apache-airflow-providers-openlineage 2.4.0__tar.gz → 2.5.0rc1__tar.gz

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/PKG-INFO

@@ -1,11 +1,11 @@
 Metadata-Version: 2.4
 Name: apache-airflow-providers-openlineage
-Version: 2.4.0
+Version: 2.5.0rc1
 Summary: Provider package apache-airflow-providers-openlineage for Apache Airflow
 Keywords: airflow-provider,openlineage,airflow,integration
 Author-email: Apache Software Foundation <dev@airflow.apache.org>
 Maintainer-email: Apache Software Foundation <dev@airflow.apache.org>
-Requires-Python: ~=3.9
+Requires-Python: ~=3.10
 Description-Content-Type: text/x-rst
 Classifier: Development Status :: 5 - Production/Stable
 Classifier: Environment :: Console
@@ -15,20 +15,19 @@ Classifier: Intended Audience :: System Administrators
 Classifier: Framework :: Apache Airflow
 Classifier: Framework :: Apache Airflow :: Provider
 Classifier: License :: OSI Approved :: Apache Software License
-Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Topic :: System :: Monitoring
-Requires-Dist: apache-airflow>=2.10.0
-Requires-Dist: apache-airflow-providers-common-sql>=1.20.0
-Requires-Dist: apache-airflow-providers-common-compat>=1.4.0
+Requires-Dist: apache-airflow>=2.10.0rc1
+Requires-Dist: apache-airflow-providers-common-sql>=1.20.0rc1
+Requires-Dist: apache-airflow-providers-common-compat>=1.4.0rc1
 Requires-Dist: attrs>=22.2
-Requires-Dist: openlineage-integration-common>=1.31.0
-Requires-Dist: openlineage-python>=1.31.0
+Requires-Dist: openlineage-integration-common>=1.34.0
+Requires-Dist: openlineage-python>=1.34.0
 Project-URL: Bug Tracker, https://github.com/apache/airflow/issues
-Project-URL: Changelog, https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.4.0/changelog.html
-Project-URL: Documentation, https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.4.0
+Project-URL: Changelog, https://airflow.staged.apache.org/docs/apache-airflow-providers-openlineage/2.5.0/changelog.html
+Project-URL: Documentation, https://airflow.staged.apache.org/docs/apache-airflow-providers-openlineage/2.5.0
 Project-URL: Mastodon, https://fosstodon.org/@airflow
 Project-URL: Slack Chat, https://s.apache.org/airflow-slack
 Project-URL: Source Code, https://github.com/apache/airflow
@@ -59,7 +58,7 @@ Project-URL: YouTube, https://www.youtube.com/channel/UCSXwxpWZQ7XZ1WL3wqevChA/
 
 Package ``apache-airflow-providers-openlineage``
 
-Release: ``2.4.0``
+Release: ``2.5.0``
 
 
 `OpenLineage <https://openlineage.io/>`__
@@ -72,7 +71,7 @@ This is a provider package for ``openlineage`` provider. All classes for this pr
 are in ``airflow.providers.openlineage`` python package.
 
 You can find package information and changelog for the provider
-in the `documentation <https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.4.0/>`_.
+in the `documentation <https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.5.0/>`_.
 
 Installation
 ------------
@@ -81,7 +80,7 @@ You can install this package on top of an existing Airflow 2 installation (see `
 for the minimum Airflow version supported) via
 ``pip install apache-airflow-providers-openlineage``
 
-The package supports the following python versions: 3.9,3.10,3.11,3.12
+The package supports the following python versions: 3.10,3.11,3.12
 
 Requirements
 ------------
@@ -93,8 +92,8 @@ PIP package                                 Version required
 ``apache-airflow-providers-common-sql``     ``>=1.20.0``
 ``apache-airflow-providers-common-compat``  ``>=1.4.0``
 ``attrs``                                   ``>=22.2``
-``openlineage-integration-common``          ``>=1.31.0``
-``openlineage-python``                      ``>=1.31.0``
+``openlineage-integration-common``          ``>=1.34.0``
+``openlineage-python``                      ``>=1.34.0``
 ==========================================  ==================
 
 Cross provider package dependencies
@@ -118,5 +117,5 @@ Dependent package
 ==================================================================================================================  =================
 
 The changelog for the provider package can be found in the
-`changelog <https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.4.0/changelog.html>`_.
+`changelog <https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.5.0/changelog.html>`_.
 

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/README.rst

@@ -23,7 +23,7 @@
 
 Package ``apache-airflow-providers-openlineage``
 
-Release: ``2.4.0``
+Release: ``2.5.0``
 
 
 `OpenLineage <https://openlineage.io/>`__
@@ -36,7 +36,7 @@ This is a provider package for ``openlineage`` provider. All classes for this pr
 are in ``airflow.providers.openlineage`` python package.
 
 You can find package information and changelog for the provider
-in the `documentation <https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.4.0/>`_.
+in the `documentation <https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.5.0/>`_.
 
 Installation
 ------------
@@ -45,7 +45,7 @@ You can install this package on top of an existing Airflow 2 installation (see `
 for the minimum Airflow version supported) via
 ``pip install apache-airflow-providers-openlineage``
 
-The package supports the following python versions: 3.9,3.10,3.11,3.12
+The package supports the following python versions: 3.10,3.11,3.12
 
 Requirements
 ------------
@@ -57,8 +57,8 @@ PIP package                                 Version required
 ``apache-airflow-providers-common-sql``     ``>=1.20.0``
 ``apache-airflow-providers-common-compat``  ``>=1.4.0``
 ``attrs``                                   ``>=22.2``
-``openlineage-integration-common``          ``>=1.31.0``
-``openlineage-python``                      ``>=1.31.0``
+``openlineage-integration-common``          ``>=1.34.0``
+``openlineage-python``                      ``>=1.34.0``
 ==========================================  ==================
 
 Cross provider package dependencies
@@ -82,4 +82,4 @@ Dependent package
 ==================================================================================================================  =================
 
 The changelog for the provider package can be found in the
-`changelog <https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.4.0/changelog.html>`_.
+`changelog <https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.5.0/changelog.html>`_.

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/docs/changelog.rst

@@ -26,6 +26,40 @@
 Changelog
 ---------
 
+2.5.0
+.....
+
+Features
+~~~~~~~~
+
+* ``[OpenLineage] Added operator_provider_version to task event (#52468)``
+
+Bug Fixes
+~~~~~~~~~
+
+* ``fix non existing 'ti.dag_run' access in openlineage provider (#51932)``
+* ``Fix type import to AbstractOperator (#51773)``
+
+Misc
+~~~~
+
+* ``Move 'BaseHook' implementation to task SDK (#51873)``
+* ``Disable UP038 ruff rule and revert mandatory 'X | Y' in insintance checks (#52644)``
+* ``Add a bunch of no-redef ignores so Mypy is happy (#52507)``
+* ``chore: use task_instance as source for all airflow identifiers used in listener (#52339)``
+* ``Drop support for Python 3.9 (#52072)``
+* ``Replace 'models.BaseOperator' to Task SDK one for Standard Provider (#52292)``
+* ``Use BaseSensorOperator from task sdk in providers (#52296)``
+* ``nit: bump openlineage libraries requirement to 1.34 (#52075)``
+* ``Fixing ruff static check failures on main (#51937)``
+
+.. Below changes are excluded from the changelog. Move them to
+   appropriate section above if needed. Do not delete the lines(!):
+   * ``Make sure all test version imports come from test_common (#52425)``
+   * ``Remove db_tests from openlineage provider (#52239)``
+   * ``Fix compatibility test for Open Lineage (#51931)``
+   * ``Fix failing openlineage test (#51928)``
+
 2.4.0
 .....
 

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/docs/index.rst

@@ -81,7 +81,7 @@ apache-airflow-providers-openlineage package
 `OpenLineage <https://openlineage.io/>`__
 
 
-Release: 2.4.0
+Release: 2.5.0
 
 Provider package
 ----------------
@@ -108,8 +108,8 @@ PIP package                                 Version required
 ``apache-airflow-providers-common-sql``     ``>=1.20.0``
 ``apache-airflow-providers-common-compat``  ``>=1.4.0``
 ``attrs``                                   ``>=22.2``
-``openlineage-integration-common``          ``>=1.31.0``
-``openlineage-python``                      ``>=1.31.0``
+``openlineage-integration-common``          ``>=1.34.0``
+``openlineage-python``                      ``>=1.34.0``
 ==========================================  ==================
 
 Cross provider package dependencies
@@ -138,5 +138,5 @@ Downloading official packages
 You can download officially released packages and verify their checksums and signatures from the
 `Official Apache Download site <https://downloads.apache.org/airflow/providers/>`_
 
-* `The apache-airflow-providers-openlineage 2.4.0 sdist package <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.4.0.tar.gz>`_ (`asc <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.4.0.tar.gz.asc>`__, `sha512 <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.4.0.tar.gz.sha512>`__)
-* `The apache-airflow-providers-openlineage 2.4.0 wheel package <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.4.0-py3-none-any.whl>`_ (`asc <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.4.0-py3-none-any.whl.asc>`__, `sha512 <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.4.0-py3-none-any.whl.sha512>`__)
+* `The apache-airflow-providers-openlineage 2.5.0 sdist package <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.5.0.tar.gz>`_ (`asc <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.5.0.tar.gz.asc>`__, `sha512 <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.5.0.tar.gz.sha512>`__)
+* `The apache-airflow-providers-openlineage 2.5.0 wheel package <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.5.0-py3-none-any.whl>`_ (`asc <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.5.0-py3-none-any.whl.asc>`__, `sha512 <https://downloads.apache.org/airflow/providers/apache_airflow_providers_openlineage-2.5.0-py3-none-any.whl.sha512>`__)

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/provider.yaml

@@ -22,12 +22,13 @@ description: |
   `OpenLineage <https://openlineage.io/>`__
 
 state: ready
-source-date-epoch: 1749896859
+source-date-epoch: 1751473929
 # Note that those versions are maintained by release manager - do not update them manually
 # with the exception of case where other provider in sources has >= new provider version.
 # In such case adding >= NEW_VERSION and bumping to NEW_VERSION in a provider have
 # to be done in the same PR
 versions:
+  - 2.5.0
   - 2.4.0
   - 2.3.0
   - 2.2.0

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/pyproject.toml

@@ -25,7 +25,7 @@ build-backend = "flit_core.buildapi"
 
 [project]
 name = "apache-airflow-providers-openlineage"
-version = "2.4.0"
+version = "2.5.0rc1"
 description = "Provider package apache-airflow-providers-openlineage for Apache Airflow"
 readme = "README.rst"
 authors = [
@@ -44,25 +44,24 @@ classifiers = [
     "Framework :: Apache Airflow",
     "Framework :: Apache Airflow :: Provider",
     "License :: OSI Approved :: Apache Software License",
-    "Programming Language :: Python :: 3.9",
     "Programming Language :: Python :: 3.10",
     "Programming Language :: Python :: 3.11",
     "Programming Language :: Python :: 3.12",
     "Topic :: System :: Monitoring",
 ]
-requires-python = "~=3.9"
+requires-python = "~=3.10"
 
 # The dependencies should be modified in place in the generated file.
 # Any change in the dependencies is preserved when the file is regenerated
 # Make sure to run ``breeze static-checks --type update-providers-dependencies --all-files``
 # After you modify the dependencies, and rebuild your Breeze CI image with ``breeze ci-image build``
 dependencies = [
-    "apache-airflow>=2.10.0",
-    "apache-airflow-providers-common-sql>=1.20.0",
-    "apache-airflow-providers-common-compat>=1.4.0",
+    "apache-airflow>=2.10.0rc1",
+    "apache-airflow-providers-common-sql>=1.20.0rc1",
+    "apache-airflow-providers-common-compat>=1.4.0rc1",
     "attrs>=22.2",
-    "openlineage-integration-common>=1.31.0",
-    "openlineage-python>=1.31.0",
+    "openlineage-integration-common>=1.34.0",
+    "openlineage-python>=1.34.0",
 ]
 
 [dependency-groups]
@@ -105,8 +104,8 @@ apache-airflow-providers-common-sql = {workspace = true}
 apache-airflow-providers-standard = {workspace = true}
 
 [project.urls]
-"Documentation" = "https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.4.0"
-"Changelog" = "https://airflow.apache.org/docs/apache-airflow-providers-openlineage/2.4.0/changelog.html"
+"Documentation" = "https://airflow.staged.apache.org/docs/apache-airflow-providers-openlineage/2.5.0"
+"Changelog" = "https://airflow.staged.apache.org/docs/apache-airflow-providers-openlineage/2.5.0/changelog.html"
 "Bug Tracker" = "https://github.com/apache/airflow/issues"
 "Source Code" = "https://github.com/apache/airflow"
 "Slack Chat" = "https://s.apache.org/airflow-slack"

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/src/airflow/providers/openlineage/__init__.py

@@ -29,7 +29,7 @@ from airflow import __version__ as airflow_version
 
 __all__ = ["__version__"]
 
-__version__ = "2.4.0"
+__version__ = "2.5.0"
 
 if packaging.version.parse(packaging.version.parse(airflow_version).base_version) < packaging.version.parse(
     "2.10.0"

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/src/airflow/providers/openlineage/extractors/base.py

@@ -19,7 +19,7 @@ from __future__ import annotations
 
 import warnings
 from abc import ABC, abstractmethod
-from typing import Generic, TypeVar, Union
+from typing import Generic, TypeVar
 
 from attrs import Factory, define
 from openlineage.client.event_v2 import Dataset as OLDataset
@@ -33,7 +33,7 @@ from airflow.utils.log.logging_mixin import LoggingMixin
 
 # this is not to break static checks compatibility with v1 OpenLineage facet classes
 DatasetSubclass = TypeVar("DatasetSubclass", bound=OLDataset)
-BaseFacetSubclass = TypeVar("BaseFacetSubclass", bound=Union[BaseFacet_V1, RunFacet, JobFacet])
+BaseFacetSubclass = TypeVar("BaseFacetSubclass", bound=BaseFacet_V1 | RunFacet | JobFacet)
 
 OL_METHOD_NAME_START = "get_openlineage_facets_on_start"
 OL_METHOD_NAME_COMPLETE = "get_openlineage_facets_on_complete"

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/src/airflow/providers/openlineage/extractors/python.py

@@ -18,7 +18,7 @@
 from __future__ import annotations
 
 import inspect
-from typing import Callable
+from collections.abc import Callable
 
 from openlineage.client.facet_v2 import source_code_job
 

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/src/airflow/providers/openlineage/plugins/listener.py

@@ -147,7 +147,7 @@ class OpenLineageListener:
                 "Skipping OpenLineage event emission for task `%s` "
                 "due to lack of explicit lineage enablement for task or DAG while "
                 "[openlineage] selective_enable is on.",
-                task.task_id,
+                task_instance.task_id,
             )
             return
 
@@ -170,14 +170,14 @@ class OpenLineageListener:
                 clear_number = dagrun.clear_number
 
             parent_run_id = self.adapter.build_dag_run_id(
-                dag_id=dag.dag_id,
+                dag_id=task_instance.dag_id,
                 logical_date=date,
                 clear_number=clear_number,
             )
 
             task_uuid = self.adapter.build_task_instance_run_id(
-                dag_id=dag.dag_id,
-                task_id=task.task_id,
+                dag_id=task_instance.dag_id,
+                task_id=task_instance.task_id,
                 try_number=task_instance.try_number,
                 logical_date=date,
                 map_index=task_instance.map_index,
@@ -199,7 +199,7 @@ class OpenLineageListener:
 
             redacted_event = self.adapter.start_task(
                 run_id=task_uuid,
-                job_name=get_job_name(task),
+                job_name=get_job_name(task_instance),
                 job_description=dag.description,
                 event_time=start_date.isoformat(),
                 nominal_start_time=data_interval_start,
@@ -278,7 +278,7 @@ class OpenLineageListener:
                 "Skipping OpenLineage event emission for task `%s` "
                 "due to lack of explicit lineage enablement for task or DAG while "
                 "[openlineage] selective_enable is on.",
-                task.task_id,
+                task_instance.task_id,
             )
             return
 
@@ -289,14 +289,14 @@ class OpenLineageListener:
                 date = dagrun.run_after
 
             parent_run_id = self.adapter.build_dag_run_id(
-                dag_id=dag.dag_id,
+                dag_id=task_instance.dag_id,
                 logical_date=date,
                 clear_number=dagrun.clear_number,
             )
 
             task_uuid = self.adapter.build_task_instance_run_id(
-                dag_id=dag.dag_id,
-                task_id=task.task_id,
+                dag_id=task_instance.dag_id,
+                task_id=task_instance.task_id,
                 try_number=task_instance.try_number,
                 logical_date=date,
                 map_index=task_instance.map_index,
@@ -321,7 +321,7 @@ class OpenLineageListener:
 
             redacted_event = self.adapter.complete_task(
                 run_id=task_uuid,
-                job_name=get_job_name(task),
+                job_name=get_job_name(task_instance),
                 end_time=end_date.isoformat(),
                 task=task_metadata,
                 # If task owner is default ("airflow"), use DAG owner instead that may have more details
@@ -409,7 +409,7 @@ class OpenLineageListener:
                 "Skipping OpenLineage event emission for task `%s` "
                 "due to lack of explicit lineage enablement for task or DAG while "
                 "[openlineage] selective_enable is on.",
-                task.task_id,
+                task_instance.task_id,
             )
             return
 
@@ -420,14 +420,14 @@ class OpenLineageListener:
                 date = dagrun.run_after
 
             parent_run_id = self.adapter.build_dag_run_id(
-                dag_id=dag.dag_id,
+                dag_id=task_instance.dag_id,
                 logical_date=date,
                 clear_number=dagrun.clear_number,
             )
 
             task_uuid = self.adapter.build_task_instance_run_id(
-                dag_id=dag.dag_id,
-                task_id=task.task_id,
+                dag_id=task_instance.dag_id,
+                task_id=task_instance.task_id,
                 try_number=task_instance.try_number,
                 logical_date=date,
                 map_index=task_instance.map_index,
@@ -452,7 +452,7 @@ class OpenLineageListener:
 
             redacted_event = self.adapter.fail_task(
                 run_id=task_uuid,
-                job_name=get_job_name(task),
+                job_name=get_job_name(task_instance),
                 end_time=end_date.isoformat(),
                 task=task_metadata,
                 error=error,
@@ -489,13 +489,13 @@ class OpenLineageListener:
         def on_state_change():
             date = dagrun.logical_date or dagrun.run_after
             parent_run_id = self.adapter.build_dag_run_id(
-                dag_id=dagrun.dag_id,
+                dag_id=ti.dag_id,
                 logical_date=date,
                 clear_number=dagrun.clear_number,
             )
 
             task_uuid = self.adapter.build_task_instance_run_id(
-                dag_id=dagrun.dag_id,
+                dag_id=ti.dag_id,
                 task_id=ti.task_id,
                 try_number=ti.try_number,
                 logical_date=date,
@@ -507,6 +507,10 @@ class OpenLineageListener:
                 "job_name": get_job_name(ti),
                 "end_time": end_date.isoformat(),
                 "task": OperatorLineage(),
+                "nominal_start_time": None,
+                "nominal_end_time": None,
+                "tags": None,
+                "owners": None,
                 "run_facets": {
                     **get_task_parent_run_facet(parent_run_id=parent_run_id, parent_job_name=ti.dag_id),
                     **get_airflow_debug_facet(),

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/src/airflow/providers/openlineage/plugins/macros.py

@@ -117,10 +117,22 @@ def lineage_root_run_id(task_instance: TaskInstance):
     return OpenLineageAdapter.build_dag_run_id(
         dag_id=task_instance.dag_id,
         logical_date=_get_logical_date(task_instance),
-        clear_number=task_instance.dag_run.clear_number,
+        clear_number=_get_dag_run_clear_number(task_instance),
     )
 
 
+def _get_dag_run_clear_number(task_instance: TaskInstance):
+    # todo: remove when min airflow version >= 3.0
+    if AIRFLOW_V_3_0_PLUS:
+        context = task_instance.get_template_context()
+        if hasattr(task_instance, "dag_run"):
+            dag_run = task_instance.dag_run
+        else:
+            dag_run = context["dag_run"]
+        return dag_run.clear_number
+    return task_instance.dag_run.clear_number
+
+
 def _get_logical_date(task_instance):
     # todo: remove when min airflow version >= 3.0
     if AIRFLOW_V_3_0_PLUS:

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/src/airflow/providers/openlineage/sqlparser.py

@@ -17,7 +17,8 @@
 from __future__ import annotations
 
 import logging
-from typing import TYPE_CHECKING, Callable, TypedDict
+from collections.abc import Callable
+from typing import TYPE_CHECKING, TypedDict
 
 import sqlparse
 from attrs import define
@@ -38,8 +39,8 @@ if TYPE_CHECKING:
     from openlineage.client.facet_v2 import JobFacet, RunFacet
     from sqlalchemy.engine import Engine
 
-    from airflow.hooks.base import BaseHook
     from airflow.providers.common.sql.hooks.sql import DbApiHook
+    from airflow.sdk import BaseHook
 
 log = logging.getLogger(__name__)
 

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/src/airflow/providers/openlineage/utils/sql.py

@@ -20,7 +20,7 @@ import logging
 from collections import defaultdict
 from contextlib import closing
 from enum import IntEnum
-from typing import TYPE_CHECKING, Optional
+from typing import TYPE_CHECKING
 
 from attrs import define
 from openlineage.client.event_v2 import Dataset
@@ -31,7 +31,7 @@ if TYPE_CHECKING:
     from sqlalchemy.engine import Engine
     from sqlalchemy.sql import ClauseElement
 
-    from airflow.hooks.base import BaseHook
+    from airflow.sdk import BaseHook
 
 
 log = logging.getLogger(__name__)
@@ -50,7 +50,7 @@ class ColumnIndex(IntEnum):
     DATABASE = 5
 
 
-TablesHierarchy = dict[Optional[str], dict[Optional[str], list[str]]]
+TablesHierarchy = dict[str | None, dict[str | None, list[str]]]
 
 
 @define

{apache_airflow_providers_openlineage-2.4.0 → apache_airflow_providers_openlineage-2.5.0rc1}/src/airflow/providers/openlineage/utils/utils.py

@@ -20,10 +20,11 @@ from __future__ import annotations
 import datetime
 import json
 import logging
+from collections.abc import Callable
 from contextlib import suppress
 from functools import wraps
 from importlib import metadata
-from typing import TYPE_CHECKING, Any, Callable
+from typing import TYPE_CHECKING, Any
 
 import attrs
 from openlineage.client.facet_v2 import parent_run
@@ -32,7 +33,7 @@ from openlineage.client.utils import RedactMixin
 from airflow import __version__ as AIRFLOW_VERSION
 
 # TODO: move this maybe to Airflow's logic?
-from airflow.models import BaseOperator, DagRun, TaskReschedule
+from airflow.models import DagRun, TaskReschedule
 from airflow.providers.openlineage import (
     __version__ as OPENLINEAGE_PROVIDER_VERSION,
     conf,
@@ -52,18 +53,17 @@ from airflow.providers.openlineage.utils.selective_enable import (
     is_task_lineage_enabled,
 )
 from airflow.providers.openlineage.version_compat import AIRFLOW_V_3_0_PLUS
-from airflow.sensors.base import BaseSensorOperator
 from airflow.serialization.serialized_objects import SerializedBaseOperator
 from airflow.utils.module_loading import import_string
 
+if AIRFLOW_V_3_0_PLUS:
+    from airflow.sdk import BaseSensorOperator
+else:
+    from airflow.sensors.base import BaseSensorOperator  # type: ignore[no-redef]
+
 if not AIRFLOW_V_3_0_PLUS:
     from airflow.utils.session import NEW_SESSION, provide_session
 
-try:
-    from airflow.sdk import BaseOperator as SdkBaseOperator
-except ImportError:
-    SdkBaseOperator = BaseOperator  # type: ignore[misc]
-
 if TYPE_CHECKING:
     from openlineage.client.event_v2 import Dataset as OpenLineageDataset
     from openlineage.client.facet_v2 import RunFacet, processing_engine_run
@@ -71,6 +71,7 @@ if TYPE_CHECKING:
     from airflow.models import TaskInstance
     from airflow.providers.common.compat.assets import Asset
     from airflow.sdk import DAG
+    from airflow.sdk.bases.operator import BaseOperator
     from airflow.sdk.definitions.mappedoperator import MappedOperator
     from airflow.sdk.execution_time.secrets_masker import (
         Redactable,
@@ -78,13 +79,15 @@ if TYPE_CHECKING:
         SecretsMasker,
         should_hide_value_for_key,
     )
+    from airflow.sdk.execution_time.task_runner import RuntimeTaskInstance
     from airflow.utils.state import DagRunState, TaskInstanceState
 else:
     try:
         from airflow.sdk import DAG
+        from airflow.sdk.bases.operator import BaseOperator
         from airflow.sdk.definitions.mappedoperator import MappedOperator
     except ImportError:
-        from airflow.models import DAG, MappedOperator
+        from airflow.models import DAG, BaseOperator, MappedOperator
 
     try:
         from airflow.providers.common.compat.assets import Asset
@@ -119,13 +122,39 @@ def try_import_from_string(string: str) -> Any:
         return import_string(string)
 
 
-def get_operator_class(task: BaseOperator | SdkBaseOperator) -> type:
+def get_operator_class(task: BaseOperator) -> type:
     if task.__class__.__name__ in ("DecoratedMappedOperator", "MappedOperator"):
         return task.operator_class
     return task.__class__
 
 
-def get_job_name(task: TaskInstance) -> str:
+def get_operator_provider_version(operator: BaseOperator | MappedOperator) -> str | None:
+    """Get the provider package version for the given operator."""
+    try:
+        class_path = get_fully_qualified_class_name(operator)
+
+        if not class_path.startswith("airflow.providers."):
+            return None
+
+        from airflow.providers_manager import ProvidersManager
+
+        providers_manager = ProvidersManager()
+
+        for package_name, provider_info in providers_manager.providers.items():
+            if package_name.startswith("apache-airflow-providers-"):
+                provider_module_path = package_name.replace(
+                    "apache-airflow-providers-", "airflow.providers."
+                ).replace("-", ".")
+                if class_path.startswith(provider_module_path + "."):
+                    return provider_info.version
+
+        return None
+
+    except Exception:
+        return None
+
+
+def get_job_name(task: TaskInstance | RuntimeTaskInstance) -> str:
     return f"{task.dag_id}.{task.task_id}"
 
 
@@ -203,7 +232,7 @@ def get_user_provided_run_facets(ti: TaskInstance, ti_state: TaskInstanceState)
     return custom_facets
 
 
-def get_fully_qualified_class_name(operator: BaseOperator | MappedOperator | SdkBaseOperator) -> str:
+def get_fully_qualified_class_name(operator: BaseOperator | MappedOperator) -> str:
     if isinstance(operator, (MappedOperator, SerializedBaseOperator)):
         # as in airflow.api_connexion.schemas.common_schema.ClassReferenceSchema
         return operator._task_module + "." + operator._task_type  # type: ignore
@@ -211,17 +240,17 @@ def get_fully_qualified_class_name(operator: BaseOperator | MappedOperator | Sdk
     return op_class.__module__ + "." + op_class.__name__
 
 
-def is_operator_disabled(operator: BaseOperator | MappedOperator | SdkBaseOperator) -> bool:
+def is_operator_disabled(operator: BaseOperator | MappedOperator) -> bool:
     return get_fully_qualified_class_name(operator) in conf.disabled_operators()
 
 
-def is_selective_lineage_enabled(obj: DAG | BaseOperator | MappedOperator | SdkBaseOperator) -> bool:
+def is_selective_lineage_enabled(obj: DAG | BaseOperator | MappedOperator) -> bool:
     """If selective enable is active check if DAG or Task is enabled to emit events."""
     if not conf.selective_enable():
         return True
     if isinstance(obj, DAG):
         return is_dag_lineage_enabled(obj)
-    if isinstance(obj, (BaseOperator, MappedOperator, SdkBaseOperator)):
+    if isinstance(obj, (BaseOperator, MappedOperator)):
         return is_task_lineage_enabled(obj)
     raise TypeError("is_selective_lineage_enabled can only be used on DAG or Operator objects")
 
@@ -508,6 +537,7 @@ class TaskInfo(InfoJsonEncodable):
         ),
         "inlets": lambda task: [AssetInfo(i) for i in task.inlets if isinstance(i, Asset)],
         "outlets": lambda task: [AssetInfo(o) for o in task.outlets if isinstance(o, Asset)],
+        "operator_provider_version": lambda task: get_operator_provider_version(task),
     }