def stream_file_to_s3(upload_name, reader, is_certified=False):
"""Stream file to S3
Args:
upload_name - file name to be used as S3 key
reader - reader object to read data from
is_certified - True if writing to the certified bucket, False otherwise (default False)
"""
path, file_name = upload_name.rsplit('/', 1)
logger.debug({
'message': 'Streaming file to S3',
'message_type': 'ValidatorDebug',
'file_name': file_name if file_name else path
})
if is_certified:
handler = S3Handler.create_file_path(upload_name, CONFIG_BROKER["certified_bucket"])
else:
handler = S3Handler.create_file_path(upload_name)
with smart_open.smart_open(handler, 'w') as writer:
while True:
chunk = reader.read(CHUNK_SIZE)
if chunk:
writer.write(chunk)
else:
break
def copy_file_from_parent_to_child(child_job, parent_job, is_local):
""" Copy the file from the parent job's bucket to the child job's bucket.
Args:
child_job: Job object for the child FileRequest
parent_job: Job object for the parent FileRequest
is_local: A boolean flag indicating whether the application is being run locally or not
"""
file_type = parent_job.file_type.letter_name
log_data = {'message': 'Copying data from parent job with job_id:{}'.format(parent_job.job_id),
'message_type': 'ValidatorInfo', 'job_id': child_job.job_id, 'file_type': parent_job.file_type.name}
if not is_local:
is_local = g.is_local
if not is_local and parent_job.filename != child_job.filename:
# Check to see if the same file exists in the child bucket
s3 = boto3.client('s3', region_name=CONFIG_BROKER["aws_region"])
response = s3.list_objects_v2(Bucket=CONFIG_BROKER['aws_bucket'], Prefix=child_job.filename)
for obj in response.get('Contents', []):
if obj['Key'] == child_job.filename:
# The file already exists in this location
log_data['message'] = 'Cached {} file CSV already exists in this location'.format(file_type)
logger.info(log_data)
return
# Copy the parent file into the child's S3 location
log_data['message'] = 'Copying the cached {} file from job {}'.format(file_type, parent_job.job_id)
logger.info(log_data)
S3Handler.copy_file(CONFIG_BROKER['aws_bucket'], CONFIG_BROKER['aws_bucket'], parent_job.filename,
child_job.filename)
def check_file_generation(job_id):
""" Check the status of a file generation
Args:
job_id: upload Job ID
Return:
Dict with keys: job_id, status, file_type, message, url, start, end
"""
sess = GlobalDB.db().session
# We want to use one_or_none() here so we can see if the job is None so we can mark the status as invalid to
# indicate that a status request is invoked for a job that isn't created yet
upload_job = sess.query(Job).filter_by(job_id=job_id).one_or_none()
response_dict = {
'job_id': job_id,
'status': '',
'file_type': '',
'message': '',
'url': '#',
'size': None
}
if upload_job is None:
response_dict['start'] = ''
response_dict['end'] = ''
response_dict['status'] = 'invalid'
response_dict[
'message'] = 'No generation job found with the specified ID'
return response_dict
response_dict['file_type'] = lookups.FILE_TYPE_DICT_LETTER[
upload_job.file_type_id]
response_dict['size'] = upload_job.file_size
response_dict['status'] = map_generate_status(sess, upload_job)
response_dict['message'] = upload_job.error_message or ''
# Generate the URL (or path) to the file
if CONFIG_BROKER['use_aws'] and response_dict[
'status'] is 'finished' and upload_job.filename:
path, file_name = upload_job.filename.split('/')
response_dict['url'] = S3Handler().get_signed_url(
path=path,
file_name=file_name,
bucket_route=None,
url_mapping=CONFIG_BROKER["submission_bucket_mapping"],
method='get_object')
elif response_dict['status'] is 'finished' and upload_job.filename:
response_dict['url'] = upload_job.filename
# Only D file generations have start and end dates
if response_dict['file_type'] in ['D1', 'D2']:
response_dict['start'] = upload_job.start_date.strftime(
"%m/%d/%Y") if upload_job.start_date is not None else ""
response_dict['end'] = upload_job.end_date.strftime(
"%m/%d/%Y") if upload_job.end_date is not None else ""
return response_dict
def generate_file(self, agency_code=None):
""" Generates a file based on the FileGeneration object and updates any Jobs referencing it """
raw_filename = (GEN_FILENAMES[self.file_type]
if not self.file_generation else
GEN_FILENAMES[self.file_type].format(
self.file_generation.agency_type))
file_name = S3Handler.get_timestamped_filename(raw_filename)
if self.is_local:
file_path = "".join([CONFIG_BROKER['broker_files'], file_name])
else:
file_path = "".join(["None/", file_name])
# Generate the file and upload to S3
log_data = {
'message': 'Finished file {} generation'.format(self.file_type),
'message_type': 'ValidatorInfo',
'file_type': self.file_type,
'file_path': file_path
}
if self.file_generation:
self.generate_d_file(file_path)
log_data.update({
'agency_code':
self.file_generation.agency_code,
'agency_type':
self.file_generation.agency_type,
'start_date':
self.file_generation.start_date,
'end_date':
self.file_generation.end_date,
'file_generation_id':
self.file_generation.file_generation_id
})
elif self.job.file_type.letter_name in ['A', 'E', 'F']:
log_data['job_id'] = self.job.job_id
mark_job_status(self.job.job_id, 'running')
if self.job.file_type.letter_name == 'A':
if not agency_code:
raise ResponseException(
'Agency code not provided for an A file generation')
self.generate_a_file(agency_code, file_path)
else:
# Call self.generate_%s_file() where %s is e or f based on the Job's file_type
file_type_lower = self.job.file_type.letter_name.lower()
getattr(self, 'generate_%s_file' % file_type_lower)()
mark_job_status(self.job.job_id, 'finished')
else:
e = 'No FileGeneration object for D file generation.' if self.file_type in ['D1', 'D2'] else \
'Cannot generate file for {} file type.'.format(self.file_type if self.file_type else 'empty')
raise ResponseException(e)
logger.info(log_data)
def generate_from_job(self, job_id, agency_code):
""" Generates a file for a specified job
Args:
job_id: ID of the upload Job
agency_code: FREC or CGAC code to generate data from
"""
mark_job_status(job_id, 'running')
with job_context(job_id, self.is_local) as context:
sess, job = context
# Ensure this is a file generation job
if job.job_type.name != 'file_upload':
raise ResponseException(
'Job ID {} is not a file generation job (job type is {})'.
format(job.job_id, job.job_type.name),
StatusCode.CLIENT_ERROR, None, ValidationError.jobError)
# Ensure there is an available agency_code
if not agency_code:
if job.submission_id:
agency_code = job.submission.frec_code if job.submission.frec_code else job.submission.cgac_code
else:
raise ResponseException(
'An agency_code must be provided to generate a file'.
format(job.job_id,
job.job_type.name), StatusCode.CLIENT_ERROR,
None, ValidationError.jobError)
# Generate timestamped file names
old_filename = job.original_filename
job.original_filename = S3Handler.get_timestamped_filename(
CONFIG_BROKER["".join([str(job.file_type.name),
"_file_name"])])
if self.is_local:
job.filename = "".join(
[CONFIG_BROKER['broker_files'], job.original_filename])
else:
job.filename = "".join(
[str(job.submission_id), "/", job.original_filename])
# Generate the file and upload to S3
if job.file_type.letter_name in ['D1', 'D2']:
# Update the validation Job if necessary
if job.submission_id:
self.update_validation_job_info(job)
generate_d_file(sess, job, agency_code, self.is_local,
old_filename)
elif job.file_type.letter_name == 'E':
generate_e_file(sess, job, self.is_local)
else:
generate_f_file(sess, job, self.is_local)
def generate_from_job(self):
""" Generates a file for a specified job """
# Mark Job as running
mark_job_status(self.job.job_id, 'running')
# Ensure this is a file generation job
job_type = self.job.job_type.name
if job_type != 'file_upload':
raise ResponseException(
'Job ID {} is not a file generation job (job type is {})'.format(self.job.job_id, job_type),
StatusCode.CLIENT_ERROR, None, ValidationError.jobError)
# Ensure there is an available agency_code
if not self.agency_code:
raise ResponseException(
'An agency_code must be provided to generate a file'.format(self.job.job_id, job_type),
StatusCode.CLIENT_ERROR, None, ValidationError.jobError)
# Retrieve any FileRequest that may have started since the Broker sent the request to SQS
skip_generation = None
if self.job.file_type.letter_name in ['D1', 'D2']:
skip_generation = retrieve_cached_file_request(self.job, self.agency_type, self.agency_code, self.is_local)
if not skip_generation:
# Generate timestamped file names
raw_filename = CONFIG_BROKER["".join([str(self.job.file_type.name), "_file_name"])]
self.job.original_filename = S3Handler.get_timestamped_filename(raw_filename)
if self.is_local:
self.job.filename = "".join([CONFIG_BROKER['broker_files'], self.job.original_filename])
else:
self.job.filename = "".join([str(self.job.submission_id), "/", self.job.original_filename])
self.sess.commit()
# Generate the file, and upload to S3
if self.job.file_type.letter_name in ['D1', 'D2']:
# Update the validation Job if necessary
update_validation_job_info(self.sess, self.job)
self.generate_d_file()
elif self.job.file_type.letter_name == 'A':
self.generate_a_file()
elif self.job.file_type.letter_name == 'E':
self.generate_e_file()
else:
self.generate_f_file()
mark_job_status(self.job.job_id, 'finished')
logger.info({
'message': 'Finished file {} generation'.format(self.job.file_type.letter_name),
'message_type': 'ValidatorInfo', 'job_id': self.job.job_id, 'agency_code': self.agency_code,
'file_type': self.job.file_type.letter_name, 'start_date': self.job.start_date,
'end_date': self.job.end_date, 'filename': self.job.original_filename
})
def get_fabs_meta(submission_id):
"""Return the total rows, valid rows, publish date, and publish file for FABS submissions"""
sess = GlobalDB.db().session
# get row counts from the DetachedAwardFinancialAssistance table
dafa = DetachedAwardFinancialAssistance
total_rows = sess.query(dafa).filter(dafa.submission_id == submission_id)
valid_rows = total_rows.filter(dafa.is_valid)
# retrieve the published data and file
submission = sess.query(Submission).filter(
Submission.submission_id == submission_id).one()
publish_date, published_file = None, None
certify_data = get_lastest_certified_date(submission, is_fabs=True)
try:
iter(certify_data)
except TypeError:
publish_date = certify_data
else:
publish_date, file_path = certify_data
if CONFIG_BROKER["use_aws"] and file_path:
path, file_name = file_path.rsplit(
'/', 1) # split by last instance of /
published_file = S3Handler().get_signed_url(
path=path,
file_name=file_name,
bucket_route=CONFIG_BROKER['certified_bucket'],
url_mapping=CONFIG_BROKER["certified_bucket_mapping"],
method="get_object")
elif file_path:
published_file = file_path
return {
'valid_rows':
valid_rows.count(),
'total_rows':
total_rows.count(),
'publish_date':
publish_date.strftime('%-I:%M%p %m/%d/%Y') if publish_date else None,
'published_file':
published_file
}
def generate_file(self, agency_code=None):
""" Generates a file based on the FileGeneration object and updates any Jobs referencing it """
raw_filename = (GEN_FILENAMES[self.file_type] if not self.file_generation else
GEN_FILENAMES[self.file_type].format(self.file_generation.agency_type))
file_name = S3Handler.get_timestamped_filename(raw_filename)
if self.is_local:
file_path = "".join([CONFIG_BROKER['broker_files'], file_name])
else:
file_path = "".join(["None/", file_name])
# Generate the file and upload to S3
log_data = {'message': 'Finished file {} generation'.format(self.file_type), 'message_type': 'ValidatorInfo',
'file_type': self.file_type, 'file_path': file_path}
if self.file_generation:
self.generate_d_file(file_path)
log_data.update({
'agency_code': self.file_generation.agency_code, 'agency_type': self.file_generation.agency_type,
'start_date': self.file_generation.start_date, 'end_date': self.file_generation.end_date,
'file_generation_id': self.file_generation.file_generation_id
})
elif self.job.file_type.letter_name in ['A', 'E', 'F']:
log_data['job_id'] = self.job.job_id
mark_job_status(self.job.job_id, 'running')
if self.job.file_type.letter_name == 'A':
if not agency_code:
raise ResponseException('Agency code not provided for an A file generation')
self.generate_a_file(agency_code, file_path)
else:
# Call self.generate_%s_file() where %s is e or f based on the Job's file_type
file_type_lower = self.job.file_type.letter_name.lower()
getattr(self, 'generate_%s_file' % file_type_lower)()
mark_job_status(self.job.job_id, 'finished')
else:
e = 'No FileGeneration object for D file generation.' if self.file_type in ['D1', 'D2'] else \
'Cannot generate file for {} file type.'.format(self.file_type if self.file_type else 'empty')
raise ResponseException(e)
logger.info(log_data)
def get_fabs_meta(submission_id):
"""Return the total rows, valid rows, publish date, and publish file for FABS submissions"""
sess = GlobalDB.db().session
# get row counts from the FABS table
total_rows = sess.query(FABS).filter(FABS.submission_id == submission_id)
valid_rows = total_rows.filter(FABS.is_valid)
# retrieve the published data and file
submission = sess.query(Submission).filter(
Submission.submission_id == submission_id).one()
publish_date, published_file = None, None
publish_data = get_latest_published_date(submission, is_fabs=True)
try:
iter(publish_data)
except TypeError:
publish_date = publish_data
else:
publish_date, file_path = publish_data
if CONFIG_BROKER['use_aws'] and file_path:
path, file_name = file_path.rsplit(
'/', 1) # split by last instance of /
published_file = S3Handler().get_signed_url(
path=path,
file_name=file_name,
bucket_route=CONFIG_BROKER['certified_bucket'],
url_mapping=CONFIG_BROKER['certified_bucket_mapping'])
elif file_path:
published_file = file_path
return {
'valid_rows':
valid_rows.count(),
'total_rows':
total_rows.count(),
'publish_date':
publish_date.strftime('%Y-%m-%dT%H:%M:%S') if publish_date else None,
'published_file':
published_file
}
def revert_to_certified(submission, file_manager):
""" Revert an updated DABS submission to its last certified state
Args:
submission: the submission to be reverted
file_manager: a FileHandler object to be used to call revert_certified_error_files and determine is_local
Returns:
A JsonResponse containing a success message
Raises:
ResponseException: if submission provided is a FABS submission or is not in an "updated" status
"""
if submission.d2_submission:
raise ResponseException('Submission must be a DABS submission.', status=StatusCode.CLIENT_ERROR)
if submission.publish_status_id != PUBLISH_STATUS_DICT['updated']:
raise ResponseException('Submission has not been certified or has not been updated since certification.',
status=StatusCode.CLIENT_ERROR)
sess = GlobalDB.db().session
move_certified_data(sess, submission.submission_id, direction='revert')
# Copy file paths from certified_files_history
max_cert_history = sess.query(func.max(CertifyHistory.certify_history_id), func.max(CertifyHistory.updated_at)).\
filter(CertifyHistory.submission_id == submission.submission_id).one()
remove_timestamp = [str(FILE_TYPE_DICT['appropriations']), str(FILE_TYPE_DICT['program_activity']),
str(FILE_TYPE_DICT['award_financial'])]
if file_manager.is_local:
filepath = CONFIG_BROKER['broker_files']
ef_path = ''
else:
filepath = '{}/'.format(submission.submission_id)
ef_path = filepath
remove_timestamp.extend([str(FILE_TYPE_DICT['executive_compensation']), str(FILE_TYPE_DICT['sub_award'])])
# Certified filename -> Job filename, original filename
# Local:
# A/B/C:
# filename -> '[broker_files dir][certified file base name]'
# original_filename -> '[certified file base name without the timestamp]'
# D1/D2:
# filename -> '[broker_files dir][certified file base name]'
# original_filename -> '[certified file base name]'
# E/F:
# filename -> '[certified file base name]'
# original_filename -> '[certified file base name]'
# Remote:
# A/B/C/E/F:
# filename -> '[submission_id]/[certified file base name]'
# original_filename -> '[certified file base name without the timestamp]'
# D1/D2:
# filename -> '[submission_id dir][certified file base name]'
# original_filename -> '[certified file base name]'
update_string = """
WITH filenames AS (
SELECT REVERSE(SPLIT_PART(REVERSE(filename), '/', 1)) AS simple_name,
file_type_id
FROM certified_files_history
WHERE certify_history_id = {history_id}
)
UPDATE job
SET filename = CASE WHEN job.file_type_id NOT IN (6, 7)
THEN '{filepath}'
ELSE '{ef_path}'
END || simple_name,
original_filename = CASE WHEN job.file_type_id NOT IN ({remove_timestamp})
THEN simple_name
ELSE substring(simple_name, position('_' in simple_name) + 1)
END
FROM filenames
WHERE job.file_type_id = filenames.file_type_id
AND job.submission_id = {submission_id};
""".format(history_id=max_cert_history[0], filepath=filepath, ef_path=ef_path,
remove_timestamp=', '.join(remove_timestamp), submission_id=submission.submission_id)
sess.execute(update_string)
# Set errors/warnings for the submission
submission.number_of_errors = 0
submission.number_of_warnings =\
sess.query(func.coalesce(func.sum(CertifiedErrorMetadata.occurrences), 0).label('total_warnings')).\
join(Job, CertifiedErrorMetadata.job_id == Job.job_id).\
filter(Job.submission_id == submission.submission_id).one().total_warnings
submission.publishable = True
# Set default numbers/status/last validation date for jobs then update warnings
sess.query(Job).filter_by(submission_id=submission.submission_id).\
update({'number_of_errors': 0, 'number_of_warnings': 0, 'job_status_id': JOB_STATUS_DICT['finished'],
'last_validated': max_cert_history[1], 'error_message': None, 'file_generation_id': None})
# Get list of jobs so we can update them
job_list = sess.query(Job).\
filter(Job.submission_id == submission.submission_id,
Job.job_type_id.in_([JOB_TYPE_DICT['csv_record_validation'], JOB_TYPE_DICT['validation']]),
Job.file_type_id.notin_([FILE_TYPE_DICT['sub_award'], FILE_TYPE_DICT['executive_compensation']])).all()
# Fixing File table
job_ids = [str(job.job_id) for job in job_list]
update_string = """
UPDATE file
SET filename = job.filename,
file_status_id = 1,
headers_missing = NULL,
headers_duplicated = NULL
FROM job
WHERE job.job_id = file.job_id
AND job.job_id IN ({job_ids});
""".format(job_ids=', '.join(job_ids))
sess.execute(update_string)
file_type_mapping = {
FILE_TYPE_DICT['appropriations']: CertifiedAppropriation,
FILE_TYPE_DICT['program_activity']: CertifiedObjectClassProgramActivity,
FILE_TYPE_DICT['award_financial']: CertifiedAwardFinancial,
FILE_TYPE_DICT['award']: CertifiedAwardFinancialAssistance,
FILE_TYPE_DICT['award_procurement']: CertifiedAwardProcurement
}
# Update the number of warnings for each job in the list
for job in job_list:
job.number_of_warnings = sess.query(func.coalesce(func.sum(CertifiedErrorMetadata.occurrences), 0).
label('total_warnings')). \
filter_by(job_id=job.job_id).one().total_warnings
# For non-cross-file jobs, also update the row count and file size
if job.job_type_id != JOB_TYPE_DICT['validation']:
file_type_model = file_type_mapping[job.file_type_id]
total_rows = sess.query(file_type_model).filter_by(submission_id=submission.submission_id).count()
job.number_of_rows = total_rows + 1
job.number_of_rows_valid = total_rows
if file_manager.is_local:
# local file size
try:
job.file_size = os.path.getsize(job.filename)
except:
logger.warning("File doesn't exist locally: %s", job.filename)
job.file_size = 0
else:
# boto file size
job.file_size = S3Handler.get_file_size(job.filename)
# Set submission to certified status
submission.publish_status_id = PUBLISH_STATUS_DICT['published']
sess.commit()
# Move warning files back non-locally and clear out error files for all environments
file_manager.revert_certified_error_files(sess, max_cert_history[0])
return JsonResponse.create(StatusCode.OK, {'message': 'Submission {} successfully reverted to certified status.'.
format(submission.submission_id)})
def run_cross_validation(self, job):
""" Cross file validation job. Test all rules with matching rule_timing.
Run each cross-file rule and create error report.
Args:
job: Current job
"""
sess = GlobalDB.db().session
job_id = job.job_id
# Create File Status object
create_file_if_needed(job_id)
# Create list of errors
error_list = ErrorInterface()
submission_id = job.submission_id
job_start = datetime.now()
logger.info({
'message':
'Beginning cross-file validations on submission_id: ' +
str(submission_id),
'message_type':
'ValidatorInfo',
'submission_id':
submission_id,
'job_id':
job.job_id,
'action':
'run_cross_validations',
'start':
job_start,
'status':
'start'
})
# Delete existing cross file errors for this submission
sess.query(ErrorMetadata).filter(
ErrorMetadata.job_id == job_id).delete()
sess.commit()
# get all cross file rules from db
cross_file_rules = sess.query(RuleSql).filter_by(
rule_cross_file_flag=True)
# for each cross-file combo, run associated rules and create error report
for c in get_cross_file_pairs():
first_file = c[0]
second_file = c[1]
combo_rules = cross_file_rules.filter(
or_(
and_(RuleSql.file_id == first_file.id,
RuleSql.target_file_id == second_file.id),
and_(RuleSql.file_id == second_file.id,
RuleSql.target_file_id == first_file.id)))
# get error file name/path
error_file_name = report_file_name(submission_id, False,
first_file.name,
second_file.name)
error_file_path = "".join(
[CONFIG_SERVICES['error_report_path'], error_file_name])
warning_file_name = report_file_name(submission_id, True,
first_file.name,
second_file.name)
warning_file_path = "".join(
[CONFIG_SERVICES['error_report_path'], warning_file_name])
# open error report and gather failed rules within it
with open(error_file_path, 'w', newline='') as error_file,\
open(warning_file_path, 'w', newline='') as warning_file:
error_csv = csv.writer(error_file,
delimiter=',',
quoting=csv.QUOTE_MINIMAL,
lineterminator='\n')
warning_csv = csv.writer(warning_file,
delimiter=',',
quoting=csv.QUOTE_MINIMAL,
lineterminator='\n')
# write headers to file
error_csv.writerow(self.crossFileReportHeaders)
warning_csv.writerow(self.crossFileReportHeaders)
# send comboRules to validator.crossValidate sql
current_cols_short_to_long = self.short_to_long_dict[
first_file.id].copy()
current_cols_short_to_long.update(
self.short_to_long_dict[second_file.id].copy())
cross_validate_sql(combo_rules.all(), submission_id,
current_cols_short_to_long, first_file.id,
second_file.id, job, error_csv, warning_csv,
error_list, job_id)
# close files
error_file.close()
warning_file.close()
# stream file to S3 when not local
if not self.is_local:
# stream error file
with open(error_file_path, 'rb') as csv_file:
with smart_open.smart_open(
S3Handler.create_file_path(
self.get_file_name(error_file_name)),
'w') as writer:
while True:
chunk = csv_file.read(CHUNK_SIZE)
if chunk:
writer.write(chunk)
else:
break
csv_file.close()
os.remove(error_file_path)
# stream warning file
with open(warning_file_path, 'rb') as warning_csv_file:
with smart_open.smart_open(
S3Handler.create_file_path(
self.get_file_name(warning_file_name)),
'w') as warning_writer:
while True:
chunk = warning_csv_file.read(CHUNK_SIZE)
if chunk:
warning_writer.write(chunk)
else:
break
warning_csv_file.close()
os.remove(warning_file_path)
# write all recorded errors to database
error_list.write_all_row_errors(job_id)
# Update error info for submission
populate_job_error_info(job)
# mark job status as "finished"
mark_job_status(job_id, "finished")
job_duration = (datetime.now() - job_start).total_seconds()
logger.info({
'message':
'Completed cross-file validations on submission_id: ' +
str(submission_id),
'message_type':
'ValidatorInfo',
'submission_id':
submission_id,
'job_id':
job.job_id,
'action':
'run_cross_validations',
'status':
'finish',
'start':
job_start,
'duration':
job_duration
})
# set number of errors and warnings for submission.
submission = populate_submission_error_info(submission_id)
# TODO: Remove temporary step below
# Temporarily set publishable flag at end of cross file, remove this once users are able to mark their
# submissions as publishable
# Publish only if no errors are present
if submission.number_of_errors == 0:
submission.publishable = True
sess.commit()
# Mark validation complete
mark_file_complete(job_id)
def run_validation(self, job):
""" Run validations for specified job
Args:
job: Job to be validated
Returns:
True if successful
"""
sess = GlobalDB.db().session
error_list = ErrorInterface()
job_id = job.job_id
submission_id = job.submission_id
row_number = 1
file_type = job.file_type.name
validation_start = datetime.now()
log_str = 'on submission_id: {}, job_id: {}, file_type: {}'.format(
str(submission_id), str(job_id), file_type)
logger.info({
'message': 'Beginning run_validation {}'.format(log_str),
'message_type': 'ValidatorInfo',
'submission_id': submission_id,
'job_id': job_id,
'file_type': file_type,
'action': 'run_validations',
'status': 'start',
'start_time': validation_start
})
# Get orm model for this file
model = [ft.model for ft in FILE_TYPE if ft.name == file_type][0]
# Delete existing file level errors for this submission
sess.query(ErrorMetadata).filter(
ErrorMetadata.job_id == job_id).delete()
sess.commit()
# Clear existing records for this submission
sess.query(model).filter_by(submission_id=submission_id).delete()
sess.commit()
# Clear existing flex fields for this job
sess.query(FlexField).filter_by(job_id=job_id).delete()
sess.commit()
# If local, make the error report directory
if self.is_local and not os.path.exists(self.directory):
os.makedirs(self.directory)
# Get bucket name and file name
file_name = job.filename
bucket_name = CONFIG_BROKER['aws_bucket']
region_name = CONFIG_BROKER['aws_region']
error_file_name = report_file_name(job.submission_id, False,
job.file_type.name)
error_file_path = "".join(
[CONFIG_SERVICES['error_report_path'], error_file_name])
warning_file_name = report_file_name(job.submission_id, True,
job.file_type.name)
warning_file_path = "".join(
[CONFIG_SERVICES['error_report_path'], warning_file_name])
# Create File Status object
create_file_if_needed(job_id, file_name)
reader = CsvReader()
# Get file size and write to jobs table
if CONFIG_BROKER["use_aws"]:
file_size = S3Handler.get_file_size(file_name)
else:
file_size = os.path.getsize(file_name)
job.file_size = file_size
sess.commit()
# Get fields for this file
fields = sess.query(FileColumn).filter(
FileColumn.file_id == FILE_TYPE_DICT[file_type]).all()
for field in fields:
sess.expunge(field)
csv_schema = {row.name_short: row for row in fields}
try:
extension = os.path.splitext(file_name)[1]
if not extension or extension.lower() not in ['.csv', '.txt']:
raise ResponseException("", StatusCode.CLIENT_ERROR, None,
ValidationError.fileTypeError)
# Count file rows: throws a File Level Error for non-UTF8 characters
temp_file = open(reader.get_filename(region_name, bucket_name,
file_name),
encoding='utf-8')
file_row_count = len(list(csv.reader(temp_file)))
try:
temp_file.close()
except AttributeError:
# File does not exist, and so does not need to be closed
pass
# Pull file and return info on whether it's using short or long col headers
reader.open_file(region_name,
bucket_name,
file_name,
fields,
bucket_name,
self.get_file_name(error_file_name),
self.long_to_short_dict[job.file_type_id],
is_local=self.is_local)
# list to keep track of rows that fail validations
error_rows = []
# While not done, pull one row and put it into staging table if it passes
# the Validator
loading_start = datetime.now()
logger.info({
'message': 'Beginning data loading {}'.format(log_str),
'message_type': 'ValidatorInfo',
'submission_id': submission_id,
'job_id': job_id,
'file_type': file_type,
'action': 'data_loading',
'status': 'start',
'start_time': loading_start
})
with open(error_file_path, 'w', newline='') as error_file,\
open(warning_file_path, 'w', newline='') as warning_file:
error_csv = csv.writer(error_file,
delimiter=',',
quoting=csv.QUOTE_MINIMAL,
lineterminator='\n')
warning_csv = csv.writer(warning_file,
delimiter=',',
quoting=csv.QUOTE_MINIMAL,
lineterminator='\n')
required_list = None
type_list = None
if file_type == "fabs":
# create a list of all required/type labels for FABS
labels = sess.query(ValidationLabel).all()
required_list = {}
type_list = {}
for label in labels:
if label.label_type == "requirement":
required_list[label.column_name] = label.label
else:
type_list[label.column_name] = label.label
# write headers to file
error_csv.writerow(self.reportHeaders)
warning_csv.writerow(self.reportHeaders)
while not reader.is_finished:
row_number += 1
if row_number % 100 == 0:
elapsed_time = (datetime.now() -
loading_start).total_seconds()
logger.info({
'message':
'Loading row: {} {}'.format(
str(row_number), log_str),
'message_type':
'ValidatorInfo',
'submission_id':
submission_id,
'job_id':
job_id,
'file_type':
file_type,
'action':
'data_loading',
'status':
'loading',
'rows_loaded':
row_number,
'start_time':
loading_start,
'elapsed_time':
elapsed_time
})
#
# first phase of validations: read record and record a
# formatting error if there's a problem
#
(record, reduceRow, skip_row, doneReading, rowErrorHere, flex_cols) = \
self.read_record(reader, error_csv, row_number, job, fields, error_list)
if reduceRow:
row_number -= 1
if rowErrorHere:
error_rows.append(row_number)
if doneReading:
# Stop reading from input file
break
elif skip_row:
# Do not write this row to staging, but continue processing future rows
continue
#
# second phase of validations: do basic schema checks
# (e.g., require fields, field length, data type)
#
# D files are obtained from upstream systems (ASP and FPDS) that perform their own basic
# validations, so these validations are not repeated here
if file_type in ["award", "award_procurement"]:
# Skip basic validations for D files, set as valid to trigger write to staging
passed_validations = True
valid = True
else:
if file_type == "fabs":
record['afa_generated_unique'] = (record['award_modification_amendme'] or '-none-') + "_" +\
(record['awarding_sub_tier_agency_c'] or '-none-') + \
"_" + (record['fain'] or '-none-') + "_" + \
(record['uri'] or '-none-')
passed_validations, failures, valid = Validator.validate(
record, csv_schema, file_type == "fabs",
required_list, type_list)
if valid:
# todo: update this logic later when we have actual validations
if file_type == "fabs":
record["is_valid"] = True
model_instance = model(job_id=job_id,
submission_id=submission_id,
valid_record=passed_validations,
**record)
skip_row = not insert_staging_model(
model_instance, job, error_csv, error_list)
if flex_cols:
sess.add_all(flex_cols)
sess.commit()
if skip_row:
error_rows.append(row_number)
continue
if not passed_validations:
fatal = write_errors(
failures, job,
self.short_to_long_dict[job.file_type_id],
error_csv, warning_csv, row_number, error_list,
flex_cols)
if fatal:
error_rows.append(row_number)
loading_duration = (datetime.now() -
loading_start).total_seconds()
logger.info({
'message':
'Completed data loading {}'.format(log_str),
'message_type':
'ValidatorInfo',
'submission_id':
submission_id,
'job_id':
job_id,
'file_type':
file_type,
'action':
'data_loading',
'status':
'finish',
'start_time':
loading_start,
'end_time':
datetime.now(),
'duration':
loading_duration,
'total_rows':
row_number
})
if file_type in ('appropriations', 'program_activity',
'award_financial'):
update_tas_ids(model, submission_id)
#
# third phase of validations: run validation rules as specified
# in the schema guidance. these validations are sql-based.
#
sql_error_rows = self.run_sql_validations(
job, file_type, self.short_to_long_dict[job.file_type_id],
error_csv, warning_csv, row_number, error_list)
error_rows.extend(sql_error_rows)
error_file.close()
warning_file.close()
# stream file to S3 when not local
if not self.is_local:
# stream error file
with open(error_file_path, 'rb') as csv_file:
with smart_open.smart_open(S3Handler.create_file_path(self.get_file_name(error_file_name)), 'w')\
as writer:
while True:
chunk = csv_file.read(CHUNK_SIZE)
if chunk:
writer.write(chunk)
else:
break
csv_file.close()
os.remove(error_file_path)
# stream warning file
with open(warning_file_path, 'rb') as warning_csv_file:
with smart_open.smart_open(S3Handler.create_file_path(self.get_file_name(warning_file_name)), 'w')\
as warning_writer:
while True:
chunk = warning_csv_file.read(CHUNK_SIZE)
if chunk:
warning_writer.write(chunk)
else:
break
warning_csv_file.close()
os.remove(warning_file_path)
# Calculate total number of rows in file
# that passed validations
error_rows_unique = set(error_rows)
total_rows_excluding_header = row_number - 1
valid_rows = total_rows_excluding_header - len(error_rows_unique)
# Update fabs is_valid rows where applicable
# Update submission to include action dates where applicable
if file_type == "fabs":
sess.query(DetachedAwardFinancialAssistance).\
filter(DetachedAwardFinancialAssistance.row_number.in_(error_rows_unique),
DetachedAwardFinancialAssistance.submission_id == submission_id).\
update({"is_valid": False}, synchronize_session=False)
sess.commit()
min_action_date, max_action_date = get_action_dates(
submission_id)
sess.query(Submission).filter(Submission.submission_id == submission_id).\
update({"reporting_start_date": min_action_date, "reporting_end_date": max_action_date},
synchronize_session=False)
# Ensure validated rows match initial row count
if file_row_count != row_number:
raise ResponseException("", StatusCode.CLIENT_ERROR, None,
ValidationError.rowCountError)
# Update job metadata
job.number_of_rows = row_number
job.number_of_rows_valid = valid_rows
sess.commit()
error_list.write_all_row_errors(job_id)
# Update error info for submission
populate_job_error_info(job)
if file_type == "fabs":
# set number of errors and warnings for detached submission
populate_submission_error_info(submission_id)
# Mark validation as finished in job tracker
mark_job_status(job_id, "finished")
mark_file_complete(job_id, file_name)
finally:
# Ensure the files always close
reader.close()
validation_duration = (datetime.now() -
validation_start).total_seconds()
logger.info({
'message':
'Completed run_validation {}'.format(log_str),
'message_type':
'ValidatorInfo',
'submission_id':
submission_id,
'job_id':
job_id,
'file_type':
file_type,
'action':
'run_validation',
'status':
'finish',
'start_time':
validation_start,
'end_time':
datetime.now(),
'duration':
validation_duration
})
return True
def copy_file_generation_to_job(job, file_generation, is_local):
""" Copy cached FileGeneration data to a Job requesting a file.
Args:
job: Job object to copy the data to
file_generation: Cached FileGeneration object to copy the data from
is_local: A boolean flag indicating whether the application is being run locally or not
"""
sess = GlobalDB.db().session
log_data = {
'message':
'Copying FileGeneration {} data to Job {}'.format(
file_generation.file_generation_id, job.job_id),
'message_type':
'BrokerInfo',
'job_id':
job.job_id,
'file_type':
job.file_type.name,
'file_generation_id':
file_generation.file_generation_id
}
logger.info(log_data)
# Do not edit submissions that have already successfully completed
sess.refresh(job)
if job.job_status_id == lookups.JOB_STATUS_DICT['finished']:
return
job.file_generation_id = file_generation.file_generation_id
# File is still being generated, just mark the FileGeneration ID in the Job and wait
# FileGeneration will update all child Jobs when it finishes
if not file_generation.file_path:
sess.commit()
return
# Generate file path for child Job's filename
filepath = CONFIG_BROKER['broker_files'] if g.is_local else "{}/".format(
str(job.submission_id))
original_filename = file_generation.file_path.split('/')[-1]
filename = '{}{}'.format(filepath, original_filename)
# Copy parent job's data
job.filename = filename
job.original_filename = original_filename
job.number_of_errors = 0
job.number_of_warnings = 0
# Change the validation job's file data when within a submission
if job.submission_id is not None:
val_job = sess.query(Job).filter(
Job.submission_id == job.submission_id,
Job.file_type_id == job.file_type_id, Job.job_type_id ==
lookups.JOB_TYPE_DICT['csv_record_validation']).one()
val_job.filename = filename
val_job.original_filename = original_filename
# Copy the data to the Submission's bucket
if not g.is_local and file_generation.file_path != job.filename:
# Check to see if the same file exists in the child bucket
s3 = boto3.client('s3', region_name=CONFIG_BROKER["aws_region"])
bucket = CONFIG_BROKER['aws_bucket']
response = s3.list_objects_v2(Bucket=bucket, Prefix=job.filename)
for obj in response.get('Contents', []):
if obj['Key'] == job.filename:
# The file already exists in this location
log_data[
'message'] = '{} file already exists in this location: {}; not overwriting.'.format(
job.file_type.name, job.filename)
logger.info(log_data)
mark_job_status(job.job_id, 'finished')
return
S3Handler.copy_file(bucket, bucket, file_generation.file_path,
job.filename)
sess.commit()
# Mark Job status last so the validation job doesn't start until everything is done
mark_job_status(job.job_id, 'finished')
def run_validation(self, job):
""" Run validations for specified job
Args:
job: Job to be validated
Returns:
True if successful
"""
sess = GlobalDB.db().session
self.job = job
self.submission_id = job.submission_id
self.file_type = job.file_type
self.file_name = job.filename
self.is_fabs = (self.file_type.name == 'fabs')
# initializing processing metadata vars for a new validation
self.reader = CsvReader()
self.error_list = ErrorInterface()
self.error_rows = []
self.max_row_number = 1
self.total_rows = 0
self.short_rows = []
self.long_rows = []
validation_start = datetime.now()
bucket_name = CONFIG_BROKER['aws_bucket']
region_name = CONFIG_BROKER['aws_region']
self.log_str = 'on submission_id: {}, job_id: {}, file_type: {}'.format(
str(self.submission_id), str(self.job.job_id), self.file_type.name)
logger.info({
'message': 'Beginning run_validation {}'.format(self.log_str),
'message_type': 'ValidatorInfo',
'submission_id': self.submission_id,
'job_id': self.job.job_id,
'file_type': self.file_type.name,
'action': 'run_validations',
'status': 'start',
'start_time': validation_start
})
# Get orm model for this file
self.model = [ft.model for ft in FILE_TYPE if ft.name == self.file_type.name][0]
# Delete existing file level errors for this submission
sess.query(ErrorMetadata).filter(ErrorMetadata.job_id == self.job.job_id).delete()
sess.commit()
# Clear existing records for this submission
sess.query(self.model).filter_by(submission_id=self.submission_id).delete()
sess.commit()
# Clear existing flex fields for this job
sess.query(FlexField).filter_by(job_id=self.job.job_id).delete()
sess.commit()
# If local, make the error report directory
if self.is_local and not os.path.exists(self.directory):
os.makedirs(self.directory)
create_file_if_needed(self.job.job_id, self.file_name)
# Get file size and write to jobs table
if CONFIG_BROKER['use_aws']:
file_size = S3Handler.get_file_size(self.file_name)
else:
file_size = os.path.getsize(self.file_name)
self.job.file_size = file_size
sess.commit()
# Get fields for this file
self.fields = sess.query(FileColumn).filter(FileColumn.file_id == FILE_TYPE_DICT[self.file_type.name])\
.order_by(FileColumn.daims_name.asc()).all()
self.expected_headers, self.parsed_fields = parse_fields(sess, self.fields)
self.csv_schema = {row.name_short: row for row in self.fields}
try:
# Loading data and initial validations
self.load_file_data(sess, bucket_name, region_name)
if self.file_type.name in ('appropriations', 'program_activity', 'award_financial'):
update_tas_ids(self.model, self.submission_id)
# SQL Validations
with open(self.error_file_path, 'a', newline='') as error_file, \
open(self.warning_file_path, 'a', newline='') as warning_file:
error_csv = csv.writer(error_file, delimiter=',', quoting=csv.QUOTE_MINIMAL, lineterminator='\n')
warning_csv = csv.writer(warning_file, delimiter=',', quoting=csv.QUOTE_MINIMAL, lineterminator='\n')
# third phase of validations: run validation rules as specified in the schema guidance. These
# validations are sql-based.
sql_error_rows = self.run_sql_validations(self.short_to_long_dict[self.file_type.file_type_id],
error_csv, warning_csv)
self.error_rows.extend(sql_error_rows)
error_file.close()
warning_file.close()
# stream file to S3 when not local
if not self.is_local:
s3_resource = boto3.resource('s3', region_name=region_name)
# stream error file
with open(self.error_file_path, 'rb') as csv_file:
s3_resource.Object(bucket_name, self.get_file_name(self.error_file_name)).put(Body=csv_file)
csv_file.close()
os.remove(self.error_file_path)
# stream warning file
with open(self.warning_file_path, 'rb') as warning_csv_file:
s3_resource.Object(bucket_name,
self.get_file_name(self.warning_file_name)).put(Body=warning_csv_file)
warning_csv_file.close()
os.remove(self.warning_file_path)
# Calculate total number of rows in file that passed validations
error_rows_unique = set(self.error_rows)
total_rows_excluding_header = self.total_rows - 1
valid_rows = total_rows_excluding_header - len(error_rows_unique)
# Update fabs is_valid rows where applicable
# Update submission to include action dates where applicable
if self.is_fabs:
sess.query(DetachedAwardFinancialAssistance). \
filter(DetachedAwardFinancialAssistance.row_number.in_(error_rows_unique),
DetachedAwardFinancialAssistance.submission_id == self.submission_id). \
update({'is_valid': False}, synchronize_session=False)
sess.commit()
min_action_date, max_action_date = get_action_dates(self.submission_id)
sess.query(Submission).filter(Submission.submission_id == self.submission_id). \
update({'reporting_start_date': min_action_date, 'reporting_end_date': max_action_date},
synchronize_session=False)
# Update job metadata
self.job.number_of_rows = self.total_rows
self.job.number_of_rows_valid = valid_rows
sess.commit()
self.error_list.write_all_row_errors(self.job.job_id)
# Update error info for submission
populate_job_error_info(self.job)
if self.is_fabs:
# set number of errors and warnings for detached submission
populate_submission_error_info(self.submission_id)
# Mark validation as finished in job tracker
mark_job_status(self.job.job_id, 'finished')
mark_file_complete(self.job.job_id, self.file_name)
except Exception:
logger.error({
'message': 'An exception occurred during validation',
'message_type': 'ValidatorInfo',
'submission_id': self.submission_id,
'job_id': self.job.job_id,
'file_type': self.file_type.name,
'traceback': traceback.format_exc()
})
raise
finally:
# Ensure the files always close
self.reader.close()
validation_duration = (datetime.now()-validation_start).total_seconds()
logger.info({
'message': 'Completed run_validation {}'.format(self.log_str),
'message_type': 'ValidatorInfo',
'submission_id': self.submission_id,
'job_id': self.job.job_id,
'file_type': self.file_type.name,
'action': 'run_validation',
'status': 'finish',
'start_time': validation_start,
'end_time': datetime.now(),
'duration': validation_duration
})
return True
def run_validation(self, job):
""" Run validations for specified job
Args:
job: Job to be validated
Returns:
True if successful
"""
sess = GlobalDB.db().session
job_id = job.job_id
error_list = ErrorInterface()
submission_id = job.submission_id
row_number = 1
file_type = job.file_type.name
validation_start = datetime.now()
logger.info(
{
'message': 'Beginning run_validation on submission_id: ' + str(submission_id) +
', job_id: ' + str(job_id) + ', file_type: ' + file_type,
'message_type': 'ValidatorInfo',
'submission_id': submission_id,
'job_id': job_id,
'file_type': file_type,
'action': 'run_validations',
'status': 'start',
'start_time': validation_start})
# Get orm model for this file
model = [ft.model for ft in FILE_TYPE if ft.name == file_type][0]
# Delete existing file level errors for this submission
sess.query(ErrorMetadata).filter(ErrorMetadata.job_id == job_id).delete()
sess.commit()
# Clear existing records for this submission
sess.query(model).filter_by(submission_id=submission_id).delete()
sess.commit()
# Clear existing flex fields for this job
sess.query(FlexField).filter_by(job_id=job_id).delete()
sess.commit()
# If local, make the error report directory
if self.isLocal and not os.path.exists(self.directory):
os.makedirs(self.directory)
# Get bucket name and file name
file_name = job.filename
bucket_name = CONFIG_BROKER['aws_bucket']
region_name = CONFIG_BROKER['aws_region']
error_file_name = self.get_file_name(report_file_name(job.submission_id, False, job.file_type.name))
warning_file_name = self.get_file_name(report_file_name(job.submission_id, True, job.file_type.name))
# Create File Status object
create_file_if_needed(job_id, file_name)
reader = self.get_reader()
# Get file size and write to jobs table
if CONFIG_BROKER["use_aws"]:
file_size = S3Handler.get_file_size(file_name)
else:
file_size = os.path.getsize(file_name)
job.file_size = file_size
sess.commit()
# Get fields for this file
fields = sess.query(FileColumn).filter(FileColumn.file_id == FILE_TYPE_DICT[file_type]).all()
for field in fields:
sess.expunge(field)
csv_schema = {row.name_short: row for row in fields}
try:
extension = os.path.splitext(file_name)[1]
if not extension or extension not in ['.csv', '.txt']:
raise ResponseException("", StatusCode.CLIENT_ERROR, None, ValidationError.fileTypeError)
# Count file rows: throws a File Level Error for non-UTF8 characters
temp_file = open(reader.get_filename(region_name, bucket_name, file_name), encoding='utf-8')
file_row_count = len(list(csv.reader(temp_file)))
try:
temp_file.close()
except AttributeError:
# File does not exist, and so does not need to be closed
pass
# Pull file and return info on whether it's using short or long col headers
reader.open_file(region_name, bucket_name, file_name, fields, bucket_name, error_file_name,
self.long_to_short_dict, is_local=self.isLocal)
# list to keep track of rows that fail validations
error_rows = []
# While not done, pull one row and put it into staging table if it passes
# the Validator
loading_start = datetime.now()
logger.info(
{
'message': 'Beginning data loading on submission_id: ' + str(submission_id) +
', job_id: ' + str(job_id) + ', file_type: ' + file_type,
'message_type': 'ValidatorInfo',
'submission_id': submission_id,
'job_id': job_id,
'file_type': file_type,
'action': 'data_loading',
'status': 'start',
'start_time': loading_start})
with self.get_writer(region_name, bucket_name, error_file_name, self.reportHeaders) as writer, \
self.get_writer(region_name, bucket_name, warning_file_name, self.reportHeaders) as warning_writer:
while not reader.is_finished:
row_number += 1
if row_number % 100 == 0:
elapsed_time = (datetime.now()-loading_start).total_seconds()
logger.info(
{
'message': 'Loading row: ' + str(row_number) + ' on submission_id: ' +
str(submission_id) + ', job_id: ' + str(job_id) + ', file_type: ' + file_type,
'message_type': 'ValidatorInfo',
'submission_id': submission_id,
'job_id': job_id,
'file_type': file_type,
'action': 'data_loading',
'status': 'loading',
'rows_loaded': row_number,
'start_time': loading_start,
'elapsed_time': elapsed_time})
#
# first phase of validations: read record and record a
# formatting error if there's a problem
#
(record, reduceRow, skip_row, doneReading, rowErrorHere, flex_cols) = \
self.read_record(reader, writer, row_number, job, fields, error_list)
if reduceRow:
row_number -= 1
if rowErrorHere:
error_rows.append(row_number)
if doneReading:
# Stop reading from input file
break
elif skip_row:
# Do not write this row to staging, but continue processing future rows
continue
#
# second phase of validations: do basic schema checks
# (e.g., require fields, field length, data type)
#
# D files are obtained from upstream systems (ASP and FPDS) that perform their own basic
# validations, so these validations are not repeated here
if file_type in ["award", "award_procurement"]:
# Skip basic validations for D files, set as valid to trigger write to staging
passed_validations = True
valid = True
else:
if file_type in ["detached_award"]:
record['afa_generated_unique'] = (record['award_modification_amendme'] or '-none-') + \
(record['awarding_sub_tier_agency_c'] or '-none-') + \
(record['fain'] or '-none-') + (record['uri'] or '-none-')
passed_validations, failures, valid = Validator.validate(record, csv_schema,
file_type in ["detached_award"])
if valid:
# todo: update this logic later when we have actual validations
if file_type in ["detached_award"]:
record["is_valid"] = True
model_instance = model(job_id=job_id, submission_id=submission_id,
valid_record=passed_validations, **record)
skip_row = not insert_staging_model(model_instance, job, writer, error_list)
if flex_cols:
sess.add_all(flex_cols)
sess.commit()
if skip_row:
error_rows.append(row_number)
continue
if not passed_validations:
fatal = write_errors(failures, job, self.short_to_long_dict, writer, warning_writer,
row_number, error_list)
if fatal:
error_rows.append(row_number)
loading_duration = (datetime.now()-loading_start).total_seconds()
logger.info(
{
'message': 'Completed data loading on submission_id: ' + str(submission_id) +
', job_id: ' + str(job_id) + ', file_type: ' + file_type,
'message_type': 'ValidatorInfo',
'submission_id': submission_id,
'job_id': job_id,
'file_type': file_type,
'action': 'data_loading',
'status': 'finish',
'start_time': loading_start,
'end_time': datetime.now(),
'duration': loading_duration,
'total_rows': row_number
})
if file_type in ('appropriations', 'program_activity', 'award_financial'):
update_tas_ids(model, submission_id)
#
# third phase of validations: run validation rules as specified
# in the schema guidance. these validations are sql-based.
#
sql_error_rows = self.run_sql_validations(job, file_type, self.short_to_long_dict, writer,
warning_writer, row_number, error_list)
error_rows.extend(sql_error_rows)
# Write unfinished batch
writer.finish_batch()
warning_writer.finish_batch()
# Calculate total number of rows in file
# that passed validations
error_rows_unique = set(error_rows)
total_rows_excluding_header = row_number - 1
valid_rows = total_rows_excluding_header - len(error_rows_unique)
# Update detached_award is_valid rows where applicable
# Update submission to include action dates where applicable
if file_type in ["detached_award"]:
sess.query(DetachedAwardFinancialAssistance).\
filter(DetachedAwardFinancialAssistance.row_number.in_(error_rows_unique),
DetachedAwardFinancialAssistance.submission_id == submission_id).\
update({"is_valid": False}, synchronize_session=False)
sess.commit()
min_action_date, max_action_date = get_action_dates(submission_id)
sess.query(Submission).filter(Submission.submission_id == submission_id).\
update({"reporting_start_date": min_action_date, "reporting_end_date": max_action_date},
synchronize_session=False)
# Ensure validated rows match initial row count
if file_row_count != row_number:
raise ResponseException("", StatusCode.CLIENT_ERROR, None, ValidationError.rowCountError)
# Update job metadata
job.number_of_rows = row_number
job.number_of_rows_valid = valid_rows
sess.commit()
error_list.write_all_row_errors(job_id)
# Update error info for submission
populate_job_error_info(job)
if file_type in ["detached_award"]:
# set number of errors and warnings for detached submission
populate_submission_error_info(submission_id)
# Mark validation as finished in job tracker
mark_job_status(job_id, "finished")
mark_file_complete(job_id, file_name)
finally:
# Ensure the file always closes
reader.close()
validation_duration = (datetime.now()-validation_start).total_seconds()
logger.info(
{
'message': 'Completed run_validation on submission_id: ' + str(submission_id) +
', job_id: ' + str(job_id) + ', file_type: ' + file_type,
'message_type': 'ValidatorInfo',
'submission_id': submission_id,
'job_id': job_id,
'file_type': file_type,
'action': 'run_validation',
'status': 'finish',
'start_time': validation_start,
'end_time': datetime.now(),
'duration': validation_duration
})
return True
def copy_file_generation_to_job(job, file_generation, is_local):
""" Copy cached FileGeneration data to a Job requesting a file.
Args:
job: Job object to copy the data to
file_generation: Cached FileGeneration object to copy the data from
is_local: A boolean flag indicating whether the application is being run locally or not
"""
sess = GlobalDB.db().session
log_data = {
'message': 'Copying FileGeneration {} data to Job {}'.format(file_generation.file_generation_id, job.job_id),
'message_type': 'BrokerInfo', 'job_id': job.job_id, 'file_type': job.file_type.name,
'file_generation_id': file_generation.file_generation_id}
logger.info(log_data)
# Do not edit submissions that have already successfully completed
sess.refresh(job)
if job.job_status_id == lookups.JOB_STATUS_DICT['finished']:
return
job.file_generation_id = file_generation.file_generation_id
# File is still being generated, just mark the FileGeneration ID in the Job and wait
# FileGeneration will update all child Jobs when it finishes
if not file_generation.file_path:
sess.commit()
return
# Generate file path for child Job's filename
filepath = CONFIG_BROKER['broker_files'] if g.is_local else "{}/".format(str(job.submission_id))
original_filename = file_generation.file_path.split('/')[-1]
filename = '{}{}'.format(filepath, original_filename)
# Copy parent job's data
job.filename = filename
job.original_filename = original_filename
job.number_of_errors = 0
job.number_of_warnings = 0
# Change the validation job's file data when within a submission
if job.submission_id is not None:
val_job = sess.query(Job).filter(Job.submission_id == job.submission_id,
Job.file_type_id == job.file_type_id,
Job.job_type_id == lookups.JOB_TYPE_DICT['csv_record_validation']).one()
val_job.filename = filename
val_job.original_filename = original_filename
# Copy the data to the Submission's bucket
if not g.is_local and file_generation.file_path != job.filename:
# Check to see if the same file exists in the child bucket
s3 = boto3.client('s3', region_name=CONFIG_BROKER["aws_region"])
bucket = CONFIG_BROKER['aws_bucket']
response = s3.list_objects_v2(Bucket=bucket, Prefix=job.filename)
for obj in response.get('Contents', []):
if obj['Key'] == job.filename:
# The file already exists in this location
log_data['message'] = '{} file already exists in this location: {}; not overwriting.'.format(
job.file_type.name, job.filename)
logger.info(log_data)
mark_job_status(job.job_id, 'finished')
return
S3Handler.copy_file(bucket, bucket, file_generation.file_path, job.filename)
sess.commit()
# Mark Job status last so the validation job doesn't start until everything is done
mark_job_status(job.job_id, 'finished')
def copy_parent_file_request_data(sess, child_job, parent_job, is_local):
"""Parent FileRequest job data to the child FileRequest job data.
Args:
sess: current DB session
child_job: Job ID for the child FileRequest object
parent_job: Job ID for the parent FileRequest object
is_local: True if in local development, False otherwise
"""
file_type = parent_job.file_type.letter_name
log_data = {
'message':
'Copying data from parent job with job_id:{}'.format(
parent_job.job_id),
'message_type':
'ValidatorInfo',
'job_id':
child_job.job_id,
'file_type':
parent_job.file_type.name
}
# Keep path but update file name
filename = '{}/{}'.format(
child_job.filename.rsplit('/', 1)[0], parent_job.original_filename)
# Copy parent job's data
child_job.from_cached = True
child_job.filename = filename
child_job.original_filename = parent_job.original_filename
child_job.number_of_errors = parent_job.number_of_errors
child_job.number_of_warnings = parent_job.number_of_warnings
child_job.error_message = parent_job.error_message
# Change the validation job's file data when within a submission
if child_job.submission_id is not None:
val_job = sess.query(Job).filter(
Job.submission_id == child_job.submission_id,
Job.file_type_id == parent_job.file_type_id,
Job.job_type_id == JOB_TYPE_DICT['csv_record_validation']).one()
val_job.filename = filename
val_job.original_filename = parent_job.original_filename
sess.commit()
if not is_local and parent_job.filename != child_job.filename:
# Check to see if the same file exists in the child bucket
s3 = boto3.client('s3', region_name=CONFIG_BROKER["aws_region"])
response = s3.list_objects_v2(Bucket=CONFIG_BROKER['aws_bucket'],
Prefix=child_job.filename)
for obj in response.get('Contents', []):
if obj['Key'] == child_job.filename:
# The file already exists in this location
log_data[
'message'] = 'Cached {} file CSV already exists in this location'.format(
file_type)
logger.info(log_data)
return
# Copy the parent file into the child's S3 location
log_data['message'] = 'Copying the cached {} file from job {}'.format(
file_type, parent_job.job_id)
logger.info(log_data)
with smart_open.smart_open(
S3Handler.create_file_path(parent_job.filename),
'r') as reader:
stream_file_to_s3(child_job.filename, reader)
# Mark job status last so the validation job doesn't start until everything is done
mark_job_status(child_job.job_id,
JOB_STATUS_DICT_ID[parent_job.job_status_id])
max_cert_id = sess.query(func.max(CertifyHistory.certify_history_id).label('cert_id')). \
filter_by(submission_id=submission_id).one()
route_vars = [
agency_code, submission.reporting_fiscal_year,
submission.reporting_fiscal_period // 3, max_pub_id.pub_id
]
new_route = '/'.join([str(var) for var in route_vars]) + '/'
if not is_local:
old_path = '{}/{}'.format(str(submission_id), filename)
new_path = new_route + filename
# Copy the file if it's a non-local submission
S3Handler().copy_file(
original_bucket=CONFIG_BROKER['aws_bucket'],
new_bucket=CONFIG_BROKER['certified_bucket'],
original_path=old_path,
new_path=new_path)
else:
new_path = "".join([CONFIG_BROKER['broker_files'], filename])
# add published history
file_history = PublishedFilesHistory(
publish_history_id=max_pub_id.pub_id,
certify_history_id=max_cert_id.cert_id,
submission_id=submission_id,
filename=new_path,
file_type_id=None,
comment=None,
warning_filename=None)
sess.add(file_history)
