Skip to content

Commit

Permalink
Update imports
Browse files Browse the repository at this point in the history
  • Loading branch information
alexmassen-hane committed Sep 21, 2023
1 parent 06c7284 commit fe657a1
Show file tree
Hide file tree
Showing 2 changed files with 20 additions and 33 deletions.
Original file line number Diff line number Diff line change
Expand Up @@ -23,23 +23,23 @@
import logging
import pendulum
from datetime import timedelta
from google.cloud import bigquery
from dataclasses import dataclass
from typing import Dict, List, Optional, Union

from google.cloud import bigquery
from google.cloud.bigquery import Table as BQTable

from academic_observatory_workflows.config import schema_folder as default_schema_folder, Tag

from observatory.platform.observatory_config import CloudWorkspace
from observatory.platform.utils.dag_run_sensor import DagRunSensor
from observatory.platform.workflows.workflow import Workflow, set_task_state, Release
from observatory.platform.bigquery import (
bq_load_from_memory,
bq_create_dataset,
bq_run_query,
bq_table_exists,
bq_select_columns,
)
from observatory.platform.observatory_config import CloudWorkspace
from observatory.platform.utils.dag_run_sensor import DagRunSensor
from observatory.platform.workflows.workflow import Workflow, set_task_state, Release


@dataclass
Expand Down Expand Up @@ -77,7 +77,7 @@ def full_table_id(self):
def make_datasets(project_id: str) -> Dict[str, List[Table]]:
"""Create a list of datasets for the DQC Workflow to process.
This list can be easily updated, and can be modified for tables outside the given project_id if needed.
This list can be easily updated, and can be modified for tables outside the given project_id if required.
:param project_id: Name of the Google project that the tables are stored in.
:return: A dictionary of key: dataset_id and the list of tables in that dataset."""
Expand Down Expand Up @@ -521,7 +521,6 @@ def perform_data_quality_check(self, release: Release, **kwargs):
dataset_id=table_to_check.dataset_id, base_name=table_to_check.name
)
else:
print(f"table_to_check.full_table_id: {table_to_check.full_table_id}")
sub_tables: List[BQTable] = [bq_get_table(table_to_check.full_table_id)]

assert (
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -16,26 +16,29 @@

import os
import pendulum


from typing import List


from datetime import timedelta
from google.cloud import bigquery
from google.cloud.bigquery import Table as BQTable

from academic_observatory_workflows.config import schema_folder as default_schema_folder


from academic_observatory_workflows.model import (
bq_load_observatory_dataset,
make_observatory_dataset,
from academic_observatory_workflows.workflows.tests.test_doi_workflow import TestDoiWorkflow
from academic_observatory_workflows.model import bq_load_observatory_dataset, make_observatory_dataset
from academic_observatory_workflows.config import test_fixtures_folder, schema_folder as default_schema_folder
from academic_observatory_workflows.workflows.data_quality_check_workflow import (
DataQualityCheckWorkflow,
Table,
create_dqc_record,
bq_count_distinct_records,
bq_count_nulls,
bq_get_table,
bq_list_tables_shards,
bq_count_duplicate_records,
create_table_hash_id,
is_in_dqc_table,
)

from observatory.platform.files import load_jsonl
from observatory.platform.observatory_config import Workflow
from observatory.platform.observatory_environment import ObservatoryEnvironment, ObservatoryTestCase
from observatory.platform.observatory_environment import (
ObservatoryEnvironment,
ObservatoryTestCase,
Expand All @@ -51,21 +54,6 @@
bq_create_dataset,
)

from academic_observatory_workflows.config import test_fixtures_folder
from academic_observatory_workflows.workflows.data_quality_check_workflow import (
DataQualityCheckWorkflow,
Table,
create_dqc_record,
bq_count_distinct_records,
bq_count_nulls,
bq_get_table,
bq_list_tables_shards,
bq_count_duplicate_records,
create_table_hash_id,
is_in_dqc_table,
)
from academic_observatory_workflows.workflows.tests.test_doi_workflow import TestDoiWorkflow


class TestDataQualityCheckWorkflow(ObservatoryTestCase):
"""Tests for the Data Quality Check Workflow"""
Expand Down

0 comments on commit fe657a1

Please sign in to comment.