def backup_media_to_s3(firebase_db):
forbidden_events = []
# get all events in firebase that do not have a 'backup_link'
for event in firebase_db.get_events_that_need_backup_iter():
event_id = event['id']
src_url = event['source_url']
media_type = event['type']
try:
filename, filepath = file_utils.download_file_from_url(src_url, media_type, event_id)
s3_utils.upload(filepath, key=filename)
# delete local file
file_utils.delete_file(filepath)
# update firebase with 'backup_link'
event['backup_link'] = filename
firebase_db.update(event)
except HTTPError as e:
if e.code == 403:
# very infrequently, a 403 is retured from the download link
# so this event is useless, delete it from firebase
logger.warn('This event has media that is forbidden. Deleting event {}'.format(event))
firebase_db.delete(event['id'])
forbidden_events.append(event)
logger.warn('Deleted event {}'.format(event))
logger.warn('Deleted {} events from firebase: {}'.format(len(forbidden_events), forbidden_events)) if forbidden_events else logger.info('')
logger.info('finished backup all media')
def __init__(self):
self.queue = Queue.Queue()
self.docker_client = client.Client()
config_path = os.environ.get(SearchAndValidate.CONFIG_ENV_NAME) or SearchAndValidate.CONFIG_FILE_PATH
self.config = yaml.safe_load(open(config_path))
self.pulled_images = []
self.failed_images = []
file_utils.delete_file()
def _save_result(self):
file_utils.delete_file()
with open('./results.txt', mode='w') as out_file:
if len(self.failed_images) > 0:
print >>out_file, 'Following images could not be pulled'
print >>out_file, '\n'.join(self.failed_images)
if len(self.pulled_images) > 0:
print >>out_file, '\nFollowing images have been pulled'
print >>out_file, '\n'.join(self.pulled_images)
def predict(path):
print('Got prediction request.')
remote_addr = request.remote_addr
filename = download_file('http://' + remote_addr + ':3001/public/' + path)
prediction = model.predict_by_path('uploads/' + filename)
prediction['exterior'] = prediction['exterior'].item()
prediction['interior'] = prediction['interior'].item()
prediction = json.dumps(prediction)
delete_file('uploads/' + filename)
print('Sending prediction response.')
return prediction, 200
def flush_old_records(self):
'''deletes old runs, associated backup files and orders (deleting runs delete cascade associated orders)'''
old_runs = self._get_old_runs()
try:
for run in old_runs:
orders_in_run = self.session.query(Order).filter_by(
run_obj=run).all()
logging.info(
f'Deleting {len(orders_in_run)} orders associated with old {run} and backup file: {run.fpath}'
)
delete_file(run.fpath)
self.session.delete(run)
self.session.commit()
except Exception as e:
logging.warning(
f'Unexpected err while flushing old records from db inside flush_old_records. Err: {e}. Last recorded run {run}'
)
def ingest():
file_utils.download_and_extract(
config.NSL_ARCHIVE_URL,
config.NSL_DOWNLOAD_LOCATION)
xml_obj = open_and_objectify(os.path.join(config.NSL_DOWNLOAD_LOCATION, "SEnsl_ssi.xml"))
nsl_other_xml_obj = open_and_objectify(os.path.join(config.NSL_DOWNLOAD_LOCATION, "SEnsl_other.xml"))
nsl_dict = NslConverter().convert(xml_obj)
nsl_other_dict = NslOtherConverter().convert(nsl_other_xml_obj)
dump_nsl(nsl_dict)
dump_nsl_other(nsl_other_dict)
dump_dictionaries()
file_utils.cleanup(config.NSL_DOWNLOAD_LOCATION)
file_utils.zipdir(config.NSL_JSON_OUTPUT_PATH, config.NSL_ARCHIVE_NAME)
file_utils.cleanup(config.NSL_JSON_OUTPUT_PATH)
file_utils.upload_file(config.NSL_DIFF_UPDATE_URL, config.NSL_ARCHIVE_NAME)
file_utils.delete_file(config.NSL_ARCHIVE_NAME)
import image_check_exceptions
__author__ = 'A.P. Rajshekhar'
import validate_images
import argparse
from image_check_exceptions import ImageCheckException
import file_utils
def parse_arg():
parser = argparse.ArgumentParser()
parser.add_argument("--env", help="the environment against which to run search")
args = parser.parse_args()
if args.env:
return args.env
else:
return 'ci'
if __name__ == '__main__':
check_image = validate_images.SearchAndValidate()
try:
check_image.start_check(parse_arg())
except ImageCheckException, e:
raise
except:
file_utils.delete_file()
file_utils.write("Build has thrown exception not related to pull of image. Please check.")
def delete_old_files(self):
'''addresses potential double loading of same LP sheets to Excel problem,
deletes csv files from Helper Files dir before new run'''
delete_file(self.lp_filename)
delete_file(self.lp_tracked_filename)
origin_url = 'https://hwcdn.libsyn.com/p/f/f/7/ff7f1a6e683c7dbb/37-_Go_East_Young_Man.mp3?c_id=1453698&cs_id=1453698&expiration=1590267009&hwt=216c6584f6ef2a42be776e773cc7d40b'
bucket = 'test-text-to-speech'
filename = 'bronze-age-collapse.mp3'
# print('Downloading file...')
file_utils.download_file_from_url(origin_url, filename)
# print('...done')
print('Uploading to s3...')
response = s3_utils.upload_file(filename, 'test-text-to-speech')
print(response)
print('...done')
print('Deleting local copy...')
file_utils.delete_file(filename)
print('...done')
job_name = 'job-%s' % uuid.uuid1()
print(job_name)
object_url = "https://%s.s3.amazonaws.com/%s" % (bucket, filename)
print(object_url)
print('Starting transcription job...')
response = sound_stuff.start_transcription_job(job_name, object_url, 'mp3')
print(response)
print('...done')
print('Waiting on transcription task...')
sound_stuff.wait_for_transaction_job(job_name)
normalize_fx = (
AudioEffectsChain()
.compand(**m_settings.compand)
.normalize()
)
# perform audio effects on infile and save results to effected files
normalized_file = file_utils.get_test_outfile_path()
normalize_fx(infile, normalized_file, sample_out=48000, channels_out=1)
# STEP 3: Isolate speech and make non-speech frames silent
isolated_speech_file = isolate_speech(normalized_file)
file_utils.delete_file(normalized_file)
# STEP 4: eq the signal
eq_fx = (
AudioEffectsChain()
.highpass(**eq_settings.highpass)
.lowshelf(**eq_settings.lowShelf)
.equalizer(**eq_settings.lowFrequency)
.equalizer(**eq_settings.midFrequency)
.highshelf(**eq_settings.highShelf)
# .reverb(**m_settings.reverb)
)
eq_file = file_utils.get_test_outfile_path()
eq_fx(isolated_speech_file, eq_file, sample_out=48000, channels_out=1)
