Internal: support iteration over LRUCache correctly

Right now LRUCache is iterable due to its __getitem__ method which Python uses
with numerical indexes. But due to the current `get()` default value of `None`
it will iterate for ever always returning `None`. For now this commit fixes
this by making iteration over a LRUCache iterate the underlying dict instead.

tests/test_utils.py

@@ -80,6 +80,15 @@ def test_lru_cache_unlimited_explicit():
     assert len(cache.lru) == 100
 
 
+def test_lru_cache_iteration_works():
+    cache = LRUCache()
+    count = 0
+    for _ in cache:
+        assert False, 'there should be no elements in the cache'
+
+    assert count == 0
+
+
 def test_catch_warning():
     class MyWarning(Warning):
         pass

tinydb/utils.py

@@ -43,6 +43,9 @@ def __delitem__(self, key):
     def __getitem__(self, key):
         return self.get(key)
 
+    def __iter__(self):
+        return iter(self.__cache)
+
     def get(self, key, default=None):
         value = self.__cache.get(key)
         if value: