fix: issue with is identified boolean clickhouse (#26604)

Co-authored-by: github-actions <41898282+github-actions[bot]@users.noreply.github.com>

posthog/hogql/database/schema/util/test/test_person_where_clause_extractor.py

@@ -194,6 +194,6 @@ def test_boolean(self):
         )
         actual = self.print_query("SELECT * FROM events WHERE person.properties.person_boolean = false")
         assert (
-            f"ifNull(equals(transform(toString(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(person.properties"
+            f"ifNull(equals(toBool(transform(toString(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(person.properties"
             in actual
         )

posthog/hogql/functions/mapping.py

@@ -432,6 +432,7 @@ def compare_types(arg_types: list[ConstantType], sig_arg_types: tuple[ConstantTy
             ((DateTimeType(), StringType()), StringType()),
         ],
     ),
+    "toBool": HogQLFunctionMeta("toBool", 1, 1),
     "toJSONString": HogQLFunctionMeta("toJSONString", 1, 1),
     "parseDateTime": HogQLFunctionMeta("parseDateTimeOrNull", 2, 3, tz_aware=True),
     "parseDateTimeBestEffort": HogQLFunctionMeta("parseDateTime64BestEffortOrNull", 1, 2, tz_aware=True),

posthog/hogql/test/test_printer.py

@@ -1603,18 +1603,18 @@ def test_field_nullable_boolean(self):
         )
         assert generated_sql_statements1 == (
             f"SELECT "
-            "ifNull(equals(transform(toString(nullIf(nullIf(events.mat_is_boolean, ''), 'null')), %(hogql_val_0)s, %(hogql_val_1)s, NULL), 1), 0), "
-            "ifNull(equals(transform(toString(nullIf(nullIf(events.mat_is_boolean, ''), 'null')), %(hogql_val_2)s, %(hogql_val_3)s, NULL), 0), 0), "
-            "isNull(transform(toString(nullIf(nullIf(events.mat_is_boolean, ''), 'null')), %(hogql_val_4)s, %(hogql_val_5)s, NULL)) "
+            "ifNull(equals(toBool(transform(toString(nullIf(nullIf(events.mat_is_boolean, ''), 'null')), %(hogql_val_0)s, %(hogql_val_1)s, NULL)), 1), 0), "
+            "ifNull(equals(toBool(transform(toString(nullIf(nullIf(events.mat_is_boolean, ''), 'null')), %(hogql_val_2)s, %(hogql_val_3)s, NULL)), 0), 0), "
+            "isNull(toBool(transform(toString(nullIf(nullIf(events.mat_is_boolean, ''), 'null')), %(hogql_val_4)s, %(hogql_val_5)s, NULL))) "
             f"FROM events WHERE equals(events.team_id, {self.team.pk}) LIMIT {MAX_SELECT_RETURNED_ROWS}"
         )
         assert context.values == {
             "hogql_val_0": ["true", "false"],
-            "hogql_val_1": [True, False],
+            "hogql_val_1": [1, 0],
             "hogql_val_2": ["true", "false"],
-            "hogql_val_3": [True, False],
+            "hogql_val_3": [1, 0],
             "hogql_val_4": ["true", "false"],
-            "hogql_val_5": [True, False],
+            "hogql_val_5": [1, 0],
         }
 
     def test_field_nullable_like(self):

posthog/hogql/transforms/property_types.py

@@ -221,12 +221,17 @@ def _field_type_to_property_call(self, node: ast.Field, field_type: str):
             return ast.Call(name="toFloat", args=[node])
         if field_type == "Boolean":
             return ast.Call(
-                name="transform",
+                name="toBool",
                 args=[
-                    ast.Call(name="toString", args=[node]),
-                    ast.Constant(value=["true", "false"]),
-                    ast.Constant(value=[True, False]),
-                    ast.Constant(value=None),
+                    ast.Call(
+                        name="transform",
+                        args=[
+                            ast.Call(name="toString", args=[node]),
+                            ast.Constant(value=["true", "false"]),
+                            ast.Constant(value=[1, 0]),
+                            ast.Constant(value=None),
+                        ],
+                    )
                 ],
             )
         return node

posthog/hogql/transforms/test/__snapshots__/test_property_types.ambr

@@ -2,7 +2,7 @@
 # name: TestPropertyTypes.test_data_warehouse_person_property_types
   '''
 
-  SELECT persons__extended_properties.string_prop AS string_prop, persons__extended_properties.int_prop AS int_prop, transform(toString(persons__extended_properties.bool_prop), %(hogql_val_8)s, %(hogql_val_9)s, NULL) AS bool_prop 
+  SELECT persons__extended_properties.string_prop AS string_prop, persons__extended_properties.int_prop AS int_prop, toBool(transform(toString(persons__extended_properties.bool_prop), %(hogql_val_8)s, %(hogql_val_9)s, NULL)) AS bool_prop 
   FROM (
   SELECT argMax(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(person.properties, %(hogql_val_0)s), ''), 'null'), '^"|"$', ''), person.version) AS persons___properties___email, person.id AS id 
   FROM person 
@@ -19,7 +19,7 @@
 # name: TestPropertyTypes.test_group_boolean_property_types
   '''
 
-  SELECT ifNull(equals(transform(toString(events__group_0.properties___group_boolean), %(hogql_val_2)s, %(hogql_val_3)s, NULL), 1), 0), ifNull(equals(transform(toString(events__group_0.properties___group_boolean), %(hogql_val_4)s, %(hogql_val_5)s, NULL), 0), 0), isNull(transform(toString(events__group_0.properties___group_boolean), %(hogql_val_6)s, %(hogql_val_7)s, NULL)) 
+  SELECT ifNull(equals(toBool(transform(toString(events__group_0.properties___group_boolean), %(hogql_val_2)s, %(hogql_val_3)s, NULL)), 1), 0), ifNull(equals(toBool(transform(toString(events__group_0.properties___group_boolean), %(hogql_val_4)s, %(hogql_val_5)s, NULL)), 0), 0), isNull(toBool(transform(toString(events__group_0.properties___group_boolean), %(hogql_val_6)s, %(hogql_val_7)s, NULL))) 
   FROM events LEFT JOIN (
   SELECT argMax(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(groups.group_properties, %(hogql_val_0)s), ''), 'null'), '^"|"$', ''), toTimeZone(groups._timestamp, %(hogql_val_1)s)) AS properties___group_boolean, groups.group_type_index AS index, groups.group_key AS key 
   FROM groups 
@@ -66,7 +66,7 @@
 # name: TestPropertyTypes.test_resolve_property_types_event
   '''
 
-  SELECT multiply(accurateCastOrNull(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(events.properties, %(hogql_val_0)s), ''), 'null'), '^"|"$', ''), %(hogql_val_1)s), accurateCastOrNull(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(events.properties, %(hogql_val_2)s), ''), 'null'), '^"|"$', ''), %(hogql_val_3)s)), transform(toString(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(events.properties, %(hogql_val_4)s), ''), 'null'), '^"|"$', '')), %(hogql_val_5)s, %(hogql_val_6)s, NULL) AS bool 
+  SELECT multiply(accurateCastOrNull(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(events.properties, %(hogql_val_0)s), ''), 'null'), '^"|"$', ''), %(hogql_val_1)s), accurateCastOrNull(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(events.properties, %(hogql_val_2)s), ''), 'null'), '^"|"$', ''), %(hogql_val_3)s)), toBool(transform(toString(replaceRegexpAll(nullIf(nullIf(JSONExtractRaw(events.properties, %(hogql_val_4)s), ''), 'null'), '^"|"$', '')), %(hogql_val_5)s, %(hogql_val_6)s, NULL)) AS bool 
   FROM events 
   WHERE equals(events.team_id, 420) 
   LIMIT 50000

posthog/hogql/transforms/test/test_property_types.py

@@ -136,7 +136,7 @@ def test_group_boolean_property_types(self):
         )
         assert printed == self.snapshot
         assert (
-            "SELECT ifNull(equals(transform(toString(events__group_0.properties___group_boolean), hogvar, hogvar, NULL), 1), 0), ifNull(equals(transform(toString(events__group_0.properties___group_boolean), hogvar, hogvar, NULL), 0), 0), isNull(transform(toString(events__group_0.properties___group_boolean), hogvar, hogvar, NULL))"
+            "SELECT ifNull(equals(toBool(transform(toString(events__group_0.properties___group_boolean), hogvar, hogvar, NULL)), 1), 0), ifNull(equals(toBool(transform(toString(events__group_0.properties___group_boolean), hogvar, hogvar, NULL)), 0), 0), isNull(toBool(transform(toString(events__group_0.properties___group_boolean), hogvar, hogvar, NULL)))"
             in re.sub(r"%\(hogql_val_\d+\)s", "hogvar", printed)
         )
 

posthog/hogql_queries/insights/funnels/base.py

@@ -37,7 +37,7 @@
 )
 from posthog.types import EntityNode, ExclusionEntityNode
 
-JOIN_ALGOS = "direct,parallel_hash,hash,full_sorting_merge"
+JOIN_ALGOS = "auto"
 
 
 class FunnelBase(ABC):

posthog/hogql_queries/insights/funnels/test/__snapshots__/test_funnel_correlation_actors_udf.ambr

@@ -60,7 +60,7 @@
            GROUP BY aggregation_target
            HAVING ifNull(greaterOrEquals(step_reached, 0), 0))
         WHERE ifNull(greaterOrEquals(step_reached, 0), 0)
-        ORDER BY aggregation_target ASC SETTINGS join_algorithm='direct,parallel_hash,hash,full_sorting_merge') AS funnel_actors ON equals(if(not(empty(event__override.distinct_id)), event__override.person_id, event.person_id), funnel_actors.actor_id)
+        ORDER BY aggregation_target ASC SETTINGS join_algorithm='auto') AS funnel_actors ON equals(if(not(empty(event__override.distinct_id)), event__override.person_id, event.person_id), funnel_actors.actor_id)
      WHERE and(equals(event.team_id, 99999), greaterOrEquals(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-01 00:00:00', 6, 'UTC'))), less(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-08 23:59:59', 6, 'UTC'))), equals(event.team_id, 99999), greater(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), funnel_actors.first_timestamp), less(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), coalesce(funnel_actors.final_timestamp, plus(toTimeZone(funnel_actors.first_timestamp, 'UTC'), toIntervalDay(14)), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-08 23:59:59', 6, 'UTC')))), notIn(event.event, ['$pageview', 'insight analyzed']), equals(event.event, 'insight loaded'), ifNull(equals(funnel_actors.steps, 2), 0))
      GROUP BY actor_id
      ORDER BY actor_id ASC) AS source
@@ -99,7 +99,7 @@
                                                              GROUP BY aggregation_target
                                                              HAVING ifNull(greaterOrEquals(step_reached, 0), 0))
                                                           WHERE ifNull(greaterOrEquals(step_reached, 0), 0)
-                                                          ORDER BY aggregation_target ASC SETTINGS join_algorithm='direct,parallel_hash,hash,full_sorting_merge') AS funnel_actors ON equals(if(not(empty(event__override.distinct_id)), event__override.person_id, event.person_id), funnel_actors.actor_id)
+                                                          ORDER BY aggregation_target ASC SETTINGS join_algorithm='auto') AS funnel_actors ON equals(if(not(empty(event__override.distinct_id)), event__override.person_id, event.person_id), funnel_actors.actor_id)
                                                        WHERE and(equals(event.team_id, 99999), greaterOrEquals(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-01 00:00:00', 6, 'UTC'))), less(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-08 23:59:59', 6, 'UTC'))), equals(event.team_id, 99999), greater(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), funnel_actors.first_timestamp), less(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), coalesce(funnel_actors.final_timestamp, plus(toTimeZone(funnel_actors.first_timestamp, 'UTC'), toIntervalDay(14)), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-08 23:59:59', 6, 'UTC')))), notIn(event.event, ['$pageview', 'insight analyzed']), equals(event.event, 'insight loaded'), ifNull(equals(funnel_actors.steps, 2), 0))
                                                        GROUP BY actor_id
                                                        ORDER BY actor_id ASC) AS source)))
@@ -192,7 +192,7 @@
            GROUP BY aggregation_target
            HAVING ifNull(greaterOrEquals(step_reached, 0), 0))
         WHERE ifNull(greaterOrEquals(step_reached, 0), 0)
-        ORDER BY aggregation_target ASC SETTINGS join_algorithm='direct,parallel_hash,hash,full_sorting_merge') AS funnel_actors ON equals(if(not(empty(event__override.distinct_id)), event__override.person_id, event.person_id), funnel_actors.actor_id)
+        ORDER BY aggregation_target ASC SETTINGS join_algorithm='auto') AS funnel_actors ON equals(if(not(empty(event__override.distinct_id)), event__override.person_id, event.person_id), funnel_actors.actor_id)
      WHERE and(equals(event.team_id, 99999), greaterOrEquals(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-01 00:00:00', 6, 'UTC'))), less(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-08 23:59:59', 6, 'UTC'))), equals(event.team_id, 99999), greater(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), funnel_actors.first_timestamp), less(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), coalesce(funnel_actors.final_timestamp, plus(toTimeZone(funnel_actors.first_timestamp, 'UTC'), toIntervalDay(14)), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-08 23:59:59', 6, 'UTC')))), notIn(event.event, ['$pageview', 'insight analyzed', 'insight updated']), equals(event.event, 'insight loaded'), ifNull(notEquals(funnel_actors.steps, 3), 1))
      GROUP BY actor_id
      ORDER BY actor_id ASC) AS source
@@ -231,7 +231,7 @@
                                                              GROUP BY aggregation_target
                                                              HAVING ifNull(greaterOrEquals(step_reached, 0), 0))
                                                           WHERE ifNull(greaterOrEquals(step_reached, 0), 0)
-                                                          ORDER BY aggregation_target ASC SETTINGS join_algorithm='direct,parallel_hash,hash,full_sorting_merge') AS funnel_actors ON equals(if(not(empty(event__override.distinct_id)), event__override.person_id, event.person_id), funnel_actors.actor_id)
+                                                          ORDER BY aggregation_target ASC SETTINGS join_algorithm='auto') AS funnel_actors ON equals(if(not(empty(event__override.distinct_id)), event__override.person_id, event.person_id), funnel_actors.actor_id)
                                                        WHERE and(equals(event.team_id, 99999), greaterOrEquals(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-01 00:00:00', 6, 'UTC'))), less(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-08 23:59:59', 6, 'UTC'))), equals(event.team_id, 99999), greater(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), funnel_actors.first_timestamp), less(toTimeZone(toDateTime(toTimeZone(event.timestamp, 'UTC'), 'UTC'), 'UTC'), coalesce(funnel_actors.final_timestamp, plus(toTimeZone(funnel_actors.first_timestamp, 'UTC'), toIntervalDay(14)), assumeNotNull(parseDateTime64BestEffortOrNull('2021-01-08 23:59:59', 6, 'UTC')))), notIn(event.event, ['$pageview', 'insight analyzed', 'insight updated']), equals(event.event, 'insight loaded'), ifNull(notEquals(funnel_actors.steps, 3), 1))
                                                        GROUP BY actor_id
                                                        ORDER BY actor_id ASC) AS source)))
@@ -325,7 +325,7 @@
            GROUP BY aggregation_target
            HAVING ifNull(greaterOrEquals(step_reached, 0), 0))
         WHERE ifNull(greaterOrEquals(step_reached, 0), 0)
-        ORDER BY aggregation_target ASC SETTINGS join_algorithm='direct,parallel_hash,hash,full_sorting_merge') AS funnel_actors
+        ORDER BY aggregation_target ASC SETTINGS join_algorithm='auto') AS funnel_actors
      WHERE ifNull(equals(funnel_actors.steps, 2), 0)
      GROUP BY funnel_actors.actor_id
      ORDER BY funnel_actors.actor_id ASC) AS source
@@ -366,7 +366,7 @@
                                                              GROUP BY aggregation_target
                                                              HAVING ifNull(greaterOrEquals(step_reached, 0), 0))
                                                           WHERE ifNull(greaterOrEquals(step_reached, 0), 0)
-                                                          ORDER BY aggregation_target ASC SETTINGS join_algorithm='direct,parallel_hash,hash,full_sorting_merge') AS funnel_actors
+                                                          ORDER BY aggregation_target ASC SETTINGS join_algorithm='auto') AS funnel_actors
                                                        WHERE ifNull(equals(funnel_actors.steps, 2), 0)
                                                        GROUP BY funnel_actors.actor_id
                                                        ORDER BY funnel_actors.actor_id ASC) AS source)))