spot_queues.remove(self._spot_queue)
-            empty_queue(self._spot_queue)
-        except:
-            pass
-        try:
-            self._heartbeat.stop()
-        except:
-            pass
-        self._spot_queue = None
+        self._sse_spot_broadcaster.unregister(self)
         super().close()
 
-    def _callback(self):
-        """Callback to check if anything has arrived in the queue, and if so send it to the client"""
+    def callback(self, spot):
+        """Callback when a new spot arrives"""
 
         try:
-            if self._spot_queue:
-                if not self._spot_queue.empty():
-                    while not self._spot_queue.empty():
-                        spot = self._spot_queue.get()
-                        # If the new spot matches our param filters, send it to the client. If not, ignore it.
-                        if spot_allowed_by_query(spot, self._query_params):
-                            if self._credentials:
-                                spot = copy.deepcopy(spot)
-                                spot.infer_missing(self._credentials)
-                            self.write_message(msg=safe_json_dumps(spot))
-
-                else:
-                    # Send a keepalive comment if the queue was empty
-                    self.write_message("keepalive", "")
-
-                if self._spot_queue not in self._sse_spot_queues:
-                    logging.error("Web server cleared up a queue of an active connection!")
-                    self.close()
+            # If the new spot matches our param filters, send it to the client. If not, ignore it.
+            if spot_allowed_by_query(spot, self._query_params):
+                if self._credentials:
+                    spot = copy.deepcopy(spot)
+                    spot.infer_missing(self._credentials)
+                self.write_message(msg=safe_json_dumps(spot))
         except Exception as e:
             logging.warning("Exception in SSE callback, connection will be closed: %s", e, exc_info=True)
             self.close()
@@ -171,7 +139,7 @@ def get_spot_list_with_filters(all_spots, query):
     # Create a shallow copy of the spot list, ordered by spot time, then filter the list to reduce it only to spots
     # that match the filter parameters in the query string. Finally, apply a limit to the number of spots returned.
     # The list of query string filters is defined in the API docs.
-    spot_ids = list(all_spots.iterkeys())
+    spot_ids = all_spots.keys()
     spots = []
     for k in spot_ids:
         s = all_spots.get(k)

server/sse_broadcaster.py

@@ -0,0 +1,36 @@
+import logging
+import threading
+
+from tornado.ioloop import IOLoop
+
+
+class SSEBroadcaster:
+    """Bridge between DataStore listener callbacks (which fire on provider threads) to Tornado's async SSE handlers
+    (which live on the IOLoop thread) to avoid any interdependency between them."""
+
+    def __init__(self):
+        self._handlers = set()
+        self._lock = threading.Lock()
+        self._loop = IOLoop.current()
+
+    def register(self, handler):
+        with self._lock:
+            self._handlers.add(handler)
+
+    def unregister(self, handler):
+        with self._lock:
+            self._handlers.discard(handler)
+
+    def publish(self, value):
+        self._loop.add_callback(self._fan_out, value)
+
+    def _fan_out(self, value):
+        with self._lock:
+            handlers = list(self._handlers)
+        for handler in handlers:
+            try:
+                handler.callback(value)
+            except Exception:
+                # Connection probably dropped, ignore and de-register the handler to stop getting future items.
+                logging.debug("Failed to push to an SSE client; dropping it")
+                self.unregister(handler)

server/webserver.py

@@ -6,7 +6,7 @@ import tornado
 from tornado.web import StaticFileHandler
 
 from core.config import ALLOW_SPOTTING, WEB_SERVER_PORT, API_ONLY_MODE, LOG_WEB_REQUESTS, BASE_URL
-from core.utils import empty_queue
+from core.data_store import DATA_STORE
 from server.handlers.api.addspot import APISpotHandler
 from server.handlers.api.alerts import APIAlertsHandler, APIAlertsStreamHandler
 from server.handlers.api.dxstats import APIDxStatsHandler
@@ -18,6 +18,7 @@ from server.handlers.api.status import APIStatusHandler
 from server.handlers.manifesthandler import ManifestHandler
 from server.handlers.metrics import PrometheusMetricsHandler
 from server.handlers.pagetemplate import PageTemplateHandler
+from server.sse_broadcaster import SSEBroadcaster
 
 _HERE = os.path.dirname(__file__ or "")
 
@@ -25,15 +26,12 @@ _HERE = os.path.dirname(__file__ or "")
 class WebServer:
     """Provides the public-facing web server."""
 
-    def __init__(self, spots, alerts, solar_conditions, status_data):
+    def __init__(self):
         """Constructor"""
 
-        self._spots = spots
-        self._alerts = alerts
-        self._solar_conditions = solar_conditions
-        self._sse_spot_queues = []
-        self._sse_alert_queues = []
-        self._status_data = status_data
+        self._data_store = DATA_STORE
+        self._spot_broadcaster = SSEBroadcaster()
+        self._alert_broadcaster = SSEBroadcaster()
         self._port = WEB_SERVER_PORT
         self._api_only_mode = API_ONLY_MODE
         self._shutdown_event = asyncio.Event()
@@ -45,6 +43,10 @@ class WebServer:
             "status": "Starting"
         }
 
+        # Listen for new spots and alerts being added to the cache, so we can notify SSE clients immediately
+        DATA_STORE.spots.add_listener(self._spot_broadcaster.publish)
+        DATA_STORE.alerts.add_listener(self._alert_broadcaster.publish)
+
     def start(self):
         """Start the web server"""
 
@@ -64,20 +66,21 @@ class WebServer:
 
         # API endpoints are always enabled
         api_routes = [
-            (r"/api/v1/spots", APISpotsHandler, {"spots": self._spots, **handler_opts}),
-            (r"/api/v1/alerts", APIAlertsHandler, {"alerts": self._alerts, **handler_opts}),
+            (r"/api/v1/spots", APISpotsHandler, {"spots": self._data_store.spots, **handler_opts}),
+            (r"/api/v1/alerts", APIAlertsHandler, {"alerts": self._data_store.alerts, **handler_opts}),
             (r"/api/v1/spots/stream", APISpotsStreamHandler,
-             {"sse_spot_queues": self._sse_spot_queues, **handler_opts}),
+             {"sse_spot_broadcaster": self._spot_broadcaster, **handler_opts}),
             (r"/api/v1/alerts/stream", APIAlertsStreamHandler,
-             {"sse_alert_queues": self._sse_alert_queues, **handler_opts}),
-            (r"/api/v1/solar", APISolarConditionsHandler, {"solar_conditions": self._solar_conditions, **handler_opts}),
-            (r"/api/v1/dxstats", APIDxStatsHandler, {"spots": self._spots, **handler_opts}),
-            (r"/api/v1/options", APIOptionsHandler, {"status_data": self._status_data, **handler_opts}),
-            (r"/api/v1/status", APIStatusHandler, {"status_data": self._status_data, **handler_opts}),
+             {"sse_alert_broadcaster": self._alert_broadcaster, **handler_opts}),
+            (r"/api/v1/solar", APISolarConditionsHandler, {"solar_conditions": self._data_store.solar_conditions,
+                                                           **handler_opts}),
+            (r"/api/v1/dxstats", APIDxStatsHandler, {"spots": self._data_store.spots, **handler_opts}),
+            (r"/api/v1/options", APIOptionsHandler, {"status_data": self._data_store.status_data, **handler_opts}),
+            (r"/api/v1/status", APIStatusHandler, {"status_data": self._data_store.status_data, **handler_opts}),
             (r"/api/v1/lookup/call", APILookupCallHandler, {**handler_opts}),
             (r"/api/v1/lookup/sigref", APILookupSIGRefHandler, {**handler_opts}),
             (r"/api/v1/lookup/grid", APILookupGridHandler, {**handler_opts}),
-            (r"/api/v1/spot", APISpotHandler, {"spots": self._spots, **handler_opts}),
+            (r"/api/v1/spot", APISpotHandler, {"spots": self._data_store.spots, **handler_opts}),
         ]
 
         # If in API-only mode, serve a basic homepage; in normal mode, serve the usual UI routes
@@ -121,55 +124,6 @@ class WebServer:
         logging.info("You can access your copy of Spothole at " + BASE_URL)
         await self._shutdown_event.wait()
 
-    def notify_new_spot(self, spot):
-        """Internal method called when a new spot is added to the system. This is used to ping any SSE clients that are
-        awaiting a server-sent message with new spots."""
-
-        for queue in self._sse_spot_queues:
-            try:
-                queue.put(spot)
-            except:
-                # Cleanup thread was probably deleting the queue, that's fine
-                pass
-        pass
-
-    def notify_new_alert(self, alert):
-        """Internal method called when a new alert is added to the system. This is used to ping any SSE clients that are
-        awaiting a server-sent message with new spots."""
-
-        for queue in self._sse_alert_queues:
-            try:
-                queue.put(alert)
-            except:
-                # Cleanup thread was probably deleting the queue, that's fine
-                pass
-        pass
-
-    def clean_up_sse_queues(self):
-        """Clean up any SSE queues that are growing too large; probably their client disconnected and we didn't catch it
-        properly for some reason."""
-
-        for q in self._sse_spot_queues:
-            try:
-                if q.full():
-                    logging.warning(
-                        "A full SSE spot queue was found, presumably because the client disconnected strangely. It has been removed.")
-                    self._sse_spot_queues.remove(q)
-                    empty_queue(q)
-            except:
-                # Probably got deleted already on another thread
-                pass
-        for q in self._sse_alert_queues:
-            try:
-                if q.full():
-                    logging.warning(
-                        "A full SSE alert queue was found, presumably because the client disconnected strangely. It has been removed.")
-                    self._sse_alert_queues.remove(q)
-                    empty_queue(q)
-            except:
-                # Probably got deleted already on another thread
-                pass
-        pass
 
 def request_log(handler):
     """Custom log function to provide more data about requests when enabled, and to provide the ability to turn off

sigrefdataproviders/arlhs.py

@@ -0,0 +1,30 @@
+import csv
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class ARLHS(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Amateur Radio Light House Society"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "ARLHS"
+    DATA_URL = "https://www.gma.rocks/download/lighthouse.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines()[1:]):
+            if "ARLHS" in row and row["ARLHS"] != "":
+                ref_id = row["ARLHS"]
+                new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=row["Name"] if "Name" in row else None,
+                                          url="https://www.cqgma.org/zinfo.php?ref=" + ref_id,
+                                          latitude=float(row["Latitude"]) if "Latitude" in row and row[
+                                              "Latitude"] != "" else None,
+                                          longitude=float(row["Longitude"]) if "Longitude" in row and row[
+                                              "Longitude"] != "" else None,
+                                          grid=row["Maidenhead Locator"]))
+
+        return new_data

sigrefdataproviders/dme.py

@@ -0,0 +1,35 @@
+import csv
+
+from pyhamtools.locator import latlong_to_locator
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.local_file_sig_ref_data_provider import LocalFileSIGRefDataProvider
+
+
+class DME(LocalFileSIGRefDataProvider):
+    """SIG ref data provider for Diploma Municipios de Espana"""
+
+    SIG = "DME"
+    PATH = "datafiles/MUNICIPIOS.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.PATH)
+
+    def _file_to_data(self, path):
+        new_data = []
+        with open(path, encoding="latin-1") as _f:
+            for row in csv.DictReader(_f, delimiter=";"):
+                ref_id = row["COD_INE"][:5]
+                latitude = float(row["LATITUD_ETRS89_REGCAN95"].replace(",", ".")) if row.get(
+                    "LATITUD_ETRS89_REGCAN95") else None
+                longitude = float(row["LONGITUD_ETRS89_REGCAN95"].replace(",", ".")) if row.get(
+                    "LONGITUD_ETRS89_REGCAN95") else None
+
+                ref = SIGRef(sig=self.SIG, id=ref_id,
+                                          name=row["NOMBRE_ACTUAL"] + ", " + row["PROVINCIA"],
+                                          latitude=latitude,
+                                          longitude=longitude)
+                if latitude and longitude:
+                    ref.grid = latlong_to_locator(latitude, longitude, 6)
+                new_data.append(ref)
+        return new_data

sigrefdataproviders/file_download_sig_ref_data_provider.py

@@ -0,0 +1,77 @@
+import logging
+from datetime import datetime
+from threading import Thread, Event
+
+import pytz
+from requests import ReadTimeout
+from requests.exceptions import ConnectionError, ConnectTimeout
+
+from core.constants import HTTP_HEADERS
+from core.url_data_cache import URL_DATA_CACHE
+from sigrefdataproviders.sig_ref_data_provider import SIGRefDataProvider
+
+
+class FileDownloadSIGRefDataProvider(SIGRefDataProvider):
+    """Generic SIG ref data provider class for providers that fetch their data from the web by downloading a file."""
+
+    def __init__(self, sig_name, provider_config, url, poll_interval):
+        """ Set up the provider, note poll_interval is in *days*."""
+        super().__init__(sig_name, provider_config)
+        self._url = url
+        self._poll_interval = poll_interval
+        self._thread = None
+        self._stop_event = Event()
+
+    def start(self):
+        # Fire off the polling thread. It will poll immediately on startup, then sleep for poll_interval between
+        # subsequent polls, so start() returns immediately and the application can continue starting.
+        logging.info(
+            "Set up query of " + self.sig_name + " SIG ref data every " + str(self._poll_interval) + " days.")
+        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 * 60 * 60 * 24):
+                break
+
+    def _poll(self):
+        try:
+            # Request data from API. Use the data cache (with a TTL of 1 day) here, not as the main mechanism for
+            # caching, but just so continual restarts of the software during testing don't hammer the servers.
+            logging.debug("Downloading " + self.sig_name + " SIG ref data...")
+            http_response = URL_DATA_CACHE.get(self._url, headers=HTTP_HEADERS)
+            # Check response code was good
+            if http_response.ok:
+                # Pass off to the subclass for processing
+                new_data = self._http_response_to_data(http_response)
+                # Submit the new spots for processing. There might not be any spots for the less popular programs.
+                if new_data:
+                    self._add_data(new_data)
+
+                self.status = "OK"
+                self.last_update_time = datetime.now(pytz.UTC)
+                logging.debug("Received SIG ref data for " + self.sig_name)
+            else:
+                self.status = "Error"
+                logging.warning(f"HTTP {http_response.status_code} when downloading SIG ref data for {self.sig_name}.")
+
+        except ConnectionError:
+            logging.warning(f"Connection error when downloading SIG ref data for {self.sig_name}.")
+        except (ConnectTimeout, ReadTimeout):
+            logging.warning(f"Timeout when downloading SIG ref data for {self.sig_name}.")
+        except Exception:
+            self.status = "Error"
+            logging.exception("Exception in HTTP SIG Ref Data Provider (" + self.sig_name + ")")
+            self._stop_event.wait(timeout=1)
+
+    def _http_response_to_data(self, http_response):
+        """Convert an HTTP response returned by the server into SIG Ref data. The whole response is provided here so the
+        subclass implementations can check for HTTP status codes if necessary, and handle the response as JSON, CSV,
+        whatever the remote file actually is."""
+
+        raise NotImplementedError("Subclasses must implement this method")

sigrefdataproviders/gma.py

@@ -0,0 +1,29 @@
+import csv
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class GMA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Global Mountain Activity"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "GMA"
+    DATA_URL = "https://www.gma.rocks/download/summits.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines()[1:]):
+            ref_id = row["Reference"]
+            new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=row["Name"] if "Name" in row else None,
+                                      url="https://www.cqgma.org/zinfo.php?ref=" + ref_id,
+                                      latitude=float(row["Latitude"]) if "Latitude" in row and row[
+                                          "Latitude"] != "" else None,
+                                      longitude=float(row["Longitude"]) if "Longitude" in row and row[
+                                          "Longitude"] != "" else None,
+                                      grid=row["Maidenhead Locator"]))
+
+        return new_data

sigrefdataproviders/illw.py

@@ -0,0 +1,30 @@
+import csv
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class ILLW(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for International Lighthouse & Lightship Weekend"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "ILLW"
+    DATA_URL = "https://www.gma.rocks/download/lighthouse.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines()[1:]):
+            if "ILLW" in row and row["ILLW"] != "":
+                ref_id = row["ILLW"]
+                new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=row["Name"] if "Name" in row else None,
+                                          url="https://www.cqgma.org/zinfo.php?ref=" + ref_id,
+                                          latitude=float(row["Latitude"]) if "Latitude" in row and row[
+                                              "Latitude"] != "" else None,
+                                          longitude=float(row["Longitude"]) if "Longitude" in row and row[
+                                              "Longitude"] != "" else None,
+                                          grid=row["Maidenhead Locator"]))
+
+        return new_data

sigrefdataproviders/iota.py

@@ -0,0 +1,36 @@
+import logging
+
+from pyhamtools.locator import latlong_to_locator
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class IOTA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Islands on the Air"""
+
+    POLL_INTERVAL_DAYS = 365
+    SIG = "IOTA"
+    DATA_URL = "https://www.iota-world.org/islands-on-the-air/downloads/download-file.html?path=groups.json"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        data = http_response.json()
+        if isinstance(data, list):
+            for ref in data:
+                ref_id = ref["refno"]
+                latitude = float(ref["latitude_min"]) + float(ref["latitude_max"]) / 2.0
+                longitude = float(ref["longitude_min"]) + float(ref["longitude_max"]) / 2.0
+                grid = None
+                try:
+                    grid = latlong_to_locator(latitude, longitude, 6)
+                except ValueError:
+                    logging.debug(f"Error converting lat/lon to locator for an IOTA reference %f %f", latitude, longitude)
+
+                new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=ref["name"],
+                                          grid=grid, latitude=latitude, longitude=longitude))
+
+        return new_data

sigrefdataproviders/llota.py

@@ -0,0 +1,32 @@
+from pyhamtools.locator import locator_to_latlong
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class LLOTA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Lagos y Lagunas on the Air"""
+
+    POLL_INTERVAL_DAYS = 7
+    SIG = "LLOTA"
+    DATA_URL = "https://llota.app/api/public/references"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        data = http_response.json()
+        if isinstance(data, list):
+            for ref in data:
+                ref_id = ref["reference_code"]
+                grid = str(ref["grid_locator"])
+                ll = locator_to_latlong(grid)
+
+                new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=str(ref["name"]),
+                                          url="https://llota.app/list/ref/" + ref_id,
+                                          grid=grid,
+                                          latitude=ll[0],
+                                          longitude=ll[1]))
+
+        return new_data

sigrefdataproviders/local_file_sig_ref_data_provider.py

@@ -0,0 +1,36 @@
+import logging
+from datetime import datetime
+
+import pytz
+
+from sigrefdataproviders.sig_ref_data_provider import SIGRefDataProvider
+
+
+class LocalFileSIGRefDataProvider(SIGRefDataProvider):
+    """Generic SIG ref data provider class for providers that fetch their data from a local file on startup."""
+
+    def __init__(self, sig, provider_config, path):
+        super().__init__(sig, provider_config)
+        self._path = path
+
+    def start(self):
+        logging.debug("Loading " + self.sig_name + " SIG ref data from file.")
+        try:
+            new_data = self._file_to_data(self._path)
+            if new_data:
+                self._add_data(new_data)
+                self.status = "OK"
+                self.last_update_time = datetime.now(pytz.UTC)
+            else:
+                logging.info("No new SIG ref data found for " + self.sig_name)
+        except Exception as e:
+            self.status = "Error"
+            logging.exception("Exception in local file SIG Ref Data Provider (" + self.sig_name + ")")
+
+    def stop(self):
+        pass
+
+    def _file_to_data(self, path):
+        """Load a file on the given path and turn it into SIG Ref data."""
+
+        raise NotImplementedError("Subclasses must implement this method")

sigrefdataproviders/mota.py

@@ -0,0 +1,29 @@
+import csv
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class MOTA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Mills on the Air"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "MOTA"
+    DATA_URL = "https://www.gma.rocks/download/mills.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines()[1:]):
+            ref_id = row["Reference"]
+            new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=row["Name"] if "Name" in row else None,
+                                      url="https://www.cqgma.org/zinfo.php?ref=" + ref_id,
+                                      latitude=float(row["Latitude"]) if "Latitude" in row and row[
+                                          "Latitude"] != "" else None,
+                                      longitude=float(row["Longitude"]) if "Longitude" in row and row[
+                                          "Longitude"] != "" else None,
+                                      grid=row["Maidenhead Locator"]))
+
+        return new_data

sigrefdataproviders/pota.py

@@ -0,0 +1,29 @@
+import csv
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class POTA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Parks on the Air"""
+
+    POLL_INTERVAL_DAYS = 7
+    SIG = "POTA"
+    DATA_URL = "https://pota.app/all_parks_ext.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines()):
+            ref_id = row["reference"]
+            new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=row["name"] if "name" in row else None,
+                                      url="https://pota.app/#/park/" + ref_id,
+                                      grid=row["grid"] if "grid" in row else None,
+                                      latitude=float(row["latitude"]) if "latitude" in row and row[
+                                          "latitude"] != "" else None,
+                                      longitude=float(row["longitude"]) if "longitude" in row and row[
+                                          "longitude"] != "" else None))
+
+        return new_data

sigrefdataproviders/sig_ref_data_provider.py

@@ -0,0 +1,40 @@
+import logging
+from datetime import datetime
+
+import pytz
+
+from core.data_store import DATA_STORE
+
+
+class SIGRefDataProvider:
+    """Generic SIG reference data provider class. Subclasses of this query the individual URLs or files for data."""
+
+    def __init__(self, sig_name, provider_config):
+        """Constructor"""
+
+        self.sig_name = sig_name
+        self.enabled = provider_config["enabled"]
+        self.last_update_time = datetime.min.replace(tzinfo=pytz.UTC)
+        self.last_spot_time = datetime.min.replace(tzinfo=pytz.UTC)
+        self.status = "Not Started" if self.enabled else "Disabled"
+        self.reference_count = 0
+
+
+    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 _add_data(self, new_data):
+        """Add all the provided reference data objects to the data store."""
+
+        for d in new_data:
+            DATA_STORE.sigrefs[self.sig_name + ":" + d.id] = d
+        self.reference_count = len(new_data)
+        logging.info(f"Loaded %d references for %s into the data store.", self.reference_count, self.sig_name)

sigrefdataproviders/siota.py

@@ -0,0 +1,26 @@
+import csv
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class SIOTA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Silos on the Air"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "SIOTA"
+    DATA_URL = "https://www.silosontheair.com/data/silos.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines()):
+            ref_id = row["SILO_CODE"]
+            new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=row["NAME"] if "NAME" in row else None,
+                                      grid=row["LOCATOR"] if "LOCATOR" in row else None,
+                                      latitude=float(row["LAT"]) if "LAT" in row else None,
+                                      longitude=float(row["LNG"]) if "LNG" in row else None))
+
+        return new_data

sigrefdataproviders/sota.py

@@ -0,0 +1,34 @@
+import csv
+
+from pyhamtools.locator import latlong_to_locator
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class SOTA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Summits on the Air"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "SOTA"
+    DATA_URL = "https://storage.sota.org.uk/summitslist.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines()[1:]):
+            ref_id = row["SummitCode"]
+            latitude = float(row["Latitude"]) if "Latitude" in row and row["Latitude"] != "" else None
+            longitude = float(row["Longitude"]) if "Longitude" in row and row["Longitude"] != "" else None
+            ref = SIGRef(sig=self.SIG, id=ref_id, name=row["SummitName"] if "SummitName" in row else None,
+                                      url="https://www.sotadata.org.uk/en/summit/" + ref_id,
+                                      latitude=latitude,
+                                      longitude=longitude,
+                                      activation_score=int(row["Points"]) if "Points" in row else None)
+            if latitude and longitude:
+                ref.grid = latlong_to_locator(latitude, longitude, 6)
+            new_data.append(ref)
+
+        return new_data

sigrefdataproviders/toilets.py

@@ -0,0 +1,24 @@
+import csv
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.local_file_sig_ref_data_provider import LocalFileSIGRefDataProvider
+
+
+class Toilets(LocalFileSIGRefDataProvider):
+    """SIG ref data provider for Toilets on the Air"""
+
+    SIG = "Toilets"
+    PATH = "datafiles/toilets.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.PATH)
+
+    def _file_to_data(self, path):
+        new_data = []
+        f = open(path)
+        csv_data = f.read()
+        dr = csv.DictReader(csv_data.splitlines())
+        for row in dr:
+            new_data.append(SIGRef(sig=self.SIG, id=row["ref"], name=row["ref"], latitude=float(row["lat"]),
+                                          longitude=float(row["lon"])))
+        return new_data

sigrefdataproviders/towers.py

@@ -0,0 +1,27 @@
+import csv
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class Towers(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Towers on the Air"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "Towers"
+    DATA_URL = "https://wwtota.com/servis/generate_csv.php?ref=&filter=all"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines(), delimiter=";"):
+            ref_id = row["Ref"]
+            new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=row["Nazev"] if "Nazev" in row else None,
+                                      url="https://wwtota.com/seznam/karta_rozhledny.php?ref=" + ref_id,
+                                      grid=row["Lokator"] if "Lokator" in row and row["Lokator"] != "" else None,
+                                      latitude=float(row["Lat"]) if "Lat" in row and row["Lat"] != "" else None,
+                                      longitude=float(row["Lon"]) if "Lon" in row and row["Lon"] != "" else None))
+
+        return new_data

sigrefdataproviders/wca.py

@@ -0,0 +1,44 @@
+import csv
+import logging
+
+from pyhamtools.locator import latlong_to_locator
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class WCA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for World Castles Award"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "WCA"
+    DATA_URL = "https://polo.ham2k.com/data/activities/wca/all-castles.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines()):
+            ref_id = row["REF"]
+
+            coordsStr = row["COORDINATES"]
+            latitude = None
+            longitude = None
+            grid = None
+            try:
+                if coordsStr:
+                    split = coordsStr.split(", ")
+                    latitude = float(split[0])
+                    longitude = float(split[1])
+                    grid = latlong_to_locator(latitude, longitude)
+            except ValueError:
+                logging.debug(f"Encountered dodgy formatting in WCA CSV, skipping location data for %s", ref_id)
+
+            new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=row["CLEAN NAME"] if "CLEAN NAME" in row else None,
+                                      url="https://www.cqgma.org/zinfo.php?ref=" + ref_id,
+                                      latitude=latitude,
+                                      longitude=longitude,
+                                      grid=grid))
+
+        return new_data

sigrefdataproviders/wota.py

@@ -0,0 +1,31 @@
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class WOTA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Wainwrights on the Air"""
+
+    POLL_INTERVAL_DAYS = 365
+    SIG = "WOTA"
+    DATA_URL = "https://www.wota.org.uk/mapping/data/summits.json"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for feature in http_response.json().get("features", []):
+            ref_id = feature["properties"]["wotaId"]
+            # Fudge WOTA URLs. Outlying fell (LDO) URLs don't match their ID numbers but require 214 to be
+            # added to them
+            url = "https://www.wota.org.uk/MM_" + ref_id
+            if ref_id.upper().startswith("LDO-"):
+                number = int(ref_id.upper().replace("LDO-", ""))
+                url = "https://www.wota.org.uk/MM_LDO-" + str(number + 214)
+
+            new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=feature["properties"]["title"], url=url,
+                                      grid=feature["properties"]["qthLocator"],
+                                      latitude=feature["geometry"]["coordinates"][1],
+                                      longitude=feature["geometry"]["coordinates"][0]))
+
+        return new_data

sigrefdataproviders/wwbota.py

@@ -0,0 +1,27 @@
+import csv
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class WWBOTA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Worldwide Bunkers on the Air"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "WWBOTA"
+    DATA_URL = "https://api.wwbota.org/bunkers/?format=CSV"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines()):
+            ref_id = row["Reference"]
+            new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=row["Name"] if "Name" in row else None,
+                                      url="https://bunkerwiki.org/?s=" + ref_id if ref_id.startswith("B/G") else None,
+                                      grid=row["Locator"] if "Locator" in row and row["Locator"] != "" else None,
+                                      latitude=float(row["Lat"]) if "Lat" in row and row["Lat"] != "" else None,
+                                      longitude=float(row["Long"]) if "Long" in row and row["Long"] != "" else None))
+
+        return new_data

sigrefdataproviders/wwff.py

@@ -0,0 +1,30 @@
+import csv
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class WWFF(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for Worldwide Flora & Fauna"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "WWFF"
+    DATA_URL = "https://wwff.co/wwff-data/wwff_directory.csv"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        for row in csv.DictReader(http_response.content.decode("utf-8-sig").splitlines()):
+            ref_id = row["reference"]
+            new_data.append(SIGRef(sig=self.SIG, id=ref_id, name=row["name"] if "name" in row else None,
+                                      url="https://wwff.co/directory/?showRef=" + ref_id,
+                                      grid=row["iaruLocator"] if "iaruLocator" in row and row[
+                                          "iaruLocator"] != "-" else None,
+                                      latitude=float(row["latitude"]) if "latitude" in row and row[
+                                          "latitude"] != "" and row["latitude"] != "-" else None,
+                                      longitude=float(row["longitude"]) if "longitude" in row and row[
+                                          "longitude"] != "" and row["longitude"] != "-" else None))
+
+        return new_data

sigrefdataproviders/zlota.py

@@ -0,0 +1,34 @@
+from pyhamtools.locator import latlong_to_locator
+
+from data.sig_ref import SIGRef
+from sigrefdataproviders.file_download_sig_ref_data_provider import FileDownloadSIGRefDataProvider
+
+
+class ZLOTA(FileDownloadSIGRefDataProvider):
+    """SIG ref data provider for New Zealand on the Air"""
+
+    POLL_INTERVAL_DAYS = 30
+    SIG = "ZLOTA"
+    DATA_URL = "https://ontheair.nz/assets/assets.json"
+
+    def __init__(self, provider_config):
+        super().__init__(self.SIG, provider_config, self.DATA_URL, self.POLL_INTERVAL_DAYS)
+
+    def _http_response_to_data(self, http_response):
+        new_data = []
+        data = http_response.json()
+        if isinstance(data, list):
+            for ref in data:
+                ref_id = ref["code"]
+                latitude = ref["y"]
+                longitude = ref["x"]
+
+                ref = SIGRef(sig=self.SIG, id=ref_id, name=ref["name"],
+                                          url="https://ontheair.nz/assets/" + ref_id.replace("/", "_"),
+                                          latitude=latitude,
+                                          longitude=longitude)
+                if latitude and longitude:
+                    ref.grid = latlong_to_locator(latitude, longitude, 6)
+                new_data.append(ref)
+
+        return new_data

solarconditionsproviders/giroionosonde.py

@@ -30,11 +30,23 @@ class GIROIonosonde(SolarConditionsProvider):
     data, but is less reliable and often offline."""
 
     def __init__(self, provider_config):
-        super().__init__(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 = []
@@ -44,21 +56,6 @@ class GIROIonosonde(SolarConditionsProvider):
                     stations.append({"ursi": row[0].strip(), "name": row[1].strip()})
         return stations
 
-    def setup(self, solar_conditions, solar_conditions_cache):
-        """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."""
-
-        super().setup(solar_conditions, solar_conditions_cache)
-        existing = 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}})
-
     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)

solarconditionsproviders/hamqsl.py

@@ -15,7 +15,7 @@ class HamQSL(HTTPSolarConditionsProvider):
     Provides solar flux index, geomagnetic indices, and HF/VHF propagation condition summaries."""
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, URL, POLL_INTERVAL)
+        super().__init__("HamQSL", provider_config, URL, POLL_INTERVAL)
 
     def _http_response_to_solar_conditions(self, http_response):
         root = ElementTree.fromstring(http_response.text)

solarconditionsproviders/http_solar_conditions_provider.py

@@ -14,8 +14,8 @@ 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, provider_config, url, poll_interval):
-        super().__init__(provider_config)
+    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

solarconditionsproviders/kc2gprop.py

@@ -24,7 +24,7 @@ class KC2GProp(SolarConditionsProvider):
     online, but has fewer stations and does not provide LUF data."""
 
     def __init__(self, provider_config):
-        super().__init__(provider_config)
+        super().__init__("KC2G Propagation Data", provider_config)
         self._thread = None
         self._stop_event = Event()
 

solarconditionsproviders/noaa3dayforecast.py

@@ -13,7 +13,7 @@ class NOAA3dayForecast(HTTPSolarConditionsProvider):
     corresponding fields in the solar conditions object.."""
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, URL, POLL_INTERVAL)
+        super().__init__("NOAA 3-day Forecast", provider_config, URL, POLL_INTERVAL)
 
     @staticmethod
     def _parse_percentage_table(lines, section_header, year):

solarconditionsproviders/solar_conditions_provider.py

@@ -2,26 +2,21 @@ 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, provider_config):
+    def __init__(self, name, provider_config):
         """Constructor"""
 
-        self._solar_conditions_cache = None
-        self.name = provider_config["name"]
+        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 = None
-
-    def setup(self, solar_conditions, solar_conditions_cache):
-        """Set up the provider, giving it the solar conditions object and its backing cache"""
-
-        self._solar_conditions = solar_conditions
-        self._solar_conditions_cache = solar_conditions_cache
+        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"""
@@ -41,4 +36,3 @@ class SolarConditionsProvider:
                 if hasattr(self._solar_conditions, key):
                     setattr(self._solar_conditions, key, value)
             self._solar_conditions.infer_descriptions()
-            self._solar_conditions_cache['solar_conditions'] = self._solar_conditions

spothole.py

@@ -1,30 +1,23 @@
 # Main script
-import importlib
 import logging
 import os
 import signal
 import sys
 
-from diskcache import Cache
-
-from core.cleanup import CleanupTimer
-from core.config import config, SERVER_OWNER_CALLSIGN, LOG_LEVEL
+from core.config import config, SERVER_OWNER_CALLSIGN, LOG_LEVEL, get_sig_ref_data_provider_from_config, \
+    get_spot_provider_from_config, get_alert_provider_from_config, get_solar_conditions_provider_from_config
 from core.constants import SOFTWARE_VERSION
+from core.data_store import DATA_STORE
 from core.lookup_helper import lookup_helper
 from core.status_reporter import StatusReporter
-from data.solar_conditions import SolarConditions
 from server.webserver import WebServer
 
 # Globals
-spots = Cache('cache/spots_cache')
-alerts = Cache('cache/alerts_cache')
-solar_conditions_cache = Cache('cache/solar_conditions_cache')
-solar_conditions = solar_conditions_cache.get('solar_conditions', SolarConditions())
 web_server = None
-status_data = {}
 spot_providers = []
 alert_providers = []
 solar_condition_providers = []
+sig_ref_data_providers = []
 cleanup_timer = None
 run = True
 
@@ -46,40 +39,13 @@ def shutdown(_signum=None, _frame=None):
     for scp in solar_condition_providers:
         if scp.enabled:
             scp.stop()
-    if cleanup_timer:
-        cleanup_timer.stop()
-    if lookup_helper:
-        lookup_helper.stop()
-    spots.close()
-    alerts.close()
-    solar_conditions_cache.close()
+    for srdp in sig_ref_data_providers:
+        if srdp.enabled:
+            srdp.stop()
+    DATA_STORE.close()
     os._exit(0)
 
 
-def get_spot_provider_from_config(config_providers_entry):
-    """Utility method to get a spot provider based on the class specified in its config entry."""
-
-    module = importlib.import_module('spotproviders.' + config_providers_entry["class"].lower())
-    provider_class = getattr(module, config_providers_entry["class"])
-    return provider_class(config_providers_entry)
-
-
-def get_alert_provider_from_config(config_providers_entry):
-    """Utility method to get an alert provider based on the class specified in its config entry."""
-
-    module = importlib.import_module('alertproviders.' + config_providers_entry["class"].lower())
-    provider_class = getattr(module, config_providers_entry["class"])
-    return provider_class(config_providers_entry)
-
-
-def get_solar_conditions_provider_from_config(config_providers_entry):
-    """Utility method to get a solar conditions provider based on the class specified in its config entry."""
-
-    module = importlib.import_module('solarconditionsproviders.' + config_providers_entry["class"].lower())
-    provider_class = getattr(module, config_providers_entry["class"])
-    return provider_class(config_providers_entry)
-
-
 # Main function
 if __name__ == '__main__':
     # Set up logging
@@ -99,17 +65,19 @@ if __name__ == '__main__':
     # Shut down gracefully on SIGINT
     signal.signal(signal.SIGINT, shutdown)
 
+    # Set up data store
+    DATA_STORE.setup()
+
     # Set up lookup helper
     lookup_helper.start()
 
     # Set up web server
-    web_server = WebServer(spots=spots, alerts=alerts, solar_conditions=solar_conditions, status_data=status_data)
+    web_server = WebServer()
 
     # Fetch, set up and start spot providers
     for entry in config["spot-providers"]:
         spot_providers.append(get_spot_provider_from_config(entry))
     for p in spot_providers:
-        p.setup(spots=spots, web_server=web_server)
         if p.enabled:
             p.start()
 
@@ -117,7 +85,6 @@ if __name__ == '__main__':
     for entry in config["alert-providers"]:
         alert_providers.append(get_alert_provider_from_config(entry))
     for p in alert_providers:
-        p.setup(alerts=alerts, web_server=web_server)
         if p.enabled:
             p.start()
 
@@ -125,18 +92,20 @@ if __name__ == '__main__':
     for entry in config.get("solar-condition-providers", []):
         solar_condition_providers.append(get_solar_conditions_provider_from_config(entry))
     for p in solar_condition_providers:
-        p.setup(solar_conditions=solar_conditions, solar_conditions_cache=solar_conditions_cache)
         if p.enabled:
             p.start()
 
-    # Set up timer to clear spot list of old data
-    cleanup_timer = CleanupTimer(spots=spots, alerts=alerts, web_server=web_server, cleanup_interval=60)
-    cleanup_timer.start()
+    # Fetch, set up and start SIG reference data providers
+    for entry in config.get("sig-ref-data-providers", []):
+        sig_ref_data_providers.append(get_sig_ref_data_provider_from_config(entry))
+    for p in sig_ref_data_providers:
+        if p.enabled:
+            p.start()
 
     # Set up status reporter
-    status_reporter = StatusReporter(status_data=status_data, spots=spots, alerts=alerts, web_server=web_server,
-                                     cleanup_timer=cleanup_timer, spot_providers=spot_providers,
+    status_reporter = StatusReporter(web_server=web_server, spot_providers=spot_providers,
                                      alert_providers=alert_providers,
+                                     sig_ref_data_providers=sig_ref_data_providers,
                                      solar_condition_providers=solar_condition_providers, run_interval=5)
     status_reporter.start()
 

spotproviders/aprsis.py

@@ -14,7 +14,7 @@ class APRSIS(SpotProvider):
     """Spot provider for the APRS-IS."""
 
     def __init__(self, provider_config):
-        super().__init__(provider_config)
+        super().__init__("APRS-IS", provider_config)
         self._thread = Thread(target=self._connect)
         self._thread.daemon = True
         self._aprsis = None

spotproviders/dxcluster.py

@@ -26,7 +26,8 @@ class DXCluster(SpotProvider):
     def __init__(self, provider_config):
         """Constructor requires hostname and port"""
 
-        super().__init__(provider_config)
+        name = provider_config["name"] if "name" in provider_config else "Cluster"
+        super().__init__(name, provider_config)
         self._hostname = provider_config["host"]
         self._port = provider_config["port"]
         self._login_prompt = provider_config["login_prompt"] if "login_prompt" in provider_config else "login:"
@@ -59,6 +60,10 @@ class DXCluster(SpotProvider):
                     self._telnet.write((self._login_callsign + "\n").encode("latin-1"))
                     connected = True
                     logging.info("DX Cluster " + self._hostname + " connected.")
+                except ConnectionRefusedError:
+                    self.status = "Error"
+                    logging.warning("Connection refused to DX cluster " + self._hostname)
+                    sleep(300)
                 except Exception:
                     self.status = "Error"
                     logging.exception("Exception while connecting to DX Cluster Provider (" + self._hostname + ").")

spotproviders/gma.py

@@ -3,8 +3,8 @@ from datetime import datetime
 
 import pytz
 
-from core.cache_utils import SEMI_STATIC_URL_DATA_CACHE
 from core.constants import HTTP_HEADERS
+from core.url_data_cache import URL_DATA_CACHE
 from data.sig_ref import SIGRef
 from data.spot import Spot
 from spotproviders.http_spot_provider import HTTPSpotProvider
@@ -26,7 +26,7 @@ class GMA(HTTPSpotProvider):
             provider_config["enabled"] = False
             logging.warning("GMA spot provider configured but no api key was provided, this API will not be queried.")
 
-        super().__init__(provider_config, self.SPOTS_URL + "?key=" + self.api_key, self.POLL_INTERVAL_SEC)
+        super().__init__("GMA", provider_config, self.SPOTS_URL + "?key=" + self.api_key, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_spots(self, http_response):
         new_spots = []
@@ -56,8 +56,8 @@ class GMA(HTTPSpotProvider):
                 # GMA doesn't give what programme (SIG) the reference is for until we separately look it up.
                 if "REF" in source_spot:
                     try:
-                        ref_response = SEMI_STATIC_URL_DATA_CACHE.get(self.REF_INFO_URL_ROOT + source_spot["REF"],
-                                                                      headers=HTTP_HEADERS)
+                        ref_response = URL_DATA_CACHE.get(self.REF_INFO_URL_ROOT + source_spot["REF"],
+                                                          headers=HTTP_HEADERS)
                         # Sometimes this is blank even if it's a 200 response, so handle that
                         if ref_response.ok and ref_response.text is not None and ref_response.text != "":
                             ref_info = ref_response.json()

spotproviders/hema.py

@@ -25,7 +25,7 @@ class HEMA(HTTPSpotProvider):
     SPOTTER_COMMENT_PATTERN = re.compile("^\\((.*)\\) (.*)$")
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOT_SEED_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("HEMA", provider_config, self.SPOT_SEED_URL, self.POLL_INTERVAL_SEC)
         self._spot_seed = ""
 
     def _http_response_to_spots(self, http_response):

spotproviders/http_spot_provider.py

@@ -15,8 +15,8 @@ class HTTPSpotProvider(SpotProvider):
     """Generic spot provider class for providers that request data via HTTP(S). Just for convenience to avoid code
     duplication. Subclasses of this query the individual APIs for data."""
 
-    def __init__(self, provider_config, url, poll_interval):
-        super().__init__(provider_config)
+    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
@@ -64,7 +64,7 @@ class HTTPSpotProvider(SpotProvider):
             logging.warning(f"Timeout when accessing {self.name} spots API.")
         except Exception:
             self.status = "Error"
-            logging.exception("Exception in HTTP JSON Spot Provider (" + self.name + ")")
+            logging.exception("Exception in HTTP Spot Provider (" + self.name + ")")
             self._stop_event.wait(timeout=1)
 
     def _http_response_to_spots(self, http_response):

spotproviders/llota.py

@@ -12,7 +12,7 @@ class LLOTA(HTTPSpotProvider):
     SPOTS_URL = "https://llota.app/api/public/spots"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("LLOTA", provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_spots(self, http_response):
         new_spots = []

spotproviders/parksnpeaks.py

@@ -17,7 +17,7 @@ class ParksNPeaks(HTTPSpotProvider):
     SIOTA_LIST_URL = "https://www.silosontheair.com/data/silos.csv"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("ParksNPeaks", provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_spots(self, http_response):
         new_spots = []

spotproviders/pota.py

@@ -14,7 +14,7 @@ class POTA(HTTPSpotProvider):
     SPOTS_URL = "https://api.pota.app/spot/activator"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("POTA", provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_spots(self, http_response):
         new_spots = []

spotproviders/rbn.py

@@ -23,7 +23,8 @@ class RBN(SpotProvider):
     def __init__(self, provider_config):
         """Constructor requires port number."""
 
-        super().__init__(provider_config)
+        name = provider_config["name"] if "name" in provider_config else "RBN"
+        super().__init__(name, provider_config)
         self._port = provider_config["port"]
         self._telnet = None
         self._thread = Thread(target=self._handle)

spotproviders/sota.py

@@ -19,11 +19,9 @@ class SOTA(HTTPSpotProvider):
     # The actual data lookup all happens after parsing and checking the epoch.
     EPOCH_URL = "https://api-db2.sota.org.uk/api/spots/epoch"
     SPOTS_URL = "https://api-db2.sota.org.uk/api/spots/60/all/all"
-    # SOTA spots don't contain lat/lon, we need a separate lookup for that
-    SUMMIT_URL_ROOT = "https://api-db2.sota.org.uk/api/summits/"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.EPOCH_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("SOTA", provider_config, self.EPOCH_URL, self.POLL_INTERVAL_SEC)
         self._api_epoch = ""
 
     def _http_response_to_spots(self, http_response):

spotproviders/spot_provider.py

@@ -2,28 +2,21 @@ from datetime import datetime
 
 import pytz
 
-from core.config import MAX_SPOT_AGE
+from core.data_store import DATA_STORE
 
 
 class SpotProvider:
     """Generic spot provider class. Subclasses of this query the individual APIs for data."""
 
-    def __init__(self, provider_config):
+    def __init__(self, name, provider_config):
         """Constructor"""
 
-        self.name = provider_config["name"]
+        self.name = name
         self.enabled = provider_config["enabled"]
         self.last_update_time = datetime.min.replace(tzinfo=pytz.UTC)
         self.last_spot_time = datetime.min.replace(tzinfo=pytz.UTC)
         self.status = "Not Started" if self.enabled else "Disabled"
-        self._spots = None
-        self._web_server = None
-
-    def setup(self, spots, web_server):
-        """Set up the provider, e.g. giving it the spot list to work from"""
-
-        self._spots = spots
-        self._web_server = web_server
+        self._spots = DATA_STORE.spots
 
     def start(self):
         """Start the provider. This should return immediately after spawning threads to access the remote resources"""
@@ -59,10 +52,7 @@ class SpotProvider:
 
     def _add_spot(self, spot):
         if not spot.expired():
-            self._spots.add(spot.id, spot, expire=MAX_SPOT_AGE)
-            # Ping the web server in case we have any SSE connections that need to see this immediately
-            if self._web_server:
-                self._web_server.notify_new_spot(spot)
+            self._spots.set(spot.id, spot)
 
     def stop(self):
         """Stop any threads and prepare for application shutdown"""

spotproviders/sse_spot_provider.py

@@ -13,8 +13,8 @@ from spotproviders.spot_provider import SpotProvider
 class SSESpotProvider(SpotProvider):
     """Spot provider using Server-Sent Events."""
 
-    def __init__(self, provider_config, url):
-        super().__init__(provider_config)
+    def __init__(self, name, provider_config, url):
+        super().__init__(name, provider_config)
         self._url = url
         self._event_source = None
         self._thread = None

spotproviders/tiles.py

@@ -12,7 +12,7 @@ class Tiles(HTTPSpotProvider):
     SPOTS_URL = "https://icneuzxitdqtofutxbla.supabase.co/functions/v1/spots?active_hours=24"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("Tiles", provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_spots(self, http_response):
         new_spots = []

spotproviders/towers.py

@@ -13,7 +13,7 @@ class Towers(HTTPSpotProvider):
     SPOTS_URL = "https://wwtota.com/api/cluster_live.php"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("Towers", provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_spots(self, http_response):
         new_spots = []
@@ -30,8 +30,8 @@ class Towers(HTTPSpotProvider):
                         dx_call=source_spot["call"].upper(),
                         freq=likely_freq,
                         comment=source_spot["comment"],
-                        sig="WWTOTA",
-                        sig_refs=[SIGRef(id=source_spot["ref"], sig="WWTOTA")],
+                        sig="Towers",
+                        sig_refs=[SIGRef(id=source_spot["ref"], sig="Towers")],
                         time=datetime.strptime(response_json["updated"][:10] + source_spot["time"],
                                                "%Y-%m-%d%H:%M").timestamp())
 

spotproviders/ukpacketnet.py

@@ -14,7 +14,7 @@ class UKPacketNet(HTTPSpotProvider):
     SPOTS_URL = "https://nodes.ukpacketradio.network/api/nodedata"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("UK Packet Net", provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_spots(self, http_response):
         new_spots = []

spotproviders/websocket_spot_provider.py

@@ -13,8 +13,8 @@ from spotproviders.spot_provider import SpotProvider
 class WebsocketSpotProvider(SpotProvider):
     """Spot provider using websockets."""
 
-    def __init__(self, provider_config, url):
-        super().__init__(provider_config)
+    def __init__(self, name, provider_config, url):
+        super().__init__(name, provider_config)
         self._url = url
         self._ws = None
         self._thread = None

spotproviders/wota.py

@@ -21,11 +21,11 @@ class WOTA(HTTPSpotProvider):
     RSS_DATE_TIME_FORMAT = "%a, %d %b %Y %H:%M:%S %z"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("WOTA", provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_spots(self, http_response):
         new_spots = []
-        rss = cast(RSS, Parser.parse(http_response.content.decode()))
+        rss = cast(RSS, Parser.parse(http_response.content.decode("utf-8-sig")))
         # Iterate through source data
         for source_spot in rss.channel.items:
 

spotproviders/wwbota.py

@@ -12,7 +12,7 @@ class WWBOTA(SSESpotProvider):
     SPOTS_URL = "https://api.wwbota.net/spots/"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOTS_URL)
+        super().__init__("WWBOTA", provider_config, self.SPOTS_URL)
 
     def _sse_message_to_spot(self, message):
         source_spot = json.loads(message)

spotproviders/wwff.py

@@ -14,7 +14,7 @@ class WWFF(HTTPSpotProvider):
     SPOTS_URL = "https://spots.wwff.co/static/spots.json"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("WWFF", provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_spots(self, http_response):
         new_spots = []

spotproviders/xota.py

@@ -1,6 +1,4 @@
-import csv
 import json
-import logging
 from datetime import datetime
 
 import pytz
@@ -12,36 +10,25 @@ from spotproviders.websocket_spot_provider import WebsocketSpotProvider
 
 class XOTA(WebsocketSpotProvider):
     """Spot provider for servers based on the "xOTA" software at https://github.com/nischu/xOTA/
-    The provider typically doesn't give us a lat/lon or SIG explicitly, so our own config provides a SIG and a reference
-    to a local CSV file with location information. This functionality is implemented for TOTA events, of which there are
-    several - so a plain lookup of a "TOTA reference" doesn't make sense, it depends on which TOTA and hence which server
-    supplied the data, which is why the CSV location lookup is here and not in sig_utils."""
+    The provider typically doesn't give us a lat/lon or SIG explicitly, so our own config provides a SIG which we can
+    then use for lookups. This functionality is implemented for Toilets on the Air events, of which there are
+    several - so a plain lookup of a "TOTA reference" doesn't make sense, it depends on which TOTA, which is why we also
+    provide a sig_ref_prefix in our config. This is applied to the reference ID, so e.g. "T-01" at C3 might become
+    "C3 T-01". This allows us to provide location lookups for TOTA at several conferences."""
 
     LOCATION_DATA = {}
     SIG = None
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, provider_config["url"])
-        locations_csv = str(provider_config["locations-csv"]) if "locations-csv" in provider_config else None
+        name = provider_config["name"] if "name" in provider_config else "xOTA"
+        super().__init__(name, provider_config, provider_config["url"])
         self.SIG = str(provider_config["sig"]) if "sig" in provider_config else None
-
-        # Load location data
-        if locations_csv:
-            try:
-                f = open(locations_csv)
-                csv_data = f.read()
-                dr = csv.DictReader(csv_data.splitlines())
-                for row in dr:
-                    self.LOCATION_DATA[row["ref"]] = {"lat": row["lat"], "lon": row["lon"]}
-            except:
-                logging.exception("Could not look up location data for XOTA source.")
+        self._sig_ref_prefix = str(provider_config["sig-ref-prefix"]) if "sig-ref-prefix" in provider_config else ""
 
     def _ws_message_to_spot(self, b):
         string = b.decode("utf-8")
         source_spot = json.loads(string)
-        ref_id = source_spot["reference"]["title"]
-        lat = float(self.LOCATION_DATA[ref_id]["lat"]) if ref_id in self.LOCATION_DATA else None
-        lon = float(self.LOCATION_DATA[ref_id]["lon"]) if ref_id in self.LOCATION_DATA else None
+        ref_id = self._sig_ref_prefix + " " + source_spot["reference"]["title"]
         spot = Spot(source=self.name,
                     source_id=source_spot["id"],
                     dx_call=source_spot["stationCallSign"].upper(),
@@ -49,10 +36,7 @@ class XOTA(WebsocketSpotProvider):
                     mode=source_spot["mode"].upper(),
                     sig=self.SIG,
                     sig_refs=[
-                        SIGRef(id=ref_id, sig=self.SIG or "", url=source_spot["reference"]["website"], latitude=lat,
-                               longitude=lon)],
+                        SIGRef(id=ref_id, sig=self.SIG or "", url=source_spot["reference"]["website"])],
                     time=datetime.now(pytz.UTC).timestamp(),
-                    dx_latitude=lat,
-                    dx_longitude=lon,
                     qrt=source_spot["state"] != "active")
         return spot

spotproviders/zlota.py

@@ -15,7 +15,7 @@ class ZLOTA(HTTPSpotProvider):
     LIST_URL = "https://ontheair.nz/assets/assets.json"
 
     def __init__(self, provider_config):
-        super().__init__(provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
+        super().__init__("ZLOTA", provider_config, self.SPOTS_URL, self.POLL_INTERVAL_SEC)
 
     def _http_response_to_spots(self, http_response):
         new_spots = []

static/apidocs/openapi.yml

@@ -23,6 +23,7 @@ info:
     * Added `comment_names` to SIGs in the `/options`, to reflect how they might be referred to in spot comments where
       it differs from their `name`.
     * Added `propagation_mode` field to spots
+    * Added `sig_ref_data_providers` to status and removed `cleanup`
 
     ### 1.3
 
@@ -795,7 +796,7 @@ components:
         - HEMA
         - WCA
         - MOTA
-        - SiOTA
+        - SIOTA
         - ARLHS
         - ILLW
         - ZLOTA
@@ -1720,6 +1721,32 @@ components:
             is zero, the provider has never updated.
           example: 1759579508
 
+    SIGRefDataProviderStatus:
+      type: object
+      properties:
+        sig_name:
+          type: string
+          description: The name of the SIG.
+          example: WWFF
+        enabled:
+          type: boolean
+          description: Whether the provider is enabled or not.
+          example: true
+        status:
+          type: string
+          description: The status of the provider.
+          example: OK
+        last_updated:
+          type: number
+          description: >
+            The last time at which this provider received data, UTC seconds since UNIX epoch. If this
+            is zero, the provider has never updated.
+          example: 1759579508
+        reference_count:
+          type: number
+          description: The number of references fetched using this provider.
+          example: 1234
+
     SpotList:
       type: array
       items:
@@ -1787,17 +1814,6 @@ components:
           type: integer
           description: Number of alerts currently in the system.
           example: 123
-        "cleanup":
-          type: object
-          properties:
-            status:
-              type: string
-              description: The status of the cleanup thread
-              example: OK
-            last_ran:
-              type: number
-              description: The last time the cleanup operation ran, UTC seconds since UNIX epoch.
-              example: 1759579508
         "webserver":
           type: object
           properties:
@@ -1828,6 +1844,11 @@ components:
           description: An array of all the solar conditions providers.
           items:
             $ref: '#/components/schemas/SolarConditionsProviderStatus'
+        sig_ref_data_providers:
+          type: array
+          description: An array of all the SIG reference data providers.
+          items:
+            $ref: '#/components/schemas/SIGRefDataProviderStatus'
 
     Options:
       type: object

static/js/status.js

@@ -12,9 +12,6 @@ function loadStatus() {
         $("#web-server-last-api").text(moment.unix(jsonData["webserver"]["last_api_access"]).utc().fromNow());
         $("#web-server-last-page").text(moment.unix(jsonData["webserver"]["last_page_access"]).utc().fromNow());
 
-        $("#cleanup-status").text(jsonData["cleanup"]["status"]);
-        $("#cleanup-last-ran").text(moment.unix(jsonData["cleanup"]["last_ran"]).utc().fromNow());
-
         jsonData["spot_providers"].forEach(p => {
             $("#spot-providers-status-container").append(`
                 <div class="row row-cols-1 row-cols-md-4 g-4 mb-2">
@@ -42,6 +39,16 @@ function loadStatus() {
                     <div class="col">Last updated: ${(p["enabled"] && p["last_updated"] > 0) ? moment.unix(p["last_updated"]).utc().fromNow() : "N/A"}</div>
                 </div>`);
         });
+
+        jsonData["sig_ref_data_providers"].forEach(p => {
+            $("#sig-ref-data-providers-status-container").append(`
+                <div class="row row-cols-1 row-cols-md-4 g-4 mb-2">
+                    <div class="col"><strong>${p["sig_name"]}</strong></div>
+                    <div class="col">Status: ${p["status"]}</div>
+                    <div class="col">Last updated: ${(p["enabled"] && p["last_updated"] > 0) ? moment.unix(p["last_updated"]).utc().fromNow() : "N/A"}</div>
+                    <div class="col">References: ${(p["enabled"] && p["reference_count"] > 0) ? p["reference_count"] : "N/A"}</div>
+                </div>`);
+        });
     });
 }
 

templates/about.html

@@ -117,7 +117,6 @@
         <li>Some SIGs, such as Worked all Britain (WAB), don't have their own spotting site and can <em>only</em> be
             identified through comments on spots retrieved from other sources.
         </li>
-        <li>SIGs have well-defined names, whereas the server owner may name the sources as they see fit.</li>
     </ol>
     <p>Spothole's web interface exists not just for the end user, but also as a reference implementation for the API, so
         I have chosen to demonstrate both methods of filtering.</p>

templates/add_spot.html

@@ -76,7 +76,7 @@
 
 </div>
 
-<script src="/static/js/add-spot.js?v=1785082221"></script>
+<script src="/static/js/add-spot.js?v=1785434214"></script>
 <script>$(document).ready(function () {
     $("#nav-link-add-spot").addClass("active");
 }); <!-- highlight active page in nav --></script>

templates/alerts.html

@@ -75,7 +75,7 @@
 
 </div>
 
-<script src="/static/js/alerts.js?v=1785082221"></script>
+<script src="/static/js/alerts.js?v=1785434213"></script>
 <script>$(document).ready(function () {
     $("#nav-link-alerts").addClass("active");
 }); <!-- highlight active page in nav --></script>

templates/bands.html

@@ -75,8 +75,8 @@
 <script>
     let spotProvidersEnabledByDefault = {% raw json_encode(web_ui_options["spot-providers-enabled-by-default"]) %};
 </script>
-<script src="/static/js/spotsbandsandmap.js?v=1785082221"></script>
-<script src="/static/js/bands.js?v=1785082221"></script>
+<script src="/static/js/spotsbandsandmap.js?v=1785434214"></script>
+<script src="/static/js/bands.js?v=1785434214"></script>
 <script>$(document).ready(function () {
     $("#nav-link-bands").addClass("active");
 }); <!-- highlight active page in nav --></script>

templates/base.html

@@ -1,6 +1,6 @@
 {% extends "skeleton.html" %}
 {% block head_extra %}
-<link rel="stylesheet" href="/static/css/style.css?v=1785082221" type="text/css">
+<link rel="stylesheet" href="/static/css/style.css?v=1785434213" type="text/css">
 <link href="/static/vendor/css/bootstrap-5.3.8.min.css" rel="stylesheet">
 <link href="/static/vendor/css/fontawesome-6.7.2.min.css" rel="stylesheet">
 <link href="/static/vendor/css/solid-6.7.2.min.css" rel="stylesheet">
@@ -10,10 +10,10 @@
 <script src="/static/vendor/js/bootstrap-5.3.8.bundle.min.js"></script>
 <script src="/static/vendor/js/tinycolor2-1.6.0.min.js"></script>
 
-<script src="/static/js/utils.js?v=1785082221"></script>
-<script src="/static/js/ui-ham.js?v=1785082221"></script>
-<script src="/static/js/geo.js?v=1785082221"></script>
-<script src="/static/js/common.js?v=1785082221"></script>
+<script src="/static/js/utils.js?v=1785434213"></script>
+<script src="/static/js/ui-ham.js?v=1785434213"></script>
+<script src="/static/js/geo.js?v=1785434213"></script>
+<script src="/static/js/common.js?v=1785434213"></script>
 {% end %}
 {% block body %}
 <div class="container">

templates/conditions.html

@@ -284,7 +284,7 @@
 </div>
 
 <script src="/static/vendor/js/chart-4.4.9.umd.min.js"></script>
-<script src="/static/js/conditions.js?v=1785082221"></script>
+<script src="/static/js/conditions.js?v=1785434213"></script>
 <script>$(document).ready(function () {
     $("#nav-link-conditions").addClass("active");
 }); <!-- highlight active page in nav --></script>

templates/map.html

@@ -108,8 +108,8 @@
 <script>
     let spotProvidersEnabledByDefault = {% raw json_encode(web_ui_options["spot-providers-enabled-by-default"]) %};
 </script>
-<script src="/static/js/spotsbandsandmap.js?v=1785082221"></script>
-<script src="/static/js/map.js?v=1785082221"></script>
+<script src="/static/js/spotsbandsandmap.js?v=1785434213"></script>
+<script src="/static/js/map.js?v=1785434213"></script>
 <script>$(document).ready(function () {
     $("#nav-link-map").addClass("active");
 }); <!-- highlight active page in nav --></script>

templates/spots.html

@@ -116,8 +116,8 @@
 <script>
     let spotProvidersEnabledByDefault = {% raw json_encode(web_ui_options["spot-providers-enabled-by-default"]) %};
 </script>
-<script src="/static/js/spotsbandsandmap.js?v=1785082221"></script>
-<script src="/static/js/spots.js?v=1785082221"></script>
+<script src="/static/js/spotsbandsandmap.js?v=1785434213"></script>
+<script src="/static/js/spots.js?v=1785434213"></script>
 <script>$(document).ready(function () {
     $("#nav-link-spots").addClass("active");
 }); <!-- highlight active page in nav --></script>

templates/status.html

@@ -24,11 +24,6 @@
             <div class="col">Last API call: <span id="web-server-last-api"></span></div>
             <div class="col">Last page req: <span id="web-server-last-page"></span></div>
         </div>
-        <div class="row row-cols-1 row-cols-md-4 g-4 mb-2">
-            <div class="col"><strong>Cleanup Service</strong></div>
-            <div class="col">Status: <span id="cleanup-status"></span></div>
-            <div class="col">Last ran: <span id="cleanup-last-ran"></span></div>
-        </div>
     </div>
 </div>
 
@@ -59,7 +54,16 @@
     </div>
 </div>
 
-<script src="/static/js/status.js?v=1785082221"></script>
+<div class="card mt-3">
+    <div class="card-header">
+        SIG Reference Data Providers
+    </div>
+    <div class="card-body" id="sig-ref-data-providers-status-container">
+
+    </div>
+</div>
+
+<script src="/static/js/status.js?v=1785434213"></script>
 <script>
     $(document).ready(function () {
         $("#nav-link-status").addClass("active");