def create_job(client,
resource_group_name,
account_name,
transform_name,
job_name,
output_asset_names,
input_asset_name=None,
description=None,
priority=None,
files=None,
base_uri=None):
from azure.mgmt.media.models import (Job, JobInputAsset, JobInputHttp,
JobOutputAsset)
if input_asset_name:
job_input = JobInputAsset(asset_name=input_asset_name, files=files)
else:
if base_uri is None and files is None:
raise CLIError(
"Missing required arguments.\nEither --input-asset-name, "
"or both --files or --base-uri must be specified.")
else:
job_input = JobInputHttp(files=files, base_uri=base_uri)
job_outputs = list(
map(lambda x: JobOutputAsset(asset_name=x), output_asset_names))
job = Job(input=job_input,
outputs=job_outputs,
description=description,
priority=priority)
return client.create(resource_group_name, account_name, transform_name,
job_name, job)
def _upload_and_encode(file):
# clean up inputs
basename = file.filename.replace('.', '-')
basename = basename.replace(' ', '-')
in_asset = "in__" + basename
print(in_asset)
# Create new input asset
client.assets.create_or_update(RESOURCE_GROUP_NAME, ACCOUNT_NAME, in_asset,
Asset())
#Get asset container and upload file to it
response = client.assets.get(RESOURCE_GROUP_NAME, ACCOUNT_NAME, in_asset)
print("got a response, know what kind of container??")
print(response.container)
blob_service.create_blob_from_stream(response.container,
basename,
stream=file.stream)
#Create output asset and add streaming locator name to it as metadata
out_asset = "out__" + basename
locator_name = "loc__" + basename
client.assets.create_or_update(RESOURCE_GROUP_NAME, ACCOUNT_NAME,
out_asset, Asset(alternate_id=locator_name))
#Check if job exists, delete it if it does
job_name = "job__" + basename
response = client.jobs.get(RESOURCE_GROUP_NAME, ACCOUNT_NAME,
TRANSFORM_NAME, job_name)
if response:
client.jobs.delete(RESOURCE_GROUP_NAME, ACCOUNT_NAME, TRANSFORM_NAME,
job_name)
#Create new job with corret inputs and outputs
inputs = JobInputAsset(asset_name=in_asset)
outputs = [JobOutputAsset(asset_name=out_asset)]
new_job = Job(input=inputs, outputs=outputs)
print("Creating new job")
client.jobs.create(RESOURCE_GROUP_NAME, ACCOUNT_NAME, TRANSFORM_NAME,
job_name, new_job)
def create_job(client,
resource_group_name,
account_name,
transform_name,
job_name,
output_assets,
input_asset_name=None,
label=None,
correlation_data=None,
description=None,
priority=None,
files=None,
base_uri=None):
from azure.mgmt.media.models import (Job, JobInputAsset, JobInputHttp)
if input_asset_name:
job_input = JobInputAsset(asset_name=input_asset_name,
files=files,
label=label)
else:
if base_uri is not None and files is not None:
job_input = JobInputHttp(files=files,
base_uri=base_uri,
label=label)
else:
raise CLIError(
"Missing required arguments.\nEither --input-asset-name, "
"or both --files or --base-uri must be specified.")
job_outputs = output_assets
job = Job(input=job_input,
outputs=job_outputs,
correlation_data=correlation_data,
description=description,
priority=priority)
return client.create(resource_group_name, account_name, transform_name,
job_name, job)
audio_job_name = 'AudioAnalyticsJob' + uniqueness
video_job_name = 'VideoAnalyticsJob' + uniqueness
# Choose which of the analyzer job name you would like to Use
# For the audio analytics job, use audio_job_name
# For the video analytics job, use video_job_name
analysis_job_name = audio_job_name
print(f"Creating Analytics job {analysis_job_name}")
files = (source_file)
# Create Job Input and Job Output Asset
input = JobInputAsset(asset_name=in_asset_name)
outputs = JobOutputAsset(asset_name=out_asset_name)
# Create a job object
the_job = Job(input=input, outputs=[outputs])
# Choose which of the analyzer Transform names you would like to use here by changing the name of the Transform to be used
# For the basic audio analyzer - pass in the audio_transform_name
# For the video analyzer - change this code to pass in the video_transform_name
analysis_transform_name = audio_transform_name
# Create a transform job
job: Job = client.jobs.create(resource_group,
account_name,
analysis_transform_name,
analysis_job_name,
parameters=the_job)
# Check Job State
job_state = client.jobs.get(resource_group, account_name,
def main(event: func.EventGridEvent):
result = json.dumps({
'id': event.id,
'data': event.get_json(),
'topic': event.topic,
'subject': event.subject,
'event_type': event.event_type,
})
logging.info('Python EventGrid trigger processed an event: %s', result)
blob_url = event.get_json().get('url')
logging.info('blob URL: %s', blob_url)
blob_name = blob_url.split("/")[-1].split("?")[0]
logging.info('blob name: %s', blob_name)
origin_container_name = blob_url.split("/")[-2].split("?")[0]
logging.info('container name: %s', origin_container_name)
storage_account_name = blob_url.split("//")[1].split(".")[0]
logging.info('storage account name: %s', storage_account_name)
ams_account_name = os.getenv('ACCOUNTNAME')
resource_group_name = os.getenv('RESOURCEGROUP')
subscription_id = os.getenv('SUBSCRIPTIONID')
client_id = os.getenv('AZURE_CLIENT_ID')
client_secret = os.getenv('AZURE_CLIENT_SECRET')
TENANT_ID = os.getenv('AZURE_TENANT_ID')
storage_blob_url = 'https://' + storage_account_name + '.blob.core.windows.net/'
transform_name = 'faceredact'
LOGIN_ENDPOINT = AZURE_PUBLIC_CLOUD.endpoints.active_directory
RESOURCE = AZURE_PUBLIC_CLOUD.endpoints.active_directory_resource_id
logging.info('login_endpoint: %s', LOGIN_ENDPOINT)
logging.info('tenant_id: %s', TENANT_ID)
out_asset_name = 'faceblurringOutput_' + datetime.utcnow().strftime(
"%m-%d-%Y_%H:%M:%S")
out_alternate_id = 'faceblurringOutput_' + datetime.utcnow().strftime(
"%m-%d-%Y_%H:%M:%S")
out_description = 'Redacted video with blurred faces'
context = adal.AuthenticationContext(LOGIN_ENDPOINT + "/" + TENANT_ID)
credentials = AdalAuthentication(
context.acquire_token_with_client_credentials, RESOURCE, client_id,
client_secret)
client = AzureMediaServices(credentials, subscription_id)
output_asset = Asset(alternate_id=out_alternate_id,
description=out_description)
client.assets.create_or_update(resource_group_name, ams_account_name,
out_asset_name, output_asset)
token_credential = DefaultAzureCredential()
datalake_service_client = DataLakeServiceClient(
account_url=storage_blob_url, credential=token_credential)
delegation_key = datalake_service_client.get_user_delegation_key(
key_start_time=datetime.utcnow(),
key_expiry_time=datetime.utcnow() + timedelta(hours=1))
sas_token = generate_file_sas(account_name=storage_account_name,
file_system_name=origin_container_name,
directory_name="",
file_name=blob_name,
credential=delegation_key,
permission=FileSasPermissions(read=True),
expiry=datetime.utcnow() +
timedelta(hours=1),
protocol="https")
sas_url = "{}?{}".format(blob_url, sas_token)
logging.info(sas_url)
job_name = 'Faceblurring-job_' + datetime.utcnow().strftime(
"%m-%d-%Y_%H:%M:%S")
job_input = JobInputHttp(label="Video_asset", files=[sas_url])
job_output = JobOutputAsset(asset_name=out_asset_name)
job_parameters = Job(input=job_input, outputs=[job_output])
client.jobs.create(resource_group_name,
ams_account_name,
transform_name,
job_name,
parameters=job_parameters)
