Merge branch 'main' into feat/support-dbt-freshness-check-test

ingestion/src/metadata/data_quality/builders/i_validator_builder.py → ingestion/src/metadata/data_quality/builders/validator_builder.py

@@ -14,8 +14,8 @@
 Validators are test classes (e.g. columnValuesToBeBetween, etc.)
 """
 
-from abc import ABC, abstractmethod
 from datetime import datetime, timezone
+from enum import Enum
 from typing import TYPE_CHECKING, Set, Type, Union
 
 from metadata.data_quality.validations.base_test_handler import BaseTestValidator
@@ -25,48 +25,67 @@
 from metadata.generated.schema.tests.testCase import TestCase, TestCaseParameterValue
 from metadata.generated.schema.type.basic import Timestamp
 from metadata.profiler.processor.runner import QueryRunner
-from metadata.utils import importer
+from metadata.utils.importer import import_test_case_class
 
 if TYPE_CHECKING:
     from pandas import DataFrame
 
 
-class IValidatorBuilder(ABC):
-    """Interface for validator builders"""
+class TestCaseImporter:
+    def import_test_case_validator(
+        self,
+        test_type: str,
+        runner_type: str,
+        test_definition: str,
+    ) -> Type[BaseTestValidator]:
+        return import_test_case_class(test_type, runner_type, test_definition)
 
-    @property
-    def test_case(self):
-        """Return the test case object"""
-        return self._test_case
 
-    @property
-    def validator(self):
-        """Return the validator object"""
-        return self._validator
+class SourceType(Enum):
+    PANDAS = "pandas"
+    SQL = "sqlalchemy"
+
+
+class ValidatorBuilder(TestCaseImporter):
+    """Interface for validator builders"""
 
     def __init__(
         self,
         runner: Union[QueryRunner, "DataFrame"],
         test_case: TestCase,
+        source_type: SourceType,
         entity_type: str,
     ) -> None:
         """Builder object for SQA validators. This builder is used to create a validator object
 
         Args:
             runner (QueryRunner): The runner object
             test_case (TestCase): The test case object
+            source_type (SourceType): The source type
             entity_type (str): one of COLUMN or TABLE -- fetched from the test definition
         """
+        super().__init__()
         self._test_case = test_case
         self.runner = runner
-        # TODO this will be removed on https://github.com/open-metadata/OpenMetadata/pull/18716
-        self.validator_cls: Type[BaseTestValidator] = importer.import_test_case_class(
+        self.validator_cls: Type[
+            BaseTestValidator
+        ] = super().import_test_case_validator(
             entity_type,
-            self._get_source_type(),
-            self.test_case.testDefinition.fullyQualifiedName,  # type: ignore
+            source_type.value,
+            self.test_case.testDefinition.fullyQualifiedName,
         )
         self.reset()
 
+    @property
+    def test_case(self):
+        """Return the test case object"""
+        return self._test_case
+
+    @property
+    def validator(self):
+        """Return the validator object"""
+        return self._validator
+
     def set_runtime_params(self, runtime_params_setters: Set[RuntimeParameterSetter]):
         """Set the runtime parameters for the validator object
 
@@ -93,8 +112,3 @@ def reset(self):
                 int(datetime.now(tz=timezone.utc).timestamp() * 1000)
             ),
         )
-
-    @abstractmethod
-    def _get_source_type(self):
-        """Get the source type"""
-        raise NotImplementedError

ingestion/src/metadata/data_quality/interface/pandas/pandas_test_suite_interface.py

@@ -13,9 +13,10 @@
 Interfaces with database for all database engine
 supporting sqlalchemy abstraction layer
 """
-from metadata.data_quality.builders.i_validator_builder import IValidatorBuilder
-from metadata.data_quality.builders.pandas_validator_builder import (
-    PandasValidatorBuilder,
+
+from metadata.data_quality.builders.validator_builder import (
+    SourceType,
+    ValidatorBuilder,
 )
 from metadata.data_quality.interface.test_suite_interface import TestSuiteInterface
 from metadata.generated.schema.entity.data.table import Table
@@ -44,13 +45,14 @@ def __init__(
         ometa_client: OpenMetadata,
         sampler: SamplerInterface,
         table_entity: Table,
-        **__,
+        **kwargs,
     ):
         super().__init__(
             service_connection_config,
             ometa_client,
             sampler,
             table_entity,
+            **kwargs,
         )
 
         (
@@ -63,5 +65,10 @@ def __init__(
 
     def _get_validator_builder(
         self, test_case: TestCase, entity_type: str
-    ) -> IValidatorBuilder:
-        return PandasValidatorBuilder(self.dataset, test_case, entity_type)
+    ) -> ValidatorBuilder:
+        return self.validator_builder_class(
+            runner=self.dataset,
+            test_case=test_case,
+            entity_type=entity_type,
+            source_type=SourceType.PANDAS,
+        )

ingestion/src/metadata/data_quality/interface/sqlalchemy/sqa_test_suite_interface.py

@@ -19,8 +19,10 @@
 from sqlalchemy.orm import DeclarativeMeta
 from sqlalchemy.orm.util import AliasedClass
 
-from metadata.data_quality.builders.i_validator_builder import IValidatorBuilder
-from metadata.data_quality.builders.sqa_validator_builder import SQAValidatorBuilder
+from metadata.data_quality.builders.validator_builder import (
+    SourceType,
+    ValidatorBuilder,
+)
 from metadata.data_quality.interface.test_suite_interface import TestSuiteInterface
 from metadata.generated.schema.entity.data.table import Table
 from metadata.generated.schema.entity.services.databaseService import DatabaseConnection
@@ -51,14 +53,14 @@ def __init__(
         ometa_client: OpenMetadata,
         sampler: SamplerInterface,
         table_entity: Table = None,
+        **kwargs,
     ):
         super().__init__(
-            service_connection_config,
-            ometa_client,
-            sampler,
-            table_entity,
+            service_connection_config, ometa_client, sampler, table_entity, **kwargs
         )
+        self.source_type = SourceType.SQL
         self.create_session()
+
         (
             self.table_sample_query,
             self.table_sample_config,
@@ -109,5 +111,10 @@ def _create_runner(self) -> QueryRunner:
 
     def _get_validator_builder(
         self, test_case: TestCase, entity_type: str
-    ) -> IValidatorBuilder:
-        return SQAValidatorBuilder(self.runner, test_case, entity_type)
+    ) -> ValidatorBuilder:
+        return self.validator_builder_class(
+            runner=self.runner,
+            test_case=test_case,
+            entity_type=entity_type,
+            source_type=self.source_type,
+        )

ingestion/src/metadata/data_quality/interface/test_suite_interface.py

@@ -17,7 +17,7 @@
 from abc import ABC, abstractmethod
 from typing import Optional, Set, Type
 
-from metadata.data_quality.builders.i_validator_builder import IValidatorBuilder
+from metadata.data_quality.builders.validator_builder import ValidatorBuilder
 from metadata.data_quality.validations.base_test_handler import BaseTestValidator
 from metadata.data_quality.validations.runtime_param_setter.param_setter import (
     RuntimeParameterSetter,
@@ -48,12 +48,14 @@ def __init__(
         ometa_client: OpenMetadata,
         sampler: SamplerInterface,
         table_entity: Table,
+        validator_builder: Type[ValidatorBuilder],
     ):
         """Required attribute for the interface"""
         self.ometa_client = ometa_client
         self.service_connection_config = service_connection_config
         self.table_entity = table_entity
         self.sampler = sampler
+        self.validator_builder_class = validator_builder
 
     @classmethod
     def create(
@@ -77,15 +79,15 @@ def create(
     @abstractmethod
     def _get_validator_builder(
         self, test_case: TestCase, entity_type: str
-    ) -> IValidatorBuilder:
+    ) -> ValidatorBuilder:
         """get the builder class for the validator. Define this in the implementation class
 
         Args:
             test_case (TestCase): test case object
             entity_type (str): type of the entity
 
         Returns:
-            IValidatorBuilder: a validator builder
+            ValidatorBuilder: a validator builder
         """
         raise NotImplementedError
 

ingestion/src/metadata/data_quality/processor/test_case_runner.py

@@ -23,10 +23,8 @@
     TestCaseResults,
     TestSuiteProcessorConfig,
 )
+from metadata.data_quality.runner.base_test_suite_source import BaseTestSuiteRunner
 from metadata.data_quality.runner.core import DataTestsRunner
-from metadata.data_quality.runner.test_suite_source_factory import (
-    test_suite_source_factory,
-)
 from metadata.generated.schema.api.tests.createTestCase import CreateTestCaseRequest
 from metadata.generated.schema.entity.data.table import Table
 from metadata.generated.schema.entity.services.ingestionPipelines.status import (
@@ -95,12 +93,7 @@ def _run(self, record: TableAndTests) -> Either:
             record.table, openmetadata_test_cases
         )
 
-        test_suite_runner = test_suite_source_factory.create(
-            record.service_type.lower(),
-            self.config,
-            self.metadata,
-            record.table,
-        ).get_data_quality_runner()
+        test_suite_runner = self.get_test_suite_runner(record.table)
 
         logger.debug(
             f"Found {len(openmetadata_test_cases)} test cases for table {record.table.fullyQualifiedName.root}"
@@ -357,3 +350,10 @@ def filter_incompatible_test_cases(
             else:
                 result.append(tc)
         return result
+
+    def get_test_suite_runner(self, table: Table):
+        return BaseTestSuiteRunner(
+            self.config,
+            self.metadata,
+            table,
+        ).get_data_quality_runner()

ingestion/src/metadata/data_quality/runner/base_test_suite_source.py

@@ -15,6 +15,7 @@
 from copy import deepcopy
 from typing import Optional, cast
 
+from metadata.data_quality.builders.validator_builder import ValidatorBuilder
 from metadata.data_quality.interface.test_suite_interface import TestSuiteInterface
 from metadata.data_quality.runner.core import DataTestsRunner
 from metadata.generated.schema.entity.data.table import Table
@@ -46,6 +47,7 @@ def __init__(
         ometa_client: OpenMetadata,
         entity: Table,
     ):
+        self.validator_builder_class = ValidatorBuilder
         self._interface = None
         self._interface_type: str = config.source.type.lower()
         self.entity = entity
@@ -125,10 +127,11 @@ def create_data_quality_interface(self) -> TestSuiteInterface:
         )
 
         self.interface: TestSuiteInterface = test_suite_class.create(
-            self.service_conn_config,
-            self.ometa_client,
-            sampler_interface,
-            self.entity,
+            service_connection_config=self.service_conn_config,
+            ometa_client=self.ometa_client,
+            sampler=sampler_interface,
+            table_entity=self.entity,
+            validator_builder=self.validator_builder_class,
         )
         return self.interface