Fix HFS limit and cursor handling

Obelisk HFS still returns a cursor to the rest of data,
even after the limit is reached.

This peculiar behaviour meant that the `limit` was never actually respected.

The workaround implemented here,
by client side limiting len(result_set),
was suggested by Obelisk HFS developers.

Author: Epse

src/obelisk/asynchronous/consumer.py

@@ -163,6 +163,16 @@ async def query(self, datasets: List[str], metrics:Optional[List[str]] = None,
             result_set.extend(result.items)
             cursor = result.cursor
 
+            if limit and len(result_set) >= limit:
+                """On Obelisk HFS, limit is actually page size,
+                so continuing to read the cursor will result in a larger than desired
+                set of results.
+
+                On the other hand, if the limit is very large,
+                we may need to iterate before we reach the desired limit after all.
+                """
+                break
+
         return result_set