acryl-datahub 1.1.0.5rc9__py3-none-any.whl → 1.1.0.5rc10__py3-none-any.whl

datahub/metadata/schemas/DataHubPageModuleProperties.avsc

@@ -23,8 +23,7 @@
           "HIERARCHY": "A module displaying a hierarchy to navigate",
           "LINK": "Link type module",
           "OWNED_ASSETS": "Module displaying assets owned by a user",
-          "RICH_TEXT": "Module containing rich text to be rendered",
-          "SUBSCRIBED_ASSETS": "Module displaying assets subscribed to by a given user"
+          "RICH_TEXT": "Module containing rich text to be rendered"
         },
         "name": "DataHubPageModuleType",
         "namespace": "com.linkedin.pegasus2avro.module",
@@ -34,7 +33,6 @@
           "ASSET_COLLECTION",
           "HIERARCHY",
           "OWNED_ASSETS",
-          "SUBSCRIBED_ASSETS",
           "DOMAINS"
         ],
         "doc": "Enum containing the types of page modules that there are"

datahub/sdk/main_client.py

@@ -66,7 +66,12 @@ class DataHubClient:
         self._graph.test_connection()
 
     @classmethod
-    def from_env(cls) -> "DataHubClient":
+    def from_env(
+        cls,
+        *,
+        client_mode: ClientMode = ClientMode.SDK,
+        datahub_component: Optional[str] = None,
+    ) -> "DataHubClient":
         """Initialize a DataHubClient from the environment variables or ~/.datahubenv file.
 
         This will first check DATAHUB_GMS_URL and DATAHUB_GMS_TOKEN. If not present,
@@ -76,6 +81,10 @@ class DataHubClient:
         If you're looking to specify the server/token in code, use the
         DataHubClient(server=..., token=...) constructor instead.
 
+        Args:
+            client_mode: [internal] The client mode to use. Defaults to "SDK".
+            datahub_component: [internal] The DataHub component name to include in the user agent.
+
         Returns:
             A DataHubClient instance.
         """
@@ -83,7 +92,10 @@ class DataHubClient:
         # Inspired by the DockerClient.from_env() method.
         # TODO: This one also reads from ~/.datahubenv, so the "from_env" name might be a bit confusing.
         # That file is part of the "environment", but is not a traditional "env variable".
-        graph = get_default_graph(ClientMode.SDK)
+        graph = get_default_graph(
+            client_mode=client_mode,
+            datahub_component=datahub_component,
+        )
 
         return cls(graph=graph)
 

datahub/sdk/search_client.py

@@ -19,6 +19,7 @@ from datahub.sdk.search_filters import (
     _OrFilters,
     _StatusFilter,
 )
+from datahub.utilities.ordered_set import OrderedSet
 
 if TYPE_CHECKING:
     from datahub.sdk.main_client import DataHubClient
@@ -80,7 +81,7 @@ def compute_entity_types(
 ) -> Optional[List[str]]:
     found_filters = False
     found_positive_filters = False
-    entity_types: List[str] = []
+    entity_types: OrderedSet[str] = OrderedSet()
     for ands in filters:
         for clause in ands["and"]:
             if clause.field == _EntityTypeFilter.ENTITY_TYPE_FIELD:
@@ -88,7 +89,7 @@ def compute_entity_types(
                 if not clause.negated:
                     found_positive_filters = True
 
-                entity_types.extend(clause.values)
+                entity_types.update(clause.values)
 
     if not found_filters:
         # If we didn't find any filters, use None so we use the default set.
@@ -100,7 +101,7 @@ def compute_entity_types(
         # still want to use the default set.
         return None
 
-    return entity_types
+    return list(entity_types)
 
 
 class SearchClient:

datahub/telemetry/telemetry.py

@@ -104,7 +104,7 @@ SENTRY_DSN: Optional[str] = os.environ.get("SENTRY_DSN", None)
 SENTRY_ENVIRONMENT: str = os.environ.get("SENTRY_ENVIRONMENT", "dev")
 
 
-def _default_telemetry_properties() -> Dict[str, Any]:
+def _default_global_properties() -> Dict[str, Any]:
     return {
         "datahub_version": nice_version_name(),
         "python_version": platform.python_version(),
@@ -122,6 +122,7 @@ class Telemetry:
     context_properties: Dict[str, Any] = {}
 
     def __init__(self):
+        self.global_properties = _default_global_properties()
         self.context_properties = {}
 
         if SENTRY_DSN:
@@ -247,6 +248,10 @@ class Telemetry:
 
         return False
 
+    def add_global_property(self, key: str, value: Any) -> None:
+        self.global_properties[key] = value
+        self._update_sentry_properties()
+
     def set_context(
         self,
         server: Optional["DataHubGraph"] = None,
@@ -257,16 +262,17 @@ class Telemetry:
             **(properties or {}),
         }
 
-        if self.sentry_enabled:
-            from sentry_sdk import set_tag
+        self._update_sentry_properties()
 
-            properties = {
-                **_default_telemetry_properties(),
-                **self.context_properties,
-            }
+    def _update_sentry_properties(self) -> None:
+        properties = {
+            **self.global_properties,
+            **self.context_properties,
+        }
+        if self.sentry_enabled:
+            import sentry_sdk
 
-            for key in properties:
-                set_tag(key, properties[key])
+            sentry_sdk.set_tags(properties)
 
     def init_capture_exception(self) -> None:
         if self.sentry_enabled:
@@ -300,7 +306,7 @@ class Telemetry:
         try:
             self.mp.people_set(
                 self.client_id,
-                _default_telemetry_properties(),
+                self.global_properties,
             )
         except Exception as e:
             logger.debug(f"Error initializing telemetry: {e}")
@@ -334,7 +340,7 @@ class Telemetry:
                 logger.debug(f"Sending telemetry for {event_name}")
 
             properties = {
-                **_default_telemetry_properties(),
+                **self.global_properties,
                 **self.context_properties,
                 **properties,
             }