Fix Normalization Rules (#894)

* Fix cross agent tests to run from anywhere

* Cover failures in rules engine with testing

Co-authored-by: Lalleh Rafeei <[email protected]>
Co-authored-by: Hannah Stepanek <[email protected]>
Co-authored-by: Uma Annamalai <[email protected]>

* Patch metrics not being properly ignored

* Patch normalization rule init default arguments

* Clean up to match other fixture setups

---------

Co-authored-by: Lalleh Rafeei <[email protected]>
Co-authored-by: Hannah Stepanek <[email protected]>
Co-authored-by: Uma Annamalai <[email protected]>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

newrelic/core/rules_engine.py

@@ -22,6 +22,27 @@
 
 
 class NormalizationRule(_NormalizationRule):
+    def __new__(
+        cls,
+        match_expression="",
+        replacement="",
+        ignore=False,
+        eval_order=0,
+        terminate_chain=False,
+        each_segment=False,
+        replace_all=False,
+    ):
+        return _NormalizationRule.__new__(
+            cls,
+            match_expression=match_expression,
+            replacement=replacement,
+            ignore=ignore,
+            eval_order=eval_order,
+            terminate_chain=terminate_chain,
+            each_segment=each_segment,
+            replace_all=replace_all,
+        )
+
     def __init__(self, *args, **kwargs):
         self.match_expression_re = re.compile(self.match_expression, re.IGNORECASE)
 

newrelic/core/stats_engine.py

@@ -1129,7 +1129,11 @@ def metric_data(self, normalizer=None):
 
         if normalizer is not None:
             for key, value in six.iteritems(self.__stats_table):
-                key = (normalizer(key[0])[0], key[1])
+                normalized_name, ignored = normalizer(key[0])
+                if ignored:
+                    continue
+
+                key = (normalized_name, key[1])
                 stats = normalized_stats.get(key)
                 if stats is None:
                     normalized_stats[key] = copy.copy(value)

tests/cross_agent/test_agent_attributes.py

@@ -40,7 +40,8 @@ def _default_settings():
         'browser_monitoring.attributes.exclude': [],
     }
 
-FIXTURE = os.path.join(os.curdir, 'fixtures', 'attribute_configuration.json')
+CURRENT_DIR = os.path.dirname(os.path.realpath(__file__))
+FIXTURE = os.path.join(CURRENT_DIR, 'fixtures', 'attribute_configuration.json')
 
 def _load_tests():
     with open(FIXTURE, 'r') as fh:

tests/cross_agent/test_datstore_instance.py

@@ -23,7 +23,8 @@
 from newrelic.core.database_node import DatabaseNode
 from newrelic.core.stats_engine import StatsEngine
 
-FIXTURE = os.path.join(os.curdir,
+CURRENT_DIR = os.path.dirname(os.path.realpath(__file__))
+FIXTURE = os.path.join(CURRENT_DIR,
         'fixtures', 'datastores', 'datastore_instances.json')
 
 _parameters_list = ['name', 'system_hostname', 'db_hostname',

tests/cross_agent/test_docker.py

@@ -19,7 +19,8 @@
 
 import newrelic.common.utilization as u
 
-DOCKER_FIXTURE = os.path.join(os.curdir, 'fixtures', 'docker_container_id')
+CURRENT_DIR = os.path.dirname(os.path.realpath(__file__))
+DOCKER_FIXTURE = os.path.join(CURRENT_DIR, 'fixtures', 'docker_container_id')
 
 
 def _load_docker_test_attributes():

tests/cross_agent/test_labels_and_rollups.py

@@ -21,7 +21,8 @@
 
 from testing_support.fixtures import override_application_settings
 
-FIXTURE = os.path.join(os.curdir, 'fixtures', 'labels.json')
+CURRENT_DIR = os.path.dirname(os.path.realpath(__file__))
+FIXTURE = os.path.join(CURRENT_DIR, 'fixtures', 'labels.json')
 
 def _load_tests():
     with open(FIXTURE, 'r') as fh:

tests/cross_agent/test_rules.py

@@ -16,23 +16,23 @@
 import os
 import pytest
 
-from newrelic.core.rules_engine import RulesEngine, NormalizationRule
+from newrelic.api.application import application_instance
+from newrelic.api.background_task import background_task
+from newrelic.api.transaction import record_custom_metric
+from newrelic.core.rules_engine import RulesEngine
+
+from testing_support.validators.validate_metric_payload import validate_metric_payload
 
 CURRENT_DIR = os.path.dirname(os.path.realpath(__file__))
 FIXTURE = os.path.normpath(os.path.join(
         CURRENT_DIR, 'fixtures', 'rules.json'))
 
+
 def _load_tests():
     with open(FIXTURE, 'r') as fh:
         js = fh.read()
     return json.loads(js)
 
-def _prepare_rules(test_rules):
-    # ensure all keys are present, if not present set to an empty string
-    for rule in test_rules:
-        for key in NormalizationRule._fields:
-            rule[key] = rule.get(key, '')
-    return test_rules
 
 def _make_case_insensitive(rules):
     # lowercase each rule
@@ -42,14 +42,14 @@ def _make_case_insensitive(rules):
             rule['replacement'] = rule['replacement'].lower()
     return rules
 
+
 @pytest.mark.parametrize('test_group', _load_tests())
 def test_rules_engine(test_group):
 
     # FIXME: The test fixture assumes that matching is case insensitive when it
     # is not. To avoid errors, just lowercase all rules, inputs, and expected
     # values.
-    insense_rules = _make_case_insensitive(test_group['rules'])
-    test_rules = _prepare_rules(insense_rules)
+    test_rules = _make_case_insensitive(test_group['rules'])
     rules_engine = RulesEngine(test_rules)
 
     for test in test_group['tests']:
@@ -66,3 +66,46 @@ def test_rules_engine(test_group):
             assert expected == ''
         else:
             assert result == expected
+
+
+@pytest.mark.parametrize('test_group', _load_tests())
+def test_rules_engine_metric_harvest(test_group):
+    # FIXME: The test fixture assumes that matching is case insensitive when it
+    # is not. To avoid errors, just lowercase all rules, inputs, and expected
+    # values.
+    test_rules = _make_case_insensitive(test_group['rules'])
+    rules_engine = RulesEngine(test_rules)
+
+    # Set rules engine on core application
+    api_application = application_instance(activate=False)
+    api_name = api_application.name
+    core_application = api_application._agent.application(api_name)
+    old_rules = core_application._rules_engine["metric"]  # save previoius rules
+    core_application._rules_engine["metric"] = rules_engine
+
+    def send_metrics():
+        # Send all metrics in this test batch in one transaction, then harvest so the normalizer is run.
+        @background_task(name="send_metrics")
+        def _test():
+            for test in test_group['tests']:
+                # lowercase each value
+                input_str = test['input'].lower()
+                record_custom_metric(input_str, {"count": 1})
+        _test()
+        core_application.harvest()
+
+    try:
+        # Create a map of all result metrics to validate after harvest
+        test_metrics = []
+        for test in test_group['tests']:
+            expected = (test['expected'] or '').lower()
+            if expected == '':  # Ignored
+                test_metrics.append((expected, None))
+            else:
+                test_metrics.append((expected, 1))
+
+        # Harvest and validate resulting payload
+        validate_metric_payload(metrics=test_metrics)(send_metrics)()
+    finally:
+        # Replace original rules engine
+        core_application._rules_engine["metric"] = old_rules

tests/cross_agent/test_rum_client_config.py

@@ -26,10 +26,11 @@
 )
 from newrelic.api.wsgi_application import wsgi_application
 
+CURRENT_DIR = os.path.dirname(os.path.realpath(__file__))
+FIXTURE = os.path.join(CURRENT_DIR, "fixtures", "rum_client_config.json")
 
 def _load_tests():
-    fixture = os.path.join(os.curdir, "fixtures", "rum_client_config.json")
-    with open(fixture, "r") as fh:
+    with open(FIXTURE, "r") as fh:
         js = fh.read()
     return json.loads(js)