adding attributes field to MetaData for GroupBy and Join

api/py/ai/chronon/group_by.py

@@ -363,6 +363,7 @@ def GroupBy(
     derivations: Optional[List[ttypes.Derivation]] = None,
     deprecation_date: Optional[str] = None,
     description: Optional[str] = None,
+    attributes: Optional[Dict[str, str]] = None,
     **kwargs,
 ) -> ttypes.GroupBy:
     """
@@ -481,6 +482,7 @@ def GroupBy(
         Additional properties that would be passed to run.py if specified under additional_args property.
         And provides an option to pass custom values to the processing logic.
     :param description: optional description of this GroupBy
+    :param attributes:  any extra attributes of this GroupBy
     :type kwargs: Dict[str, str]
     :return:
         A GroupBy object containing specified aggregations.
@@ -560,6 +562,7 @@ def _normalize_source(source):
         offlineSchedule=offline_schedule,
         deprecationDate=deprecation_date,
         description=description,
+        attributes=attributes,
     )
 
     group_by = ttypes.GroupBy(

api/py/ai/chronon/join.py

@@ -396,6 +396,7 @@ def Join(
     tags: Optional[Dict[str, str]] = None,
     description: Optional[str] = None,
     model_transforms: Optional[api.ModelTransforms] = None,
+    attributes: Optional[Dict[str, str]] = None,
     **kwargs,
 ) -> api.Join:
     """
@@ -498,6 +499,7 @@ def Join(
     :param description: optional description of this Join
     :param model_transforms:
         A list of model transforms to convert derivation outputs to model outputs using model-based transformations
+    :param attributes:  any extra attributes of this Join
     :return:
         A join object that can be used to backfill or serve data. For ML use-cases this should map 1:1 to model.
     """
@@ -607,6 +609,7 @@ def Join(
         historicalBackfill=historical_backfill,
         deprecationDate=deprecation_date,
         description=description,
+        attributes=attributes,
     )
 
     return api.Join(

api/py/test/test_group_by.py

@@ -295,6 +295,26 @@ def test_group_by_with_description():
     assert gb.metaData.description == "GroupBy description"
 
 
+def test_group_by_with_attributes():
+  gb = group_by.GroupBy(
+      sources=[
+        ttypes.EventSource(
+            table="event_table1",
+            query=query.Query(
+                selects=None,
+                time_column="ts"
+            )
+        )
+      ],
+      keys=["key1", "key2"],
+      aggregations=[group_by.Aggregation(input_column="event_id", operation=ttypes.Operation.SUM)],
+      name="test.additional_metadata_gb",
+      attributes=["attr1", "attr1Val"]
+  )
+  assert gb.metaData.attributes == ["attr1", "attr1Val"]
+
+
+
 def test_derivation():
     derivation = Derivation(name="derivation_name", expression="derivation_expression")
     expected_derivation = ttypes.Derivation(

api/py/test/test_join.py

@@ -62,6 +62,14 @@ def test_join_with_description():
     )
     assert join.metaData.description == "Join description"
 
+def test_join_with_attributes():
+  join = Join(
+      left=event_source("sample_namespace.sample_table"),
+      right_parts=[right_part(event_source("sample_namespace.another_table"))],
+      attributes=["attr1", "attr1Val"]
+  )
+  assert join.metaData.attributes == ["attr1", "attr1Val"]
+
 def test_deduped_dependencies():
     """
     Check left and right dependencies are deduped in metadata.

api/src/main/scala/ai/chronon/api/Builders.scala

@@ -274,7 +274,8 @@ object Builders {
         tableProperties: Map[String, String] = Map.empty,
         historicalBackill: Boolean = true,
         deprecationDate: String = null,
-        description: String = null
+        description: String = null,
+        attributes: Map[String, String] = Map.empty
     ): MetaData = {
       val result = new MetaData()
       result.setName(name)
@@ -297,6 +298,9 @@ object Builders {
       if (description != null) {
         result.setDescription(description)
       }
+      if (attributes != null) {
+        result.setAttributes(attributes.toJava)
+      }
       result
     }
   }

api/src/test/scala/ai/chronon/api/test/ExtensionsTest.scala

@@ -248,4 +248,26 @@ class ExtensionsTest {
     assertEquals(join1.semanticHash(excludeTopic = true), join2.semanticHash(excludeTopic = true))
     assertEquals(join1.semanticHash(excludeTopic = false), join2.semanticHash(excludeTopic = false))
   }
+
+  @Test
+  def semanticHashIgnoresMetadataAttributes(): Unit = {
+    val metadata = Builders.MetaData(name = "test", attributes = Map("owner" -> "Lucie"))
+    val groupBy = Builders.GroupBy(
+      sources = Seq(Builders.Source.events(query = null, table = "db.gb_table", topic = "test.gb_topic")),
+      keyColumns = Seq("a", "c"),
+      metaData = metadata)
+    val join1 = Builders.Join(
+      left = Builders.Source.events(query = null, table = "db.join_table", topic = "test.join_topic"),
+      joinParts = Seq(Builders.JoinPart(groupBy = groupBy)),
+      metaData = metadata,
+      derivations = Seq(Builders.Derivation(name = "*", expression = "*", metaData = metadata))
+    )
+    val updatedMetadata = Builders.MetaData(name = "test", attributes = Map("owner" -> "Katie"))
+    val join2 = join1.deepCopy()
+    join2.setMetaData(updatedMetadata)
+    join2.joinParts.get(0).groupBy.setMetaData(updatedMetadata)
+    join2.derivations.get(0).setMetaData(updatedMetadata)
+    assertEquals(join1.semanticHash(excludeTopic = true), join2.semanticHash(excludeTopic = true))
+    assertEquals(join1.semanticHash(excludeTopic = false), join2.semanticHash(excludeTopic = false))
+  }
 }

api/thrift/api.thrift

@@ -263,6 +263,7 @@ struct MetaData {
     15: optional string deprecationDate
     // Description for the object holding this metadata
     16: optional string description
+    17: optional map<string, string> attributes
 }
 
 // Equivalent to a FeatureSet in chronon terms