failing tests

ee/clickhouse/models/test/test_property.py

@@ -1515,8 +1515,16 @@ def test_events(db, team) -> List[UUID]:
             event="$pageview",
             team=team,
             distinct_id="whatever",
+            # new line character shouldn't be matched by a single regex dot
             properties={"email": "test@post\nhog.com"},
         ),
+        _create_event(
+            event="$pageview",
+            team=team,
+            distinct_id="whatever",
+            # not a new line character - instead a single character - should match
+            properties={"email": "[email protected]"},
+        ),
     ]
 
 
@@ -1756,7 +1764,8 @@ def clean_up_materialised_columns():
         id="can match before date only values",
     ),
     # Regression test, we were previously matching on newline characters
-    pytest.param(Property(key="email", value=r"[email protected]", operator="regex"), []),
+    # this should match one of two possibles (how are you supposed to figure out the expected index 🙈)
+    pytest.param(Property(key="email", value=r"[email protected]", operator="regex"), [28]),
 ]
 
 

posthog/api/test/__snapshots__/test_query.ambr

@@ -100,32 +100,6 @@
                     allow_experimental_object_type=1
   '
 ---
-# name: TestQuery.test_event_property_regex_is_patched_for_dotall_setting
-  '
-  /* user_id:0 request:_snapshot_ */
-  SELECT replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(events.properties, 'example_value'), ''), 'null'), '^"|"$', '') AS example_value
-  FROM events
-  WHERE and(equals(events.team_id, 2), match(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(events.properties, 'example_value'), ''), 'null'), '^"|"$', ''), '(?-s)a.b'), equals(events.event, 'demonstrate dot all'), less(toTimeZone(events.timestamp, 'UTC'), toDateTime64('2020-01-10 12:14:05.000000', 6, 'UTC')), greater(toTimeZone(events.timestamp, 'UTC'), toDateTime64('2020-01-09 12:14:00.000000', 6, 'UTC')))
-  ORDER BY replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(events.properties, 'example_value'), ''), 'null'), '^"|"$', '') ASC
-  LIMIT 101
-  OFFSET 0 SETTINGS readonly=2,
-                    max_execution_time=60,
-                    allow_experimental_object_type=1
-  '
----
-# name: TestQuery.test_event_property_regex_is_patched_for_dotall_setting_materialized
-  '
-  /* user_id:0 request:_snapshot_ */
-  SELECT nullIf(nullIf(events.mat_example_value, ''), 'null') AS example_value
-  FROM events
-  WHERE and(equals(events.team_id, 2), match(nullIf(nullIf(events.mat_example_value, ''), 'null'), '(?-s)a.b'), equals(events.event, 'demonstrate dot all'), less(toTimeZone(events.timestamp, 'UTC'), toDateTime64('2020-01-10 12:14:05.000000', 6, 'UTC')), greater(toTimeZone(events.timestamp, 'UTC'), toDateTime64('2020-01-09 12:14:00.000000', 6, 'UTC')))
-  ORDER BY nullIf(nullIf(events.mat_example_value, ''), 'null') ASC
-  LIMIT 101
-  OFFSET 0 SETTINGS readonly=2,
-                    max_execution_time=60,
-                    allow_experimental_object_type=1
-  '
----
 # name: TestQuery.test_events_query_all_time_date
   '
   /* user_id:0 request:_snapshot_ */

posthog/api/test/test_query.py

@@ -331,8 +331,8 @@ def test_event_property_regex_is_patched_for_dotall_setting(self):
                 properties=[{"key": "example_value", "value": "a.b", "operator": "regex", "type": "event"}],
             )
             response = self.client.post(f"/api/projects/{self.team.id}/query/", {"query": query.dict()}).json()
-            # assert [x[0] for x in response["results"]] == ['aab', 'abb']
-            assert [x[0] for x in response["results"]] == []
+            assert "(?-s)" in response["hogql"]
+            assert [x[0] for x in response["results"]] == ["aab", "abb"]
 
     @also_test_with_materialized_columns(event_properties=["key"], person_properties=["email"])
     @snapshot_clickhouse_queries

posthog/hogql/printer.py

@@ -542,16 +542,16 @@ def visit_compare_operation(self, node: ast.CompareOperation):
         elif node.op == ast.CompareOperationOp.GlobalNotIn:
             op = f"globalNotIn({left}, {right})"
         elif node.op == ast.CompareOperationOp.Regex:
-            op = f"match({left}, {right})"
+            op = f"match({left}, concat('(?-s)', {right}))"
             value_if_both_sides_are_null = True
         elif node.op == ast.CompareOperationOp.NotRegex:
-            op = f"not(match({left}, {right}))"
+            op = f"not(match({left}, concat('(?-s)', {right})))"
             value_if_one_side_is_null = True
         elif node.op == ast.CompareOperationOp.IRegex:
-            op = f"match({left}, concat('(?i)', {right}))"
+            op = f"match({left}, concat('(?i-s)', {right}))"
             value_if_both_sides_are_null = True
         elif node.op == ast.CompareOperationOp.NotIRegex:
-            op = f"not(match({left}, concat('(?i)', {right})))"
+            op = f"not(match({left}, concat('(?i-s)', {right})))"
             value_if_one_side_is_null = True
         elif node.op == ast.CompareOperationOp.Gt:
             op = f"greater({left}, {right})"

posthog/hogql/test/test_property.py

@@ -145,7 +145,7 @@ def test_property_to_expr_event(self):
         )
         self.assertEqual(
             self._property_to_expr({"type": "event", "key": "a", "value": ".*", "operator": "not_regex"}),
-            self._parse_expr("not(match(properties.a, '.*'))"),
+            self._parse_expr("not(match(properties.a, '(?-s).*'))"),
         )
         self.assertEqual(
             self._property_to_expr({"type": "event", "key": "a", "value": [], "operator": "exact"}),