Merge branch 'wip-martin-review-cleanup' into 'master'

review notes and some cleanup

See merge request webgroup/fuzzycat!7

2 files changed, 37 insertions, 9 deletions

diff --git a/tests/test_matching.py b/tests/test_matching.py
index 7ab7b11..90d1fee 100644
--- a/tests/test_matching.py
+++ b/tests/test_matching.py
@@ -9,7 +9,8 @@ from fatcat_openapi_client import ReleaseEntity
 from fuzzycat.entities import entity_from_dict
 from fuzzycat.matching import anything_to_entity, match_release_fuzzy
 
-warnings.filterwarnings("ignore") # InsecureRequestWarning: Unverified HTTPS request is being made to host ...
+warnings.filterwarnings(
+    "ignore")  # InsecureRequestWarning: Unverified HTTPS request is being made to host ...
 
 from fuzzycat.matching import anything_to_entity, match_release_fuzzy
 from fuzzycat.config import settings
@@ -28,6 +29,7 @@ FATCAT_SEARCH_URL = settings.get("FATCAT_SEARCH_URL", "https://search.fatcat.wik
 def is_not_reachable(url, timeout=3):
     return not is_reachable(url)
 
+
 def is_reachable(url, timeout=3):
     """
     Return true, if URL is reachable and returns HTTP 200.
@@ -37,14 +39,21 @@ def is_reachable(url, timeout=3):
     except Exception:
         return False
 
+
 @pytest.fixture
 def es_client():
     return elasticsearch.Elasticsearch([FATCAT_SEARCH_URL])
 
 
-@pytest.mark.skipif(is_not_reachable(FATCAT_SEARCH_URL),
-                    reason="{} not reachable, use e.g. FUZZYCAT_FATCAT_SEARCH_URL=localhost:9200 to override".format(FATCAT_SEARCH_URL))
+@pytest.mark.skipif(
+    is_not_reachable(FATCAT_SEARCH_URL),
+    reason="{} not reachable, use e.g. FUZZYCAT_FATCAT_SEARCH_URL=localhost:9200 to override".
+    format(FATCAT_SEARCH_URL))
 def test_match_release_fuzzy(es_client, caplog):
+    """
+    This test is tied to the current index contents, so if that changes, this
+    test may fail as well.
+    """
     cases = (
         ("wtv64ahbdzgwnan7rllwr3nurm", 1),
         ("eqcgtpav3na5jh56o5vjsvb4ei", 1),
@@ -63,17 +72,35 @@ def test_match_release_fuzzy(es_client, caplog):
             "ext_ids": {}
         }, 5),
         ({
-            "title": "The Future of Digital Scholarship",
-            "contribs": [{
-                "raw_name": "Costantino Thanos"
-            }],
+            "title": "unlikelytitle",
+            "ext_ids": {}
+        }, 0),
+        ({
+            "title": "Imminent dystopia",
+            "ext_ids": {}
+        }, 2),
+        ({
+            "title": "",
+            "contribs": [{"raw_name": "Aristoteles"}],
             "ext_ids": {}
         }, 5),
+        # ({
+        #     "title": "Letter",
+        #     "contribs": [{"raw_name": "Claudel"}],
+        #     "ext_ids": {}
+        # }, 1),
+        # ({
+        #     "title": "The Future of Digital Scholarship",
+        #     "contribs": [{
+        #         "raw_name": "Costantino Thanos"
+        #     }],
+        #     "ext_ids": {}
+        # }, 5),
     )
     for i, (doc, count) in enumerate(cases):
         entity = entity_from_dict(doc, ReleaseEntity)
         result = match_release_fuzzy(entity, es=es_client)
         with caplog.at_level(logging.INFO):
-            logging.info("[{}] given {}, found {}, {}".format(i, entity.title, len(result),
+            logging.info("[{}] given title '{}', found {}, {}".format(i, entity.title, len(result),
                                                               [v.title for v in result]))
-        assert len(result) == count
+        assert len(result) == count, doc
diff --git a/tests/test_utils.py b/tests/test_utils.py
index 21b85a4..957203f 100644
--- a/tests/test_utils.py
+++ b/tests/test_utils.py
@@ -124,6 +124,7 @@ def test_es_compat_hits_total():
     for r, expected in cases:
         assert es_compat_hits_total(r) == expected
 
+
 def test_clean_doi():
     assert clean_doi(None) == None
     assert clean_doi("blah") == None