apache-airflow-providers-apache-hive 7.0.1rc1__tar.gz → 8.0.0__tar.gz

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: apache-airflow-providers-apache-hive
-Version: 7.0.1rc1
+Version: 8.0.0
 Summary: Provider package apache-airflow-providers-apache-hive for Apache Airflow
 Keywords: airflow-provider,apache.hive,airflow,integration
 Author-email: Apache Software Foundation <dev@airflow.apache.org>
@@ -19,9 +19,10 @@ Classifier: Programming Language :: Python :: 3.8
 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-providers-common-sql>=1.3.1.dev0
-Requires-Dist: apache-airflow>=2.6.0.dev0
+Requires-Dist: apache-airflow-providers-common-sql>=1.3.1
+Requires-Dist: apache-airflow>=2.6.0
 Requires-Dist: hmsclient>=0.1.0
 Requires-Dist: pandas>=1.2.5,<2.2
 Requires-Dist: pyhive[hive_pure_sasl]>=0.7.0
@@ -34,8 +35,8 @@ Requires-Dist: apache-airflow-providers-presto ; extra == "presto"
 Requires-Dist: apache-airflow-providers-samba ; extra == "samba"
 Requires-Dist: apache-airflow-providers-vertica ; extra == "vertica"
 Project-URL: Bug Tracker, https://github.com/apache/airflow/issues
-Project-URL: Changelog, https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/7.0.1/changelog.html
-Project-URL: Documentation, https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/7.0.1
+Project-URL: Changelog, https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/8.0.0/changelog.html
+Project-URL: Documentation, https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/8.0.0
 Project-URL: Slack Chat, https://s.apache.org/airflow-slack
 Project-URL: Source Code, https://github.com/apache/airflow
 Project-URL: Twitter, https://twitter.com/ApacheAirflow
@@ -92,7 +93,7 @@ Provides-Extra: vertica
 
 Package ``apache-airflow-providers-apache-hive``
 
-Release: ``7.0.1.rc1``
+Release: ``8.0.0``
 
 
 `Apache Hive <https://hive.apache.org/>`__
@@ -105,7 +106,7 @@ This is a provider package for ``apache.hive`` provider. All classes for this pr
 are in ``airflow.providers.apache.hive`` python package.
 
 You can find package information and changelog for the provider
-in the `documentation <https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/7.0.1/>`_.
+in the `documentation <https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/8.0.0/>`_.
 
 Installation
 ------------
@@ -114,7 +115,7 @@ 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-apache-hive``
 
-The package supports the following python versions: 3.8,3.9,3.10,3.11
+The package supports the following python versions: 3.8,3.9,3.10,3.11,3.12
 
 Requirements
 ------------
@@ -156,4 +157,4 @@ Dependent package
 ======================================================================================================================  ===================
 
 The changelog for the provider package can be found in the
-`changelog <https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/7.0.1/changelog.html>`_.
+`changelog <https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/8.0.0/changelog.html>`_.

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/README.rst

@@ -42,7 +42,7 @@
 
 Package ``apache-airflow-providers-apache-hive``
 
-Release: ``7.0.1.rc1``
+Release: ``8.0.0``
 
 
 `Apache Hive <https://hive.apache.org/>`__
@@ -55,7 +55,7 @@ This is a provider package for ``apache.hive`` provider. All classes for this pr
 are in ``airflow.providers.apache.hive`` python package.
 
 You can find package information and changelog for the provider
-in the `documentation <https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/7.0.1/>`_.
+in the `documentation <https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/8.0.0/>`_.
 
 Installation
 ------------
@@ -64,7 +64,7 @@ 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-apache-hive``
 
-The package supports the following python versions: 3.8,3.9,3.10,3.11
+The package supports the following python versions: 3.8,3.9,3.10,3.11,3.12
 
 Requirements
 ------------
@@ -106,4 +106,4 @@ Dependent package
 ======================================================================================================================  ===================
 
 The changelog for the provider package can be found in the
-`changelog <https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/7.0.1/changelog.html>`_.
+`changelog <https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/8.0.0/changelog.html>`_.

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/__init__.py

@@ -27,7 +27,7 @@ import packaging.version
 
 __all__ = ["__version__"]
 
-__version__ = "7.0.1"
+__version__ = "8.0.0"
 
 try:
     from airflow import __version__ as airflow_version

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/get_provider_info.py

@@ -30,6 +30,7 @@ def get_provider_info():
         "state": "ready",
         "source-date-epoch": 1709554960,
         "versions": [
+            "8.0.0",
             "7.0.1",
             "7.0.0",
             "6.4.2",

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/hooks/hive.py

@@ -23,12 +23,9 @@ import re
 import socket
 import subprocess
 import time
-import warnings
 from tempfile import NamedTemporaryFile, TemporaryDirectory
 from typing import TYPE_CHECKING, Any, Iterable, Mapping
 
-from airflow.exceptions import AirflowProviderDeprecationWarning
-
 if TYPE_CHECKING:
     import pandas as pd
 
@@ -116,6 +113,7 @@ class HiveCliHook(BaseHook):
         self.mapred_queue_priority = mapred_queue_priority
         self.mapred_job_name = mapred_job_name
         self.proxy_user = proxy_user
+        self.high_availability = self.conn.extra_dejson.get("high_availability", False)
 
     @classmethod
     def get_connection_form_widgets(cls) -> dict[str, Any]:
@@ -125,11 +123,12 @@ class HiveCliHook(BaseHook):
         from wtforms import BooleanField, StringField
 
         return {
-            "use_beeline": BooleanField(lazy_gettext("Use Beeline"), default=False),
+            "use_beeline": BooleanField(lazy_gettext("Use Beeline"), default=True),
             "proxy_user": StringField(lazy_gettext("Proxy User"), widget=BS3TextFieldWidget(), default=""),
             "principal": StringField(
                 lazy_gettext("Principal"), widget=BS3TextFieldWidget(), default="hive/_HOST@EXAMPLE.COM"
             ),
+            "high_availability": BooleanField(lazy_gettext("High Availability"), default=False),
         }
 
     @classmethod
@@ -159,7 +158,12 @@ class HiveCliHook(BaseHook):
         if self.use_beeline:
             hive_bin = "beeline"
             self._validate_beeline_parameters(conn)
-            jdbc_url = f"jdbc:hive2://{conn.host}:{conn.port}/{conn.schema}"
+            if self.high_availability:
+                jdbc_url = f"jdbc:hive2://{conn.host}/{conn.schema}"
+                self.log.info("High Availability set, setting JDBC url as %s", jdbc_url)
+            else:
+                jdbc_url = f"jdbc:hive2://{conn.host}:{conn.port}/{conn.schema}"
+                self.log.info("High Availability not set, setting JDBC url as %s", jdbc_url)
             if conf.get("core", "security") == "kerberos":
                 template = conn.extra_dejson.get("principal", "hive/_HOST@EXAMPLE.COM")
                 if "_HOST" in template:
@@ -170,6 +174,10 @@ class HiveCliHook(BaseHook):
                 if ";" in proxy_user:
                     raise RuntimeError("The proxy_user should not contain the ';' character")
                 jdbc_url += f";principal={template};{proxy_user}"
+                if self.high_availability:
+                    if not jdbc_url.endswith(";"):
+                        jdbc_url += ";"
+                    jdbc_url += "serviceDiscoveryMode=zooKeeper;ssl=true;zooKeeperNamespace=hiveserver2"
             elif self.auth:
                 jdbc_url += ";auth=" + self.auth
 
@@ -186,18 +194,28 @@ class HiveCliHook(BaseHook):
         return [hive_bin, *cmd_extra, *hive_params_list]
 
     def _validate_beeline_parameters(self, conn):
-        if ":" in conn.host or "/" in conn.host or ";" in conn.host:
-            raise Exception(
+        if self.high_availability:
+            if ";" in conn.schema:
+                raise ValueError(
+                    f"The schema used in beeline command ({conn.schema}) should not contain ';' character)"
+                )
+            return
+        elif ":" in conn.host or "/" in conn.host or ";" in conn.host:
+            raise ValueError(
                 f"The host used in beeline command ({conn.host}) should not contain ':/;' characters)"
             )
         try:
             int_port = int(conn.port)
             if not 0 < int_port <= 65535:
-                raise Exception(f"The port used in beeline command ({conn.port}) should be in range 0-65535)")
+                raise ValueError(
+                    f"The port used in beeline command ({conn.port}) should be in range 0-65535)"
+                )
         except (ValueError, TypeError) as e:
-            raise Exception(f"The port used in beeline command ({conn.port}) should be a valid integer: {e})")
+            raise ValueError(
+                f"The port used in beeline command ({conn.port}) should be a valid integer: {e})"
+            )
         if ";" in conn.schema:
-            raise Exception(
+            raise ValueError(
                 f"The schema used in beeline command ({conn.schema}) should not contain ';' character)"
             )
 
@@ -518,11 +536,13 @@ class HiveMetastoreHook(BaseHook):
     def __getstate__(self) -> dict[str, Any]:
         # This is for pickling to work despite the thrift hive client not
         # being picklable
+        """Serialize object and omit non-serializable attributes."""
         state = dict(self.__dict__)
         del state["metastore"]
         return state
 
     def __setstate__(self, d: dict[str, Any]) -> None:
+        """Deserialize object and restore non-serializable attributes."""
         self.__dict__.update(d)
         self.__dict__["metastore"] = self.get_metastore_client()
 
@@ -538,15 +558,6 @@ class HiveMetastoreHook(BaseHook):
         if not host:
             raise AirflowException("Failed to locate the valid server.")
 
-        if "authMechanism" in conn.extra_dejson:
-            warnings.warn(
-                "The 'authMechanism' option is deprecated. Please use 'auth_mechanism'.",
-                AirflowProviderDeprecationWarning,
-                stacklevel=2,
-            )
-            conn.extra_dejson["auth_mechanism"] = conn.extra_dejson["authMechanism"]
-            del conn.extra_dejson["authMechanism"]
-
         auth_mechanism = conn.extra_dejson.get("auth_mechanism", "NOSASL")
 
         if conf.get("core", "security") == "kerberos":
@@ -849,15 +860,6 @@ class HiveServer2Hook(DbApiHook):
 
         db = self.get_connection(self.hiveserver2_conn_id)  # type: ignore
 
-        if "authMechanism" in db.extra_dejson:
-            warnings.warn(
-                "The 'authMechanism' option is deprecated. Please use 'auth_mechanism'.",
-                AirflowProviderDeprecationWarning,
-                stacklevel=2,
-            )
-            db.extra_dejson["auth_mechanism"] = db.extra_dejson["authMechanism"]
-            del db.extra_dejson["authMechanism"]
-
         auth_mechanism = db.extra_dejson.get("auth_mechanism", "NONE")
         if auth_mechanism == "NONE" and db.login is None:
             # we need to give a username

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/operators/hive.py

@@ -22,10 +22,7 @@ import re
 from functools import cached_property
 from typing import TYPE_CHECKING, Any, Sequence
 
-from deprecated.classic import deprecated
-
 from airflow.configuration import conf
-from airflow.exceptions import AirflowProviderDeprecationWarning
 from airflow.models import BaseOperator
 from airflow.providers.apache.hive.hooks.hive import HiveCliHook
 from airflow.utils import operator_helpers
@@ -131,11 +128,6 @@ class HiveOperator(BaseOperator):
             proxy_user=self.proxy_user,
         )
 
-    @deprecated(reason="use `hook` property instead.", category=AirflowProviderDeprecationWarning)
-    def get_hook(self) -> HiveCliHook:
-        """Get Hive cli hook."""
-        return self.hook
-
     def prepare_template(self) -> None:
         if self.hiveconf_jinja_translate:
             self.hql = re.sub(r"(\$\{(hiveconf:)?([ a-zA-Z0-9_]*)\})", r"{{ \g<3> }}", self.hql)

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/operators/hive_stats.py

@@ -18,7 +18,6 @@
 from __future__ import annotations
 
 import json
-import warnings
 from typing import TYPE_CHECKING, Any, Callable, Sequence
 
 from airflow.exceptions import AirflowException
@@ -80,15 +79,6 @@ class HiveStatsCollectionOperator(BaseOperator):
         dttm: str = "{{ logical_date.isoformat() }}",
         **kwargs: Any,
     ) -> None:
-        if "col_blacklist" in kwargs:
-            warnings.warn(
-                f"col_blacklist kwarg passed to {self.__class__.__name__} "
-                f"(task_id: {kwargs.get('task_id')}) is deprecated, "
-                f"please rename it to excluded_columns instead",
-                category=FutureWarning,
-                stacklevel=2,
-            )
-            excluded_columns = kwargs.pop("col_blacklist")
         super().__init__(**kwargs)
         self.table = table
         self.partition = partition

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/transfers/hive_to_mysql.py

@@ -16,6 +16,7 @@
 # specific language governing permissions and limitations
 # under the License.
 """This module contains an operator to move data from Hive to MySQL."""
+
 from __future__ import annotations
 
 from tempfile import NamedTemporaryFile

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/transfers/hive_to_samba.py

@@ -16,6 +16,7 @@
 # specific language governing permissions and limitations
 # under the License.
 """This module contains an operator to move data from Hive to Samba."""
+
 from __future__ import annotations
 
 from tempfile import NamedTemporaryFile

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/transfers/mssql_to_hive.py

@@ -16,6 +16,7 @@
 # specific language governing permissions and limitations
 # under the License.
 """This module contains an operator to move data from MSSQL to Hive."""
+
 from __future__ import annotations
 
 import csv

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/transfers/mysql_to_hive.py

@@ -16,6 +16,7 @@
 # specific language governing permissions and limitations
 # under the License.
 """This module contains an operator to move data from MySQL to Hive."""
+
 from __future__ import annotations
 
 import csv

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/transfers/s3_to_hive.py

@@ -16,6 +16,7 @@
 # specific language governing permissions and limitations
 # under the License.
 """This module contains an operator to move data from an S3 bucket to Hive."""
+
 from __future__ import annotations
 
 import bz2

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/airflow/providers/apache/hive/transfers/vertica_to_hive.py

@@ -16,6 +16,7 @@
 # specific language governing permissions and limitations
 # under the License.
 """This module contains an operator to move data from Vertica to Hive."""
+
 from __future__ import annotations
 
 import csv

{apache_airflow_providers_apache_hive-7.0.1rc1 → apache_airflow_providers_apache_hive-8.0.0}/pyproject.toml

@@ -28,7 +28,7 @@ build-backend = "flit_core.buildapi"
 
 [project]
 name = "apache-airflow-providers-apache-hive"
-version = "7.0.1.rc1"
+version = "8.0.0"
 description = "Provider package apache-airflow-providers-apache-hive for Apache Airflow"
 readme = "README.rst"
 authors = [
@@ -51,12 +51,13 @@ classifiers = [
     "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 = "~=3.8"
 dependencies = [
-    "apache-airflow-providers-common-sql>=1.3.1.dev0",
-    "apache-airflow>=2.6.0.dev0",
+    "apache-airflow-providers-common-sql>=1.3.1",
+    "apache-airflow>=2.6.0",
     "hmsclient>=0.1.0",
     "pandas>=1.2.5,<2.2",
     "pyhive[hive_pure_sasl]>=0.7.0",
@@ -64,8 +65,8 @@ dependencies = [
 ]
 
 [project.urls]
-"Documentation" = "https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/7.0.1"
-"Changelog" = "https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/7.0.1/changelog.html"
+"Documentation" = "https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/8.0.0"
+"Changelog" = "https://airflow.apache.org/docs/apache-airflow-providers-apache-hive/8.0.0/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"