Fixes Scrapinghub jobs.list Method Called with Incorrect Filter Param…

…eter (#446)

* Fixes Scrapinghub jobs.list Method Called with Incorrect Filter Parameter

* Return empty list on empty cases

* Add _get_tags_to_filter tests

spidermon/contrib/scrapy/monitors/monitors.py

@@ -569,7 +569,7 @@ def _get_jobs(self, states, number_of_jobs):
                 start=start,
                 state=states,
                 count=count,
-                filters=dict(has_tag=tags) if tags else None,
+                has_tag=tags or None,
             )
             total_jobs.extend(current_jobs)
 
@@ -584,19 +584,19 @@ def _get_jobs(self, states, number_of_jobs):
 
     def _get_tags_to_filter(self):
         """
-        Return the intersect of the desired tags to filter and
+        Return a list of tags with the intersection of the desired tags to filter and
         the ones from the current job.
         """
         desired_tags = self.crawler.settings.getlist(SPIDERMON_JOBS_COMPARISON_TAGS)
         if not desired_tags:
-            return {}
+            return []
 
         current_tags = json.loads(os.environ.get("SHUB_JOB_DATA", "{}")).get("tags")
         if not current_tags:
-            return {}
+            return []
 
         tags_to_filter = set(desired_tags) & set(current_tags)
-        return sorted(tags_to_filter)
+        return list(sorted(tags_to_filter))
 
     def get_threshold(self):
         number_of_jobs = self.crawler.settings.getint(SPIDERMON_JOBS_COMPARISON)

tests/contrib/scrapy/monitors/test_jobs_comparison_monitor.py

@@ -90,6 +90,26 @@ def runTest():
         pass
 
 
+def test_jobs_comparison_monitor_get_tags_to_filter(monkeypatch):
+    mock_data = Mock()
+
+    monitor = TestZyteJobsComparisonMonitor()
+    monitor.data = mock_data
+
+    # Empty SPIDERMON_JOBS_COMPARISON_TAGS
+    mock_data.crawler.settings.getlist.return_value = None
+    assert monitor._get_tags_to_filter() == []
+
+    # Empty SHUB_JOB_DATA.tags
+    mock_data.crawler.settings.getlist.return_value = ["tag1", "tag2"]
+    assert monitor._get_tags_to_filter() == []
+
+    # Sorted intersection
+    mock_data.crawler.settings.getlist.return_value = ["tag2", "tag1", "tag3"]
+    monkeypatch.setenv("SHUB_JOB_DATA", '{"tags": ["tag1", "tag2"]}')
+    assert monitor._get_tags_to_filter() == ["tag1", "tag2"]
+
+
 def test_jobs_comparison_monitor_get_jobs():
     mock_client = Mock()
     with patch(
@@ -197,7 +217,7 @@ def test_arguments_passed_to_zyte_client(
             state=list(states),
             # Count goes from pending number of jobs up to 1000
             count=min(number_of_jobs - n * 1000, 1000),
-            filters={"has_tag": list(tags)},
+            has_tag=list(tags),
         )
         # One call to api every 1000 expected jobs
         for n in range(0, math.ceil(number_of_jobs / 1000))