def add_citation(
self,
comment: str,
batch_request: Optional[
Union[str, Dict[str, Union[str, Dict[str, Any]]]]
] = None,
batch_definition: Optional[dict] = None,
batch_spec: Optional[dict] = None,
batch_kwargs: Optional[dict] = None,
batch_markers: Optional[dict] = None,
batch_parameters: Optional[dict] = None,
citation_date: Optional[Union[str, datetime.datetime]] = None,
):
if "citations" not in self.meta:
self.meta["citations"] = []
if isinstance(citation_date, str):
citation_date = parse_string_to_datetime(datetime_string=citation_date)
citation_date = citation_date or datetime.datetime.now(datetime.timezone.utc)
self.meta["citations"].append(
{
"citation_date": get_datetime_string_from_strftime_format(
format_str="%Y-%m-%dT%H:%M:%S.%fZ", datetime_obj=citation_date
),
"batch_request": batch_request,
"batch_definition": batch_definition,
"batch_spec": batch_spec,
"batch_kwargs": batch_kwargs,
"batch_markers": batch_markers,
"batch_parameters": batch_parameters,
"comment": comment,
}
)
def add_citation(
self,
comment: str,
batch_request: Optional[Union[str, Dict[str,
Union[str,
Dict[str,
Any]]]]] = None,
batch_definition: Optional[dict] = None,
batch_spec: Optional[dict] = None,
batch_kwargs: Optional[dict] = None,
batch_markers: Optional[dict] = None,
batch_parameters: Optional[dict] = None,
profiler_config: Optional[dict] = None,
citation_date: Optional[Union[str, datetime.datetime]] = None,
):
if "citations" not in self.meta:
self.meta["citations"] = []
if isinstance(citation_date, str):
citation_date = parse_string_to_datetime(
datetime_string=citation_date)
citation_date = citation_date or datetime.datetime.now(
datetime.timezone.utc)
citation: Dict[str,
Any] = {
"citation_date":
get_datetime_string_from_strftime_format(
format_str="%Y-%m-%dT%H:%M:%S.%fZ",
datetime_obj=citation_date),
"batch_request":
batch_request,
"batch_definition":
batch_definition,
"batch_spec":
batch_spec,
"batch_kwargs":
batch_kwargs,
"batch_markers":
batch_markers,
"batch_parameters":
batch_parameters,
"profiler_config":
profiler_config,
"comment":
comment,
}
ge.util.filter_properties_dict(properties=citation,
clean_falsy=True,
inplace=True)
self.meta["citations"].append(citation)
def run(
self,
template_name: Optional[str] = None,
run_name_template: Optional[str] = None,
expectation_suite_name: Optional[str] = None,
batch_request: Optional[Union[BatchRequest, dict]] = None,
action_list: Optional[List[dict]] = None,
evaluation_parameters: Optional[dict] = None,
runtime_configuration: Optional[dict] = None,
validations: Optional[List[dict]] = None,
profilers: Optional[List[dict]] = None,
run_id=None,
run_name=None,
run_time=None,
result_format=None,
**kwargs,
) -> CheckpointResult:
assert not (run_id and run_name) and not (
run_id and run_time
), "Please provide either a run_id or run_name and/or run_time."
run_time = run_time or datetime.now()
runtime_configuration: dict = runtime_configuration or {}
result_format: Optional[
dict] = result_format or runtime_configuration.get("result_format")
if result_format is None:
result_format = {"result_format": "SUMMARY"}
runtime_kwargs = {
"template_name": template_name,
"run_name_template": run_name_template,
"expectation_suite_name": expectation_suite_name,
"batch_request": batch_request,
"action_list": action_list,
"evaluation_parameters": evaluation_parameters,
"runtime_configuration": runtime_configuration,
"validations": validations,
"profilers": profilers,
}
substituted_runtime_config: CheckpointConfig = self.get_substituted_config(
runtime_kwargs=runtime_kwargs)
run_name_template: Optional[
str] = substituted_runtime_config.run_name_template
validations: list = substituted_runtime_config.validations
run_results = {}
if run_name is None and run_name_template is not None:
run_name: str = get_datetime_string_from_strftime_format(
format_str=run_name_template, datetime_obj=run_time)
run_id = run_id or RunIdentifier(run_name=run_name, run_time=run_time)
for idx, validation_dict in enumerate(validations):
try:
substituted_validation_dict: dict = get_substituted_validation_dict(
substituted_runtime_config=substituted_runtime_config,
validation_dict=validation_dict,
)
batch_request: BatchRequest = substituted_validation_dict.get(
"batch_request")
expectation_suite_name: str = substituted_validation_dict.get(
"expectation_suite_name")
action_list: list = substituted_validation_dict.get(
"action_list")
validator: Validator = self.data_context.get_validator(
batch_request=batch_request,
expectation_suite_name=expectation_suite_name,
)
action_list_validation_operator: ActionListValidationOperator = (
ActionListValidationOperator(
data_context=self.data_context,
action_list=action_list,
result_format=result_format,
name=f"{self.name}-checkpoint-validation[{idx}]",
))
val_op_run_result: ValidationOperatorResult = (
action_list_validation_operator.run(
assets_to_validate=[validator],
run_id=run_id,
evaluation_parameters=substituted_validation_dict.get(
"evaluation_parameters"),
result_format=result_format,
))
run_results.update(val_op_run_result.run_results)
except CheckpointError as e:
raise CheckpointError(
f"Exception occurred while running validation[{idx}] of checkpoint '{self.name}': {e.message}"
)
return CheckpointResult(run_id=run_id,
run_results=run_results,
checkpoint_config=self.config)
def run(
self,
template_name: Optional[str] = None,
run_name_template: Optional[str] = None,
expectation_suite_name: Optional[str] = None,
batch_request: Optional[Union[BatchRequestBase, dict]] = None,
action_list: Optional[List[dict]] = None,
evaluation_parameters: Optional[dict] = None,
runtime_configuration: Optional[dict] = None,
validations: Optional[List[dict]] = None,
profilers: Optional[List[dict]] = None,
run_id: Optional[Union[str, RunIdentifier]] = None,
run_name: Optional[str] = None,
run_time: Optional[Union[str, datetime.datetime]] = None,
result_format: Optional[Union[str, dict]] = None,
expectation_suite_ge_cloud_id: Optional[str] = None,
) -> CheckpointResult:
assert not (run_id and run_name) and not (
run_id and run_time
), "Please provide either a run_id or run_name and/or run_time."
run_time = run_time or datetime.datetime.now()
runtime_configuration = runtime_configuration or {}
result_format = result_format or runtime_configuration.get(
"result_format")
batch_request = get_batch_request_as_dict(batch_request=batch_request)
validations = get_validations_with_batch_request_as_dict(
validations=validations)
runtime_kwargs: dict = {
"template_name": template_name,
"run_name_template": run_name_template,
"expectation_suite_name": expectation_suite_name,
"batch_request": batch_request or {},
"action_list": action_list or [],
"evaluation_parameters": evaluation_parameters or {},
"runtime_configuration": runtime_configuration or {},
"validations": validations or [],
"profilers": profilers or [],
"expectation_suite_ge_cloud_id": expectation_suite_ge_cloud_id,
}
substituted_runtime_config: dict = self.get_substituted_config(
runtime_kwargs=runtime_kwargs)
run_name_template = substituted_runtime_config.get("run_name_template")
batch_request = substituted_runtime_config.get("batch_request")
validations = substituted_runtime_config.get("validations") or []
if len(validations) == 0 and not batch_request:
raise ge_exceptions.CheckpointError(
f'Checkpoint "{self.name}" must contain either a batch_request or validations.'
)
if run_name is None and run_name_template is not None:
run_name = get_datetime_string_from_strftime_format(
format_str=run_name_template, datetime_obj=run_time)
run_id = run_id or RunIdentifier(run_name=run_name, run_time=run_time)
# Use AsyncExecutor to speed up I/O bound validations by running them in parallel with multithreading (if
# concurrency is enabled in the data context configuration) -- please see the below arguments used to initialize
# AsyncExecutor and the corresponding AsyncExecutor docstring for more details on when multiple threads are
# used.
with AsyncExecutor(self.data_context.concurrency,
max_workers=len(validations)) as async_executor:
# noinspection PyUnresolvedReferences
async_validation_operator_results: List[
AsyncResult[ValidationOperatorResult]] = []
if len(validations) > 0:
for idx, validation_dict in enumerate(validations):
self._run_validation(
substituted_runtime_config=substituted_runtime_config,
async_validation_operator_results=
async_validation_operator_results,
async_executor=async_executor,
result_format=result_format,
run_id=run_id,
idx=idx,
validation_dict=validation_dict,
)
else:
self._run_validation(
substituted_runtime_config=substituted_runtime_config,
async_validation_operator_results=
async_validation_operator_results,
async_executor=async_executor,
result_format=result_format,
run_id=run_id,
)
run_results: dict = {}
for async_validation_operator_result in async_validation_operator_results:
run_results.update(
async_validation_operator_result.result().run_results)
return CheckpointResult(
run_id=run_id,
run_results=run_results,
checkpoint_config=self.config,
)
def resolve_config_using_acceptable_arguments(
checkpoint: "Checkpoint", # noqa: F821
template_name: Optional[str] = None,
run_name_template: Optional[str] = None,
expectation_suite_name: Optional[str] = None,
batch_request: Optional[Union[BatchRequest, RuntimeBatchRequest,
dict]] = None,
action_list: Optional[List[dict]] = None,
evaluation_parameters: Optional[dict] = None,
runtime_configuration: Optional[dict] = None,
validations: Optional[List[dict]] = None,
profilers: Optional[List[dict]] = None,
run_id: Optional[Union[str, RunIdentifier]] = None,
run_name: Optional[str] = None,
run_time: Optional[Union[str, datetime.datetime]] = None,
result_format: Optional[Union[str, dict]] = None,
expectation_suite_ge_cloud_id: Optional[str] = None,
) -> dict:
"""
This method reconciles the Checkpoint configuration (e.g., obtained from the Checkpoint store) with dynamically
supplied arguments in order to obtain that Checkpoint specification that is ready for running validation on it.
This procedure is necessecitated by the fact that the Checkpoint configuration is hierarchical in its form,
which was established for the purposes of making the specification of different Checkpoint capabilities easy.
In particular, entities, such as BatchRequest, expectation_suite_name, and action_list, can be specified at the
top Checkpoint level with the suitable ovverrides provided at lower levels (e.g., in the validations section).
Reconciling and normalizing the Checkpoint configuration is essential for usage statistics, because the exact
values of the entities in their formally validated form (e.g., BatchRequest) is the required level of detail.
"""
assert not (run_id and run_name) and not (
run_id and run_time
), "Please provide either a run_id or run_name and/or run_time."
run_time = run_time or datetime.datetime.now()
runtime_configuration = runtime_configuration or {}
batch_request = get_batch_request_as_dict(batch_request=batch_request)
validations = get_validations_with_batch_request_as_dict(
validations=validations)
runtime_kwargs: dict = {
"template_name": template_name,
"run_name_template": run_name_template,
"expectation_suite_name": expectation_suite_name,
"batch_request": batch_request,
"action_list": action_list,
"evaluation_parameters": evaluation_parameters,
"runtime_configuration": runtime_configuration,
"validations": validations,
"profilers": profilers,
"expectation_suite_ge_cloud_id": expectation_suite_ge_cloud_id,
}
substituted_runtime_config: dict = checkpoint.get_substituted_config(
runtime_kwargs=runtime_kwargs)
run_name_template = substituted_runtime_config.get("run_name_template")
validations = substituted_runtime_config.get("validations") or []
batch_request = substituted_runtime_config.get("batch_request")
if len(validations) == 0 and not batch_request:
raise ge_exceptions.CheckpointError(
f'Checkpoint "{checkpoint.name}" must contain either a batch_request or validations.'
)
if run_name is None and run_name_template is not None:
run_name = get_datetime_string_from_strftime_format(
format_str=run_name_template, datetime_obj=run_time)
run_id = run_id or RunIdentifier(run_name=run_name, run_time=run_time)
validation_dict: dict
for validation_dict in validations:
substituted_validation_dict: dict = get_substituted_validation_dict(
substituted_runtime_config=substituted_runtime_config,
validation_dict=validation_dict,
)
validation_batch_request: Union[
BatchRequest,
RuntimeBatchRequest] = substituted_validation_dict.get(
"batch_request")
validation_dict["batch_request"] = validation_batch_request
validation_expectation_suite_name: str = substituted_validation_dict.get(
"expectation_suite_name")
validation_dict[
"expectation_suite_name"] = validation_expectation_suite_name
validation_expectation_suite_ge_cloud_id: str = (
substituted_validation_dict.get(
"expectation_suite_ge_cloud_id"))
validation_dict[
"expectation_suite_ge_cloud_id"] = validation_expectation_suite_ge_cloud_id
validation_action_list: list = substituted_validation_dict.get(
"action_list")
validation_dict["action_list"] = validation_action_list
return substituted_runtime_config
