openmetadata-ingestion 1.8.6.0__py3-none-any.whl → 1.8.6.1__py3-none-any.whl

metadata/ingestion/lineage/sql_lineage.py

@@ -14,7 +14,7 @@ Helper functions to handle SQL lineage operations
 import itertools
 import traceback
 from collections import defaultdict
-from typing import Any, Iterable, List, Optional, Tuple, Union
+from typing import Any, Dict, Iterable, List, Optional, Tuple, Union
 
 import networkx as nx
 from collate_sqllineage.core.models import Column, DataFunction
@@ -894,7 +894,7 @@ def _get_paths_from_subtree(subtree: DiGraph) -> List[List[Any]]:
 
 
 def get_lineage_by_graph(
-    graph: DiGraph,
+    graph: Optional[DiGraph],
     metadata: OpenMetadata,
 ) -> Iterable[Either[AddLineageRequest]]:
     """
@@ -922,3 +922,27 @@ def get_lineage_by_graph(
         subtree = graph.subgraph(component).copy()
         for path in _get_paths_from_subtree(subtree):
             yield from _process_sequence(path, subtree, metadata)
+
+
+def get_lineage_by_procedure_graph(
+    procedure_graph_map: Optional[Dict],
+    metadata: OpenMetadata,
+) -> Iterable[Either[AddLineageRequest]]:
+    """
+    Generate lineage information from a directed graph.
+    """
+    if procedure_graph_map is None:
+        return
+
+    for procedure_and_procedure_graph in procedure_graph_map.values():
+        for either_lineage in get_lineage_by_graph(
+            graph=procedure_and_procedure_graph.graph,
+            metadata=metadata,
+        ):
+            if either_lineage.left is None and either_lineage.right.edge.lineageDetails:
+                either_lineage.right.edge.lineageDetails.pipeline = EntityReference(
+                    id=procedure_and_procedure_graph.procedure.id,
+                    type="storedProcedure",
+                )
+
+            yield either_lineage

metadata/ingestion/models/custom_pydantic.py

@@ -199,6 +199,7 @@ class _CustomSecretStr(SecretStr):
             and SecretsManagerFactory().get_secrets_manager()
         ):
             secret_id = self._secret_value.replace(SECRET, "")
+            logger.info(f"Getting secret value for {secret_id}")
             try:
                 return (
                     SecretsManagerFactory()

metadata/ingestion/ometa/ometa_api.py

@@ -59,7 +59,12 @@ from metadata.ingestion.ometa.mixins.user_mixin import OMetaUserMixin
 from metadata.ingestion.ometa.mixins.version_mixin import OMetaVersionMixin
 from metadata.ingestion.ometa.models import EntityList
 from metadata.ingestion.ometa.routes import ROUTES
-from metadata.ingestion.ometa.utils import get_entity_type, model_str, quote
+from metadata.ingestion.ometa.utils import (
+    decode_jwt_token,
+    get_entity_type,
+    model_str,
+    quote,
+)
 from metadata.utils.logger import ometa_logger
 from metadata.utils.secrets.secrets_manager_factory import SecretsManagerFactory
 from metadata.utils.ssl_registry import get_verify_ssl_fn
@@ -155,6 +160,7 @@ class OpenMetadata(
         ).get_secrets_manager()
 
         self._auth_provider = OpenMetadataAuthenticationProvider.create(self.config)
+        self.log_user_name_from_jwt_token()
 
         get_verify_ssl = get_verify_ssl_fn(self.config.verifySSL)
 
@@ -174,6 +180,29 @@ class OpenMetadata(
         if self.config.enableVersionValidation:
             self.validate_versions()
 
+    def log_user_name_from_jwt_token(self) -> None:
+        """
+        Log user name from JWT token.
+        """
+        # Log user name from JWT token if authProvider is openmetadata
+        if (
+            self.config.authProvider
+            and self.config.authProvider.value == "openmetadata"
+        ):
+            try:
+                # Get the JWT token from the auth provider
+                jwt_token, _ = self._auth_provider.get_access_token()
+                if jwt_token:
+                    # Decode the JWT token to extract user information
+                    payload = decode_jwt_token(jwt_token)
+                    if payload:
+                        if payload.get("sub"):
+                            logger.debug(f"Authenticated user: {payload.get('sub')}")
+                        else:
+                            logger.debug("Could not extract user name from JWT token")
+            except Exception as e:
+                logger.debug(f"Error processing JWT token: {e}")
+
     @classmethod
     def from_env(cls) -> "OpenMetadata":
         settings = OpenMetadataSettings()

metadata/ingestion/ometa/utils.py

@@ -12,9 +12,11 @@
 Helper functions to handle OpenMetadata Entities' properties
 """
 
+import base64
+import json
 import re
 import string
-from typing import Any, Type, TypeVar, Union
+from typing import Any, Dict, Optional, Type, TypeVar, Union
 
 from pydantic import BaseModel
 from requests.utils import quote as url_quote
@@ -98,3 +100,32 @@ def build_entity_reference(entity: T) -> EntityReference:
         description=entity.description,
         href=entity.href,
     )
+
+
+def decode_jwt_token(jwt_token: str) -> Optional[Dict[str, Any]]:
+    """
+    Decode JWT token to extract payload without verification.
+    JWT tokens have three parts: header.payload.signature
+    We only decode the payload part to get user information.
+    """
+    from metadata.utils.logger import ometa_logger
+
+    logger = ometa_logger()
+    try:
+        # Split the JWT token into its three parts
+        parts = jwt_token.split(".")
+        if len(parts) != 3:
+            logger.debug("Invalid JWT token format: expected 3 parts")
+            return None
+
+        # Decode the payload (second part)
+        payload = parts[1]
+        # Add padding if needed for base64 decoding
+        payload += "=" * (4 - len(payload) % 4) if len(payload) % 4 else ""
+
+        # Decode base64 and parse JSON
+        decoded_payload = base64.urlsafe_b64decode(payload)
+        return json.loads(decoded_payload.decode("utf-8"))
+    except Exception as e:
+        logger.debug(f"Failed to decode JWT token: {e}")
+        return None

metadata/ingestion/source/database/lineage_source.py

@@ -41,6 +41,7 @@ from metadata.ingestion.lineage.models import ConnectionTypeDialectMapper, Diale
 from metadata.ingestion.lineage.sql_lineage import (
     get_column_fqn,
     get_lineage_by_graph,
+    get_lineage_by_procedure_graph,
     get_lineage_by_query,
 )
 from metadata.ingestion.models.ometa_lineage import OMetaLineageRequest
@@ -49,7 +50,7 @@ from metadata.ingestion.source.database.query_parser_source import QueryParserSo
 from metadata.ingestion.source.models import TableView
 from metadata.utils import fqn
 from metadata.utils.db_utils import get_view_lineage
-from metadata.utils.filters import filter_by_table
+from metadata.utils.filters import filter_by_database, filter_by_schema, filter_by_table
 from metadata.utils.logger import ingestion_logger
 
 logger = ingestion_logger()
@@ -314,9 +315,19 @@ class LineageSource(QueryParserSource, ABC):
     ) -> Iterable[Either[AddLineageRequest]]:
         try:
             for view in views:
-                if filter_by_table(
-                    self.source_config.tableFilterPattern,
-                    view.table_name,
+                if (
+                    filter_by_database(
+                        self.source_config.databaseFilterPattern,
+                        view.db_name,
+                    )
+                    or filter_by_schema(
+                        self.source_config.schemaFilterPattern,
+                        view.schema_name,
+                    )
+                    or filter_by_table(
+                        self.source_config.tableFilterPattern,
+                        view.table_name,
+                    )
                 ):
                     self.status.filter(
                         view.table_name,
@@ -441,10 +452,21 @@ class LineageSource(QueryParserSource, ABC):
         Based on the query logs, prepare the lineage
         and send it to the sink
         """
+        if (
+            self.procedure_graph_map is None
+            and self.source_config.enableTempTableLineage
+        ):
+            # Create a dictionary to store the directed graph for each procedure
+            self.procedure_graph_map = {}
+
         if self.source_config.processViewLineage:
             yield from self.yield_view_lineage() or []
         if self.source_config.processStoredProcedureLineage:
             yield from self.yield_procedure_lineage() or []
+            yield from get_lineage_by_procedure_graph(
+                procedure_graph_map=self.procedure_graph_map,
+                metadata=self.metadata,
+            )
         if self.source_config.processQueryLineage:
             if hasattr(self.service_connection, "supportsLineageExtraction"):
                 yield from self.yield_query_lineage() or []

metadata/ingestion/source/database/query_parser_source.py

@@ -64,6 +64,7 @@ class QueryParserSource(Source, ABC):
         self.source_config = self.config.sourceConfig.config
         self.start, self.end = get_start_and_end(self.source_config.queryLogDuration)
         self.graph = None
+        self.procedure_graph_map = None
 
         self.engine = None
         if get_engine:

metadata/ingestion/source/database/stored_procedures_mixin.py

@@ -19,6 +19,7 @@ from collections import defaultdict
 from datetime import datetime
 from typing import Dict, Iterable, List, Optional, Union
 
+from networkx import DiGraph
 from pydantic import BaseModel, ConfigDict, Field
 from sqlalchemy.engine import Engine
 
@@ -41,7 +42,12 @@ from metadata.ingestion.lineage.sql_lineage import get_lineage_by_query
 from metadata.ingestion.models.ometa_lineage import OMetaLineageRequest
 from metadata.ingestion.models.topology import Queue
 from metadata.ingestion.ometa.ometa_api import OpenMetadata
-from metadata.utils.filters import filter_by_stored_procedure
+from metadata.utils.filters import (
+    filter_by_database,
+    filter_by_schema,
+    filter_by_stored_procedure,
+)
+from metadata.utils.helpers import pprint_format_object
 from metadata.utils.logger import ingestion_logger
 from metadata.utils.stored_procedures import get_procedure_name_from_call
 from metadata.utils.time_utils import datetime_to_timestamp
@@ -80,6 +86,17 @@ class ProcedureAndQuery(BaseModel):
     model_config = ConfigDict(populate_by_name=True)
 
 
+class ProcedureAndProcedureGraph(BaseModel):
+    """
+    Model to hold the procedure and its graph
+    """
+
+    model_config = ConfigDict(arbitrary_types_allowed=True)
+
+    procedure: StoredProcedure
+    graph: DiGraph
+
+
 class StoredProcedureLineageMixin(ABC):
     """
     The full flow is:
@@ -137,6 +154,17 @@ class StoredProcedureLineageMixin(ABC):
                     )
                 )
 
+        queries_count_per_procedure = {
+            procedure_name: len(queries)
+            for procedure_name, queries in queries_dict.items()
+        }
+        logger.info(
+            f"Count of queries executed for stored procedures: {sum(queries_count_per_procedure.values())}"
+        )
+        logger.info(
+            f"Count of queries per stored procedure: {pprint_format_object(queries_count_per_procedure)}"
+        )
+
         return queries_dict
 
     @staticmethod
@@ -166,6 +194,18 @@ class StoredProcedureLineageMixin(ABC):
             query_type=query_by_procedure.query_type,
             query_text=query_by_procedure.query_text,
         ):
+            graph = None
+            if self.source_config.enableTempTableLineage:
+                if not self.procedure_graph_map.get(procedure.fullyQualifiedName.root):
+                    # Map to store the directed graph for each procedure with its FQN as key
+                    self.procedure_graph_map[
+                        procedure.fullyQualifiedName.root
+                    ] = ProcedureAndProcedureGraph(procedure=procedure, graph=DiGraph())
+
+                graph = self.procedure_graph_map.get(
+                    procedure.fullyQualifiedName.root
+                ).graph
+
             self.stored_procedure_query_lineage = True
             for either_lineage in get_lineage_by_query(
                 self.metadata,
@@ -178,6 +218,7 @@ class StoredProcedureLineageMixin(ABC):
                 ),
                 timeout_seconds=self.source_config.parsingTimeoutLimit,
                 lineage_source=LineageSource.QueryLineage,
+                graph=graph,
             ):
                 if (
                     either_lineage.left is None
@@ -290,9 +331,19 @@ class StoredProcedureLineageMixin(ABC):
             or []
         ):
             if procedure:
-                if filter_by_stored_procedure(
-                    self.source_config.storedProcedureFilterPattern,
-                    procedure.name.root,
+                if (
+                    filter_by_database(
+                        self.source_config.databaseFilterPattern,
+                        procedure.database.name,
+                    )
+                    or filter_by_schema(
+                        self.source_config.schemaFilterPattern,
+                        procedure.databaseSchema.name,
+                    )
+                    or filter_by_stored_procedure(
+                        self.source_config.storedProcedureFilterPattern,
+                        procedure.name.root,
+                    )
                 ):
                     self.status.filter(
                         procedure.name.root,

metadata/ingestion/source/pipeline/airflow/metadata.py

@@ -187,36 +187,44 @@ class AirflowSource(PipelineServiceSource):
         """
         Return the DagRuns of given dag
         """
-        dag_run_list = (
-            self.session.query(
-                DagRun.dag_id,
-                DagRun.run_id,
-                DagRun.queued_at,
-                DagRun.execution_date,
-                DagRun.start_date,
-                DagRun.state,
+        try:
+            dag_run_list = (
+                self.session.query(
+                    DagRun.dag_id,
+                    DagRun.run_id,
+                    DagRun.queued_at,
+                    DagRun.execution_date,
+                    DagRun.start_date,
+                    DagRun.state,
+                )
+                .filter(DagRun.dag_id == dag_id)
+                .order_by(DagRun.execution_date.desc())
+                .limit(self.config.serviceConnection.root.config.numberOfStatus)
+                .all()
             )
-            .filter(DagRun.dag_id == dag_id)
-            .order_by(DagRun.execution_date.desc())
-            .limit(self.config.serviceConnection.root.config.numberOfStatus)
-            .all()
-        )
 
-        dag_run_dict = [dict(elem) for elem in dag_run_list]
-
-        # Build DagRun manually to not fall into new/old columns from
-        # different Airflow versions
-        return [
-            DagRun(
-                dag_id=elem.get("dag_id"),
-                run_id=elem.get("run_id"),
-                queued_at=elem.get("queued_at"),
-                execution_date=elem.get("execution_date"),
-                start_date=elem.get("start_date"),
-                state=elem.get("state"),
+            dag_run_dict = [dict(elem) for elem in dag_run_list]
+
+            # Build DagRun manually to not fall into new/old columns from
+            # different Airflow versions
+            return [
+                DagRun(
+                    dag_id=elem.get("dag_id"),
+                    run_id=elem.get("run_id"),
+                    queued_at=elem.get("queued_at"),
+                    execution_date=elem.get("execution_date"),
+                    start_date=elem.get("start_date"),
+                    state=elem.get("state"),
+                )
+                for elem in dag_run_dict
+            ]
+        except Exception as exc:
+            logger.debug(traceback.format_exc())
+            logger.warning(
+                f"Could not get pipeline status for {dag_id}. "
+                f"This might be due to Airflow version incompatibility - {exc}"
             )
-            for elem in dag_run_dict
-        ]
+            return []
 
     def get_task_instances(
         self, dag_id: str, run_id: str, serialized_tasks: List[AirflowTask]
@@ -369,16 +377,27 @@ class AirflowSource(PipelineServiceSource):
                 break
             for serialized_dag in results:
                 try:
-                    dag_model = (
-                        self.session.query(DagModel)
-                        .filter(DagModel.dag_id == serialized_dag[0])
-                        .one_or_none()
-                    )
-                    pipeline_state = (
-                        PipelineState.Active.value
-                        if dag_model and not dag_model.is_paused
-                        else PipelineState.Inactive.value
-                    )
+                    # Query only the is_paused column from DagModel
+                    try:
+                        is_paused_result = (
+                            self.session.query(DagModel.is_paused)
+                            .filter(DagModel.dag_id == serialized_dag[0])
+                            .scalar()
+                        )
+                        pipeline_state = (
+                            PipelineState.Active.value
+                            if not is_paused_result
+                            else PipelineState.Inactive.value
+                        )
+                    except Exception as exc:
+                        logger.debug(traceback.format_exc())
+                        logger.warning(
+                            f"Could not query DagModel.is_paused for {serialized_dag[0]}. "
+                            f"Using default pipeline state - {exc}"
+                        )
+                        # If we can't query is_paused, assume the pipeline is active
+                        pipeline_state = PipelineState.Active.value
+
                     data = serialized_dag[1]["dag"]
                     dag = AirflowDagDetails(
                         dag_id=serialized_dag[0],

metadata/ingestion/source/pipeline/airflow/utils.py

@@ -23,6 +23,7 @@ from metadata.utils.logger import ingestion_logger
 logger = ingestion_logger()
 
 
+# pylint: disable=too-many-branches,too-many-return-statements,too-many-nested-blocks
 def get_schedule_interval(pipeline_data: Dict[str, Any]) -> Optional[str]:
     """
     Fetch Schedule Intervals from Airflow Dags
@@ -40,7 +41,49 @@ def get_schedule_interval(pipeline_data: Dict[str, Any]) -> Optional[str]:
 
             expression_class = timetable.get("__type")
             if expression_class:
-                return import_from_module(expression_class)().summary
+                try:
+                    # Try to instantiate the timetable class safely
+                    timetable_class = import_from_module(expression_class)
+
+                    # Handle special cases for classes that require arguments
+                    if "DatasetTriggeredTimetable" in expression_class:
+                        # DatasetTriggeredTimetable requires datasets argument
+                        # For now, return a descriptive string since we can't instantiate it properly
+                        return "Dataset Triggered"
+                    if "CronDataIntervalTimetable" in expression_class:
+                        # Handle cron-based timetables
+                        try:
+                            return timetable_class().summary
+                        except (TypeError, AttributeError):
+                            return "Cron Based"
+                    else:
+                        # Try to instantiate with no arguments
+                        try:
+                            return timetable_class().summary
+                        except (TypeError, AttributeError):
+                            # If summary attribute doesn't exist, try to get a string representation
+                            try:
+                                instance = timetable_class()
+                                return str(instance)
+                            except TypeError:
+                                # If instantiation fails, return the class name
+                                return f"Custom Timetable ({expression_class.split('.')[-1]})"
+                except ImportError as import_error:
+                    logger.debug(
+                        f"Could not import timetable class {expression_class}: {import_error}"
+                    )
+                    return f"Custom Timetable ({expression_class.split('.')[-1]})"
+                except TypeError as type_error:
+                    # If instantiation fails due to missing arguments, log and continue
+                    logger.debug(
+                        f"Could not instantiate timetable class {expression_class}: {type_error}"
+                    )
+                    return f"Custom Timetable ({expression_class.split('.')[-1]})"
+                except Exception as inst_error:
+                    logger.debug(
+                        f"Error instantiating timetable class {expression_class}: {inst_error}"
+                    )
+                    return f"Custom Timetable ({expression_class.split('.')[-1]})"
 
         if schedule:
             if isinstance(schedule, str):
@@ -57,7 +100,6 @@ def get_schedule_interval(pipeline_data: Dict[str, Any]) -> Optional[str]:
 
     except Exception as exc:
         logger.debug(traceback.format_exc())
-        logger.warning(
-            f"Couldn't fetch schedule interval for dag {pipeline_data.get('_dag_id'): [{exc}]}"
-        )
+        dag_id = pipeline_data.get("_dag_id", "unknown")
+        logger.warning(f"Couldn't fetch schedule interval for dag {dag_id}: {exc}")
     return None

metadata/utils/helpers.py

@@ -17,6 +17,7 @@ from __future__ import annotations
 
 import hashlib
 import itertools
+import pprint
 import re
 import shutil
 import sys
@@ -568,3 +569,10 @@ def evaluate_threshold(threshold: int, operator: str, result: int) -> bool:
         f"Invalid threshold: {threshold}, "
         "Allowed format: <, >, <=, >=, ==, !=. Example: >5"
     )
+
+
+def pprint_format_object(data: Any) -> str:
+    """
+    Pretty print an object in a format that is easy to read
+    """
+    return pprint.pformat(data, width=150)