refactor!: make RequestQueueClient.is_finished abstract

docs/upgrading/upgrading_to_v2.md

@@ -16,3 +16,7 @@ We drop support for Python 3.10, which reached its end of life in October 2026.
 ## `StatisticsState.crawler_runtime` is read-only
 
 `StatisticsState.crawler_runtime` is computed from the crawler's start and finish timestamps. In v1, assigning to it only emitted a `DeprecationWarning` and the value was ignored. In v2, the property has no setter, so an assignment raises `AttributeError`. Remove any such assignments.
+
+## `RequestQueueClient.is_finished` is abstract
+
+`RequestQueueClient.is_finished` is now an abstract method. In v1, it fell back to `is_empty`, which ignores requests still being processed and could report the queue as finished too early. If you have a custom request queue client, implement `is_finished` so that it returns `True` only when the queue is empty and no requests are in progress. The built-in clients already implement it.

src/crawlee/storage_clients/_base/_request_queue_client.py

@@ -125,20 +125,12 @@ async def is_empty(self) -> bool:
             True if the request queue is empty, False otherwise.
         """
 
+    @abstractmethod
     async def is_finished(self) -> bool:
         """Check if the request queue is finished.
 
         A finished queue is empty and has no requests currently being processed.
 
-        Warning:
-            This default only checks `is_empty`, which reports whether requests are available to fetch, not
-            whether requests are still being processed. It can therefore return `True` while requests are in
-            progress. Subclasses that track in-progress requests should override this method; all built-in
-            clients already do.
-
         Returns:
             True if the request queue is finished, False otherwise.
         """
-        # TODO: Make this method abstract.
-        # https://github.com/apify/crawlee-python/issues/1985
-        return await self.is_empty()