Improve use of URL cache #118

core/url_data_cache.py

@@ -3,21 +3,23 @@ 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),
+from core.data_store import CACHE_DIR
+
+
+class URLDataCache(CachedSession):
+    """Cache for "semi-static" data retrieved from a URL. This is an extra layer 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. It should only be used for data that isn't
+    expected to change on a sub-daily basis, so never for spots & alerts. It also implements a lock to allow it to be
+    used across multiple threads, though note that URL lookups will block each other this way, so it is still better to
+    create one of these objects per thread if possible."""
+
+    _lock = threading.Lock()
+
+    def __init__(self, name):
+        super().__init__(CACHE_DIR + "urls/" + name, 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()
+        with self._lock:
+            return super().get(*args, **kwargs)