urlstore: add method get_all_counts()

courlan/urlstore.py

@@ -386,6 +386,10 @@ def get_crawl_delay(self, website: str, default: float = 5) -> float:
 
     # GENERAL INFO
 
+    def get_all_counts(self) -> List[int]:
+        "Return all download counts for the hosts in store."
+        return [self.urldict[d].count for d in self.urldict]
+
     def total_url_number(self) -> int:
         "Find number of all URLs in store."
         return sum(len(self.urldict[d].tuples) for d in self.urldict)

tests/urlstore_tests.py

@@ -122,6 +122,7 @@ def test_urlstore():
         urls
     )
     assert my_urls.total_url_number() == len(urls)
+    assert my_urls.get_all_counts() == [0, 0]
 
     if my_urls.compressed is False:
         assert sum(len(v.tuples) for _, v in my_urls.urldict.items()) == len(urls)
@@ -170,6 +171,7 @@ def test_urlstore():
     assert my_urls.urldict[example_domain].count == 2
     assert timestamp != my_urls.urldict[example_domain].timestamp
     assert url2 not in set(my_urls.find_unvisited_urls(example_domain))
+    assert my_urls.get_all_counts() == [2, 0, 0]
 
     # as_visited=False
     timestamp = my_urls.urldict[example_domain].timestamp