feat(errors): Add YAML definitions for new features columns

snuba/datasets/configuration/discover/storages/discover.yaml

@@ -111,6 +111,8 @@ query_processors:
       hash_map_name: _tags_hash_map
       killswitch: tags_hash_map_enabled
   - processor: EmptyTagConditionProcessor
+    args:
+      column_name: tags.key
   - processor: ArrayJoinKeyValueOptimizer
     args:
       column_name: tags

snuba/datasets/configuration/events/entities/events.yaml

@@ -49,6 +49,24 @@ schema:
           inner_type: { type: UInt, args: { size: 64 } },
         },
     },
+    {
+      name: features,
+      type: Nested,
+      args:
+        {
+          subcolumns:
+            [{ name: key, type: String }, { name: value, type: String }],
+        },
+    },
+    {
+      name: _features_hash_map,
+      type: Array,
+      args:
+        {
+          schema_modifiers: [readonly],
+          inner_type: { type: UInt, args: { size: 64 } },
+        },
+    },
     {
       name: contexts,
       type: Nested,
@@ -297,6 +315,18 @@ storages:
             from_col_name: tags_value
             to_function_name: arrayJoin
             to_function_column: tags.value
+        - mapper: ColumnToFunctionOnColumn
+          args:
+            from_table_name: null
+            from_col_name: features_key
+            to_function_name: arrayJoin
+            to_function_column: features.key
+        - mapper: ColumnToFunctionOnColumn
+          args:
+            from_table_name: null
+            from_col_name: features_value
+            to_function_name: arrayJoin
+            to_function_column: features.value
       subscriptables:
         - mapper: SubscriptableMapper
           args:
@@ -314,6 +344,14 @@ storages:
             to_nested_col_name: contexts
             value_subcolumn_name: value
             nullable: false
+        - mapper: SubscriptableMapper
+          args:
+            from_column_table: null
+            from_column_name: features
+            to_nested_col_table: null
+            to_nested_col_name: features
+            value_subcolumn_name: value
+            nullable: false
   - storage: errors
     is_writable: true
     translation_mappers:
@@ -408,6 +446,18 @@ storages:
             from_col_name: tags_value
             to_function_name: arrayJoin
             to_function_column: tags.value
+        - mapper: ColumnToFunctionOnColumn
+          args:
+            from_table_name: null
+            from_col_name: features_key
+            to_function_name: arrayJoin
+            to_function_column: features.key
+        - mapper: ColumnToFunctionOnColumn
+          args:
+            from_table_name: null
+            from_col_name: features_value
+            to_function_name: arrayJoin
+            to_function_column: features.value
       subscriptables:
         - mapper: SubscriptableMapper
           args:
@@ -425,6 +475,14 @@ storages:
             to_nested_col_name: contexts
             value_subcolumn_name: value
             nullable: false
+        - mapper: SubscriptableMapper
+          args:
+            from_column_table: null
+            from_column_name: features
+            to_nested_col_table: null
+            to_nested_col_name: features
+            value_subcolumn_name: value
+            nullable: false
 storage_selector:
   selector: ErrorsQueryStorageSelector
 query_processors:

snuba/datasets/configuration/events/storages/errors.yaml

@@ -76,6 +76,24 @@ schema:
             inner_type: { type: UInt, args: { size: 64 } },
           },
       },
+      {
+        name: features,
+        type: Nested,
+        args:
+          {
+            subcolumns:
+              [{ name: key, type: String }, { name: value, type: String }],
+          },
+      },
+      {
+        name: _features_hash_map,
+        type: Array,
+        args:
+          {
+            schema_modifiers: [readonly],
+            inner_type: { type: UInt, args: { size: 64 } },
+          },
+      },
       {
         name: contexts,
         type: Nested,
@@ -310,10 +328,23 @@ query_processors:
       column_name: tags
       hash_map_name: _tags_hash_map
       killswitch: events_tags_hash_map_enabled
+  - processor: MappingOptimizer
+    args:
+      column_name: features
+      hash_map_name: _features_hash_map
+      killswitch: events_features_hash_map_enabled
   - processor: EmptyTagConditionProcessor
+    args:
+      column_name: tags.key.key
+  - processor: EmptyTagConditionProcessor
+    args:
+      column_name: features.key
   - processor: ArrayJoinKeyValueOptimizer
     args:
       column_name: tags
+  - processor: ArrayJoinKeyValueOptimizer
+    args:
+      column_name: features
   - processor: PrewhereProcessor
     args:
       omit_if_final:

snuba/datasets/configuration/events/storages/errors_ro.yaml

@@ -76,6 +76,24 @@ schema:
             inner_type: { type: UInt, args: { size: 64 } },
           },
       },
+      {
+        name: features,
+        type: Nested,
+        args:
+          {
+            subcolumns:
+              [{ name: key, type: String }, { name: value, type: String }],
+          },
+      },
+      {
+        name: _features_hash_map,
+        type: Array,
+        args:
+          {
+            schema_modifiers: [readonly],
+            inner_type: { type: UInt, args: { size: 64 } },
+          },
+      },
       {
         name: contexts,
         type: Nested,
@@ -306,10 +324,20 @@ query_processors:
       column_name: tags
       hash_map_name: _tags_hash_map
       killswitch: events_tags_hash_map_enabled
+  - processor: MappingOptimizer
+    args:
+      column_name: features
+      hash_map_name: _features_hash_map
+      killswitch: events_features_hash_map_enabled
   - processor: EmptyTagConditionProcessor
+    args:
+      column_name: tags.key
   - processor: ArrayJoinKeyValueOptimizer
     args:
       column_name: tags
+  - processor: ArrayJoinKeyValueOptimizer
+    args:
+      column_name: features
   - processor: PrewhereProcessor
     args:
       omit_if_final:

snuba/datasets/configuration/issues/storages/search_issues.yaml

@@ -105,6 +105,8 @@ query_processors:
       hash_map_name: _tags_hash_map
       killswitch: tags_hash_map_enabled
   - processor: EmptyTagConditionProcessor
+    args:
+      column_name: tags.key
   - processor: ArrayJoinKeyValueOptimizer
     args:
       column_name: tags

snuba/datasets/configuration/replays/storages/replays.yaml

@@ -190,6 +190,8 @@ query_processors:
       hash_map_name: _tags_hash_map
       killswitch: tags_hash_map_enabled
   - processor: EmptyTagConditionProcessor
+    args:
+      column_name: tags.key
   - processor: ArrayJoinKeyValueOptimizer
     args:
       column_name: tags

snuba/datasets/configuration/spans/storages/spans.yaml

@@ -156,6 +156,8 @@ query_processors:
       hash_map_name: _measurements_hash_map
       killswitch: measurements_hash_map_enabled
   - processor: EmptyTagConditionProcessor
+    args:
+      column_name: tags.key
   - processor: ArrayJoinKeyValueOptimizer
     args:
       column_name: tags

snuba/datasets/configuration/transactions/storages/transactions.yaml

@@ -215,6 +215,8 @@ query_processors:
       hash_map_name: _tags_hash_map
       killswitch: tags_hash_map_enabled
   - processor: EmptyTagConditionProcessor
+    args:
+      column_name: tags.key
   - processor: ArrayJoinKeyValueOptimizer
     args:
       column_name: tags

snuba/query/processors/physical/empty_tag_condition_processor.py

@@ -30,18 +30,24 @@ class EmptyTagConditionProcessor(ClickhouseQueryProcessor):
     the `has` function.
     """
 
+    def __init__(self, column_name: str) -> None:
+        self.column_name = column_name
+
     def process_query(self, query: Query, query_settings: QuerySettings) -> None:
         def process_condition(exp: Expression) -> Expression:
             result = CONDITION_PATTERN.match(exp)
             if result is not None:
                 key_column = result.optional_string(KEY_COL_MAPPING_PARAM)
-                if key_column == "tags.key":
+                if key_column == self.column_name:
                     rhs = result.optional_string(KEY_MAPPING_PARAM)
                     table_name = result.optional_string(TABLE_MAPPING_PARAM)
                     replacement = FunctionCall(
                         exp.alias,
                         "has",
-                        (Column(None, table_name, "tags.key"), Literal(None, rhs)),
+                        (
+                            Column(None, table_name, self.column_name),
+                            Literal(None, rhs),
+                        ),
                     )
 
                     assert isinstance(exp, FunctionCall)

tests/query/processors/test_mapping_optimizer_no_useless_conditions.py

@@ -269,8 +269,12 @@ def test_useless_has_condition(
 
     # change the existence expression to be a has(tags, 'my_tag') expression for boh queries
     # this allows reuse of the previous test cases
-    EmptyTagConditionProcessor().process_query(input_query, HTTPQuerySettings())
-    EmptyTagConditionProcessor().process_query(expected_query, HTTPQuerySettings())
+    EmptyTagConditionProcessor("tags.key").process_query(
+        input_query, HTTPQuerySettings()
+    )
+    EmptyTagConditionProcessor("tags.key").process_query(
+        expected_query, HTTPQuerySettings()
+    )
 
     MappingOptimizer(
         column_name="tags",