def test_get_metadata_for_raw_files_discovered_after_datetime(self):
with freeze_time('2015-01-02T03:05:05'):
raw_unprocessed_path_1 = self._make_unprocessed_path(
'bucket/file_tag.csv',
GcsfsDirectIngestFileType.RAW_DATA,
dt=datetime.datetime.utcnow())
self.metadata_manager.register_new_file(raw_unprocessed_path_1)
self.metadata_manager_other_region.register_new_file(
raw_unprocessed_path_1)
with freeze_time('2015-01-02T03:06:06'):
raw_unprocessed_path_2 = self._make_unprocessed_path(
'bucket/other_tag.csv',
GcsfsDirectIngestFileType.RAW_DATA,
dt=datetime.datetime.utcnow())
self.metadata_manager.register_new_file(raw_unprocessed_path_2)
with freeze_time('2015-01-02T03:07:07'):
raw_unprocessed_path_3 = self._make_unprocessed_path(
'bucket/file_tag.csv',
GcsfsDirectIngestFileType.RAW_DATA,
dt=datetime.datetime.utcnow())
self.metadata_manager.register_new_file(raw_unprocessed_path_3)
expected_list = [
DirectIngestRawFileMetadata.new_with_defaults(
region_code=self.metadata_manager.region_code,
file_tag='file_tag',
discovery_time=datetime.datetime(2015, 1, 2, 3, 5, 5),
normalized_file_name=
'unprocessed_2015-01-02T03:05:05:000000_raw_file_tag.csv',
datetimes_contained_upper_bound_inclusive=datetime.datetime(
2015, 1, 2, 3, 5, 5)),
DirectIngestRawFileMetadata.new_with_defaults(
region_code=self.metadata_manager.region_code,
file_tag='file_tag',
discovery_time=datetime.datetime(2015, 1, 2, 3, 7, 7),
normalized_file_name=
'unprocessed_2015-01-02T03:07:07:000000_raw_file_tag.csv',
datetimes_contained_upper_bound_inclusive=datetime.datetime(
2015, 1, 2, 3, 7, 7))
]
self.assertEqual(
expected_list,
self.metadata_manager.
get_metadata_for_raw_files_discovered_after_datetime(
'file_tag', discovery_time_lower_bound_exclusive=None))
expected_list = expected_list[-1:]
self.assertEqual(
expected_list,
self.metadata_manager.
get_metadata_for_raw_files_discovered_after_datetime(
'file_tag',
discovery_time_lower_bound_exclusive=datetime.datetime(
2015, 1, 2, 3, 7, 0)))
def test_get_raw_file_metadata_unique_to_state(self):
# Arrange
raw_unprocessed_path = self._make_unprocessed_path(
'bucket/file_tag.csv', GcsfsDirectIngestFileType.RAW_DATA)
self.metadata_manager_other_region.mark_file_as_discovered(
raw_unprocessed_path)
# Act
self.metadata_manager.mark_file_as_discovered(raw_unprocessed_path)
metadata = self.metadata_manager.get_file_metadata(
raw_unprocessed_path)
# Assert
expected_metadata = DirectIngestRawFileMetadata.new_with_defaults(
region_code=self.metadata_manager.region_code,
file_tag='file_tag',
discovery_time=datetime.datetime(2015, 1, 2, 3, 4, 6),
normalized_file_name=
'unprocessed_2015-01-02T03:03:03:000003_raw_file_tag.csv',
processed_time=None,
datetimes_contained_upper_bound_inclusive=datetime.datetime(
2015, 1, 2, 3, 3, 3, 3))
self.assertIsInstance(metadata, DirectIngestRawFileMetadata)
self.assertIsNotNone(metadata.file_id)
self.assertEqual(expected_metadata, metadata)
def _metadata_for_unprocessed_file_path(
self, path: GcsfsFilePath) -> DirectIngestRawFileMetadata:
parts = filename_parts_from_path(path)
return DirectIngestRawFileMetadata(
region_code=self.test_region.region_code,
file_tag=parts.file_tag,
file_id=123,
processed_time=None,
normalized_file_name=path.file_name,
discovery_time=datetime.datetime.now(),
datetimes_contained_upper_bound_inclusive=parts.
utc_upload_datetime,
)
def test_getIngestViewExportTaskArgs_happy(self) -> None:
# Arrange
region = self.create_fake_region(ingest_view_exports_enabled=True)
export_manager = self.create_export_manager(region)
export_manager.file_metadata_manager.get_ingest_view_metadata_for_most_recent_valid_job = Mock( # type: ignore
return_value=DirectIngestIngestFileMetadata(
file_id=_ID,
region_code=region.region_code,
file_tag="ingest_view",
normalized_file_name="normalized_file_name",
processed_time=_DATE_1,
is_invalidated=False,
is_file_split=False,
job_creation_time=_DATE_1,
export_time=_DATE_1,
datetimes_contained_lower_bound_exclusive=_DATE_1,
datetimes_contained_upper_bound_inclusive=_DATE_1,
discovery_time=_DATE_1,
))
export_manager.file_metadata_manager.get_metadata_for_raw_files_discovered_after_datetime = Mock( # type: ignore
return_value=[
DirectIngestRawFileMetadata(
file_id=2,
region_code=region.region_code,
file_tag="ingest_view",
discovery_time=_DATE_2,
normalized_file_name=
"unprocessed_2015-01-02T03:03:03:000003_raw_file_tag.csv",
processed_time=None,
datetimes_contained_upper_bound_inclusive=_DATE_2,
)
])
# Act
args = export_manager.get_ingest_view_export_task_args()
# Assert
self.assertListEqual(
args,
[
GcsfsIngestViewExportArgs(
ingest_view_name="ingest_view",
upper_bound_datetime_prev=_DATE_1,
upper_bound_datetime_to_export=_DATE_2,
)
],
)
def test_getIngestViewExportTaskArgs_rawCodeTableOlderThanLastExport(
self) -> None:
# Arrange
CODE_TABLE_TAG = "RECIDIVIZ_REFERENCE_ingest_view"
region = self.create_fake_region(ingest_view_exports_enabled=True)
export_manager = self.create_export_manager(
region, controller_file_tags=[CODE_TABLE_TAG])
export_manager.file_metadata_manager.get_ingest_view_metadata_for_most_recent_valid_job = Mock( # type: ignore
return_value=DirectIngestIngestFileMetadata(
file_id=_ID,
region_code=region.region_code,
file_tag=CODE_TABLE_TAG,
normalized_file_name="normalized_file_name",
processed_time=_DATE_2,
is_invalidated=False,
is_file_split=False,
job_creation_time=_DATE_2,
export_time=_DATE_2,
datetimes_contained_lower_bound_exclusive=_DATE_2 -
datetime.timedelta(days=7),
datetimes_contained_upper_bound_inclusive=_DATE_2,
discovery_time=_DATE_2,
))
export_manager.file_metadata_manager.get_metadata_for_raw_files_discovered_after_datetime = Mock( # type: ignore
return_value=[
DirectIngestRawFileMetadata(
file_id=2,
region_code=region.region_code,
file_tag=CODE_TABLE_TAG,
discovery_time=_DATE_1,
normalized_file_name=
"unprocessed_2015-01-02T03:03:03:000003_raw_file_tag.csv",
processed_time=None,
datetimes_contained_upper_bound_inclusive=_DATE_1,
)
])
# Act
args = export_manager.get_ingest_view_export_task_args()
# Assert
# New code tables are backdated but don't need to be re-ingested, so ignore them.
self.assertListEqual(args, [])
def test_getIngestViewExportTaskArgs_rawFileOlderThanLastExport(
self) -> None:
# Arrange
region = self.create_fake_region(ingest_view_exports_enabled=True)
export_manager = self.create_export_manager(region)
export_manager.file_metadata_manager.get_ingest_view_metadata_for_most_recent_valid_job = Mock( # type: ignore
return_value=DirectIngestIngestFileMetadata(
file_id=_ID,
region_code=region.region_code,
file_tag="ingest_view",
normalized_file_name="normalized_file_name",
processed_time=_DATE_2,
is_invalidated=False,
is_file_split=False,
job_creation_time=_DATE_2,
export_time=_DATE_2,
datetimes_contained_lower_bound_exclusive=_DATE_2,
datetimes_contained_upper_bound_inclusive=_DATE_2,
discovery_time=_DATE_2,
))
export_manager.file_metadata_manager.get_metadata_for_raw_files_discovered_after_datetime = Mock( # type: ignore
return_value=[
DirectIngestRawFileMetadata(
file_id=2,
region_code=region.region_code,
file_tag="ingest_view",
discovery_time=_DATE_1,
normalized_file_name=
"unprocessed_2015-01-02T03:03:03:000003_raw_file_tag.csv",
processed_time=None,
datetimes_contained_upper_bound_inclusive=_DATE_1,
)
])
# Act
with pytest.raises(
ValueError,
match=r"upper bound date.*before the last valid export"):
export_manager.get_ingest_view_export_task_args()
def do_upload(
state_code: StateCode,
sandbox_dataset_prefix: str,
source_bucket: GcsfsBucketPath,
file_tag_filter: Optional[str],
) -> None:
"""Imports a set of raw data files in the given source bucket into a sandbox
dataset.
"""
input_str = input(
f"Have you already uploaded raw files to [{source_bucket.uri()}] using script "
f"`recidiviz.tools.ingest.operations.upload_raw_state_files_to_ingest_bucket_with_date` "
f"with arg `--destination-bucket {source_bucket.bucket_name}`?. [y/n] "
)
if input_str.upper() != "Y":
return
import_manager = SandboxDirectIngestRawFileImportManager(
state_code=state_code,
sandbox_dataset_prefix=sandbox_dataset_prefix,
test_ingest_bucket=source_bucket,
)
bq_client = BigQueryClientImpl()
# Create the dataset up front with table expiration
bq_client.create_dataset_if_necessary(
bq_client.dataset_ref_for_id(dataset_id=import_manager.sandbox_dataset),
default_table_expiration_ms=TEMP_DATASET_DEFAULT_TABLE_EXPIRATION_MS,
)
raw_files_to_import = import_manager.get_unprocessed_raw_files_to_import()
failures_by_exception = defaultdict(list)
for i, file_path in enumerate(raw_files_to_import):
parts = filename_parts_from_path(file_path)
if file_tag_filter and not re.search(file_tag_filter, parts.file_tag):
logging.info("** Skipping file with tag [%s] **", parts.file_tag)
continue
logging.info("Running file with tag [%s]", parts.file_tag)
try:
import_manager.import_raw_file_to_big_query(
file_path,
DirectIngestRawFileMetadata(
file_id=i,
region_code=state_code.value,
file_tag=parts.file_tag,
processed_time=None,
discovery_time=datetime.datetime.now(),
normalized_file_name=file_path.file_name,
datetimes_contained_upper_bound_inclusive=parts.utc_upload_datetime,
),
)
except Exception as e:
logging.exception(e)
failures_by_exception[str(e)].append(file_path.abs_path())
if failures_by_exception:
logging.error("************************* FAILURES ************************")
total_files = 0
all_failed_paths = []
for error, file_list in failures_by_exception.items():
total_files += len(file_list)
all_failed_paths += file_list
logging.error(
"Failed [%s] files with error [%s]: %s",
len(file_list),
error,
file_list,
)
logging.error("***********************************************************")
raise ValueError(f"Failed to import [{total_files}] files: {all_failed_paths}")
