Commit 6c87fa83 by Qubad786

Integrate 3PlayMedia Transcription API

parent 1b69155e
......@@ -39,4 +39,10 @@ urlpatterns = [
view=transcripts.Cielo24CallbackHandlerView.as_view(),
name='cielo24_transcript_completed'
),
# 3PlayMedia callback handler view
url(
regex=r'^3playmedia/transcripts/handle/(?P<token>[\w]+)$',
view=transcripts.ThreePlayMediaCallbackHandlerView.as_view(),
name='3play_media_callback'
)
]
......@@ -647,6 +647,8 @@ class TranscriptProcessMetadata(TimeStampedModel):
video = models.ForeignKey(Video)
provider = models.CharField('Transcript provider', max_length=50, choices=TranscriptProvider.CHOICES)
process_id = models.CharField('Process id', max_length=255)
# To keep track of 3Play Translations.
translation_id = models.CharField('Translation id', max_length=255, null=True, blank=True)
lang_code = models.CharField('Language code', max_length=3)
status = models.CharField(
'Transcript status',
......
......@@ -9,9 +9,11 @@ import boto
import django.dispatch
import requests
from boto.s3.key import Key
from django.db.models import Q
from pysrt import SubRipFile
from requests.packages.urllib3.exceptions import InsecurePlatformWarning
from rest_framework import status
from rest_framework.parsers import FormParser
from rest_framework.permissions import AllowAny
from rest_framework.response import Response
from rest_framework.views import APIView
......@@ -26,11 +28,34 @@ requests.packages.urllib3.disable_warnings(InsecurePlatformWarning)
logging.basicConfig()
LOGGER = logging.getLogger(__name__)
VALID_TRANSCRIPT_PROVIDERS = [
'',
'',
]
# 3PlayMedia possible send-along statuses for a transcription callback.
COMPLETE = 'complete'
ERROR = 'error'
# Transcript format
TRANSCRIPT_SJSON = 'sjson'
CIELO24_TRANSCRIPT_COMPLETED = django.dispatch.Signal(providing_args=['job_id', 'lang_code', 'org', 'video_id'])
CIELO24_GET_CAPTION_URL = 'https://api.cielo24.com/api/job/get_caption'
CONFIG = utils.get_config()
# 3PlayMedia callback signal
THREE_PLAY_TRANSCRIPTION_DONE = django.dispatch.Signal(
providing_args=['org', 'lang_code', 'edx_video_id', 'file_id', 'status', 'error_description']
)
# 3PlayMedia API URLs.
THREE_PLAY_TRANSCRIPT_URL = u'https://static.3playmedia.com/files/{file_id}/transcript.srt'
THREE_PLAY_TRANSLATION_SERVICES_URL = u'https://static.3playmedia.com/translation_services'
THREE_PLAY_ORDER_TRANSLATION_URL = u'https://api.3playmedia.com/files/{file_id}/translations/order'
THREE_PLAY_TRANSLATION_STATUS_URL = u'https://static.3playmedia.com/files/{file_id}/translations/{translation_id}'
THREE_PLAY_TRANSLATION_DOWNLOAD_URL = (u'https://static.3playmedia.com/files/{file_id}/translations/{translation_id}/'
u'captions.srt')
class TranscriptError(Exception):
"""
......@@ -46,6 +71,13 @@ class TranscriptFetchError(TranscriptError):
pass
class TranscriptTranslationError(TranscriptError):
"""
An error occurred during the translation attempt on 3PlayMedia.
"""
pass
class TranscriptConversionError(TranscriptError):
"""
An error occurred during srt to sjson conversion.
......@@ -266,3 +298,494 @@ def upload_sjson_to_s3(config, sjson_data):
)
k.set_contents_from_string(json.dumps(sjson_data))
return k.key
class ThreePlayMediaCallbackHandlerView(APIView):
"""
View to handle 3PlayMedia callback requests.
"""
parser_classes = (FormParser,)
permission_classes = (AllowValidTranscriptProvider,)
def post(self, request, **kwargs):
"""
Handle 3PlayMedia callback request.
"""
required_attrs = ['file_id', 'status', 'org', 'edx_video_id']
received_attributes = request.data.keys() + request.query_params.keys()
missing = [attr for attr in required_attrs if attr not in received_attributes]
if missing:
LOGGER.warning(
u'[3PlayMedia Callback] process_id=%s Received Attributes=%s Missing Attributes=%s',
request.data.get('file_id'),
received_attributes,
missing,
)
return Response(status=status.HTTP_200_OK)
# Dispatch 3playMedia transcription signal
THREE_PLAY_TRANSCRIPTION_DONE.send_robust(
sender=self,
org=request.query_params['org'],
edx_video_id=request.query_params['edx_video_id'],
lang_code='en',
file_id=request.data['file_id'],
status=request.data['status'],
# Following is going to be an error description if an error occurs during
# 3playMedia transcription process
error_description=request.data.get('error_description'),
)
return Response(status=status.HTTP_200_OK)
def order_translations(file_id, api_key, api_secret, target_languages):
"""
Order translations on 3PlayMedia for all the target languages.
Process:
* Fetch all the pending translations process for a file
* Fetch all the translation services from 3PlayMedia
* For each process,
- Find suitable translation service
- Order translation from that service
- Move the process to `in progress` and update it with the
translation id received from 3Play.
Arguments:
file_id(unicode): File identifier
api_key(unicode): API key
api_secret(unicode): API Secret
target_languages(list): List of language codes
Raises:
TranscriptTranslationError: when an error occurred while fetching the translation services.
"""
translation_processes = TranscriptProcessMetadata.objects.filter(
process_id=file_id,
provider=TranscriptProvider.THREE_PLAY,
status=TranscriptStatus.PENDING,
lang_code__in=target_languages,
)
response = requests.get(utils.build_url(THREE_PLAY_TRANSLATION_SERVICES_URL, apikey=api_key))
if not response.ok:
# Fail all the pending translation processes associated with this file id.
translation_processes.update(status=TranscriptStatus.FAILED)
raise TranscriptTranslationError(
u'[3PlayMedia Callback] Error while fetching the translation services -- {status}, {response}'.format(
status=response.status_code,
response=response.text,
)
)
# Response should be a list containing services, details:
# http://support.3playmedia.com/hc/en-us/articles/227729988-Translations-API-Methods
available_services = json.loads(response.text)
if not isinstance(available_services, list):
# Fail all the pending translation processes associated with this file id.
translation_processes.update(status=TranscriptStatus.FAILED)
raise TranscriptTranslationError(
u'[3PlayMedia Callback] Expected list but got: -- {response}.'.format(
response=response.text,
)
)
for target_language in target_languages:
try:
translation_process = translation_processes.filter(lang_code=target_language).latest()
except TranscriptProcessMetadata.DoesNotExist:
LOGGER.warning(
u'[3PlayMedia Callback] process not found for target language %s -- process id %s',
target_language,
file_id,
)
continue
# 1 - Find a standard service for translation in the target language.
translation_service_id = None
for service in available_services:
service_found = (
service['target_language_iso_639_1_code'] == target_language and
service['service_level'] == 'standard'
)
if service_found:
translation_service_id = service['id']
break
if translation_service_id is None:
# Fail the process
translation_process.status = TranscriptStatus.FAILED
translation_process.save()
LOGGER.error(
'[3PlayMedia Callback] No translation service found for target language %s -- process id %s',
target_language,
file_id,
)
continue
# 2 - At this point, we've got our service ready to use. Now, place an order for the translation.
response = requests.post(THREE_PLAY_ORDER_TRANSLATION_URL.format(file_id=file_id), data={
'apikey': api_key,
'api_secret_key': api_secret,
'translation_service_id': translation_service_id,
})
if not response.ok:
# Fail the process
translation_process.status = TranscriptStatus.FAILED
translation_process.save()
LOGGER.error(
'[3PlayMedia Callback] An error occurred during translation, target language=%s, file_id=%s, status=%s',
target_language,
file_id,
response.status_code,
)
continue
# Translation Order API returns `success` attribute specifying whether the order has been placed
# successfully: http://support.3playmedia.com/hc/en-us/articles/227729988-Translations-API-Methods
translation_order = json.loads(response.text)
if translation_order.get('success'):
translation_process.status = TranscriptStatus.IN_PROGRESS
translation_process.translation_id = translation_order['translation_id']
translation_process.save()
else:
translation_process.status = TranscriptStatus.FAILED
translation_process.save()
LOGGER.error(
'[3PlayMedia Callback] Translation failed fot target language=%s, file_id=%s, response=%s',
target_language,
file_id,
response.text,
)
@django.dispatch.receiver(THREE_PLAY_TRANSCRIPTION_DONE, dispatch_uid="three_play_transcription_done")
def three_play_transcription_callback(sender, **kwargs):
"""
Arguments:
sender: sender of the signal
kwargs(dict): video transcription metadata
Process:
* download transcript(SRT) from 3PlayMedia
* convert SRT to SJSON
* upload SJSON to AWS S3
* order translations for all the preferred languages
* update transcript status in VAL
"""
# Extract all the must have attributes
org = kwargs['org']
edx_video_id = kwargs['edx_video_id']
lang_code = kwargs['lang_code']
file_id = kwargs['file_id']
state = kwargs['status']
try:
process = TranscriptProcessMetadata.objects.filter(
provider=TranscriptProvider.THREE_PLAY,
process_id=file_id,
lang_code=lang_code,
).latest()
except TranscriptProcessMetadata.DoesNotExist:
LOGGER.exception(
u'[3PlayMedia Callback] Unable to get transcript process for org=%s, edx_video_id=%s, file_id=%s.',
org,
edx_video_id,
file_id,
)
return
if state == COMPLETE:
# Indicates that the default video speech transcription has been done successfully.
try:
transcript_secrets = TranscriptPreferences.objects.get(org=org, provider=TranscriptProvider.THREE_PLAY)
except TranscriptPreferences.DoesNotExist:
# Fail the process
process.status = TranscriptStatus.FAILED
process.save()
# Log the failure
LOGGER.exception(
u'[3PlayMedia Callback] Unable to get transcript secrets for org=%s, edx_video_id=%s, file_id=%s.',
org,
edx_video_id,
file_id,
)
return
# Fetch the transcript from 3PlayMedia
try:
srt_transcript = fetch_srt_data(
THREE_PLAY_TRANSCRIPT_URL.format(file_id=file_id),
apikey=transcript_secrets.api_key,
)
except TranscriptFetchError:
process.status = TranscriptStatus.FAILED
process.save()
LOGGER.exception(
'[3PlayMedia Callback] Fetch request failed for video=%s -- lang=%s -- process_id=%s',
edx_video_id,
lang_code,
file_id
)
return
# fetched transcript is going to be SRT content and if this is not so, it'll be a json response
# describing the error.
try:
json.loads(srt_transcript)
# Fail the process and log all the details.
process.status = TranscriptStatus.FAILED
process.save()
LOGGER.error(
'[3PlayMedia Task] Transcript fetch error for video=%s -- lang_code=%s -- process=%s -- response=%s',
process.video.studio_id,
process.lang_code,
process.process_id,
srt_transcript,
)
return
except ValueError:
pass
# We've got the transcript from 3PlayMedia, now update process status accordingly.
process.status = TranscriptStatus.READY
process.save()
try:
sjson_transcript = convert_srt_to_sjson(srt_transcript)
sjson_file = upload_sjson_to_s3(CONFIG, sjson_transcript)
except Exception:
# in case of any exception, log and raise.
LOGGER.exception(
u'[3PlayMedia Callback] Request failed for video=%s -- lang_code=%s -- process_id=%s',
edx_video_id,
lang_code,
file_id,
)
raise
# Update edx-val with completed transcript information
val_api = VALAPICall(video_proto=None, val_status=None)
val_api.update_val_transcript(
video_id=process.video.studio_id,
lang_code=lang_code,
name=sjson_file,
transcript_format=TRANSCRIPT_SJSON,
provider=TranscriptProvider.THREE_PLAY,
)
# Translation Phase
target_languages = list(process.video.preferred_languages)
# Remove the language that is already processed - in our case, its en.
target_languages.remove(lang_code)
# Check if the translations are needed.
if target_languages:
# Create the translation tracking processes for all the target languages.
for target_language in target_languages:
TranscriptProcessMetadata.objects.create(
video=process.video,
provider=TranscriptProvider.THREE_PLAY,
process_id=file_id,
lang_code=target_language,
status=TranscriptStatus.PENDING,
)
try:
# Order translations for target languages
order_translations(file_id, transcript_secrets.api_key, transcript_secrets.api_secret, target_languages)
except TranscriptTranslationError:
LOGGER.exception(
u'[3PlayMedia Callback] Translation could not be performed - org=%s, edx_video_id=%s, file_id=%s.',
org,
edx_video_id,
file_id,
)
except Exception:
LOGGER.exception(
(u'[3PlayMedia Callback] Error while translating the transcripts - org=%s, edx_video_id=%s, '
u'file_id=%s.'),
org,
edx_video_id,
file_id,
)
raise
# in case if there is only one language which has already been processed.
if not target_languages:
val_api.update_video_status(
process.video.studio_id, VideoStatus.TRANSCRIPTION_READY
)
# On success, a happy farewell log.
LOGGER.info(
u'[3PlayMedia Callback] Video speech transcription was successful for video=%s -- lang_code=%s -- '
u'process_id=%s',
edx_video_id,
lang_code,
file_id,
)
elif state == ERROR:
# Fail the process
process.status = TranscriptStatus.FAILED
process.save()
# Log the error information
LOGGER.error(
u'[3PlayMedia Callback] Error while transcription - error=%s, org=%s, edx_video_id=%s, file_id=%s.',
kwargs['error_description'],
org,
edx_video_id,
file_id,
)
else:
# Status must be either 'complete' or 'error'
# more details on http://support.3playmedia.com/hc/en-us/articles/227729828-Files-API-Methods
LOGGER.error(
u'[3PlayMedia Callback] Got invalid status - status=%s, org=%s, edx_video_id=%s, file_id=%s.',
state,
org,
edx_video_id,
file_id,
)
def retrieve_three_play_translations():
"""
Checks translation status on 3PlayMedia for all the progressing processes, fetches them if they're complete.
Retrieval flow:
1. Fetches 3PlayMedia translation processes whose status is `in progress`
2. For each process, retrieve the org-wide api keys
3. Check translation status through 3PlayMedia
4. If its done, mark the process as complete, fetch translated transcript, convert to sjson, upload it to s3 and
finally, update it in edx-val.
"""
translation_processes = TranscriptProcessMetadata.objects.filter(
provider=TranscriptProvider.THREE_PLAY,
status=TranscriptStatus.IN_PROGRESS,
).exclude(Q(translation_id__isnull=True) | Q(translation_id__exact=''))
for translation_process in translation_processes:
course_id = translation_process.video.inst_class.local_storedir.split(',')[0]
org = utils.extract_course_org(course_id=course_id)
try:
three_play_secrets = TranscriptPreferences.objects.get(org=org, provider=TranscriptProvider.THREE_PLAY)
except TranscriptPreferences.DoesNotExist:
LOGGER.exception(
u'[3PlayMedia Task] 3Play secrets not found for video=%s -- lang_code=%s -- process_id=%s',
translation_process.video.studio_id,
translation_process.lang_code,
translation_process.process_id,
)
continue
translation_status_url = utils.build_url(
THREE_PLAY_TRANSLATION_STATUS_URL.format(
file_id=translation_process.process_id,
translation_id=translation_process.translation_id,
),
apikey=three_play_secrets.api_key
)
response = requests.get(translation_status_url)
if not response.ok:
LOGGER.error(
(u'[3PlayMedia Task] Translation status request failed for video=%s -- '
u'lang_code=%s -- process_id=%s -- status=%s'),
translation_process.video.studio_id,
translation_process.lang_code,
translation_process.process_id,
response.status_code,
)
continue
translation_status = json.loads(response.text)
if translation_status.get('iserror'):
translation_process.status = TranscriptStatus.FAILED
translation_process.save()
LOGGER.error(
(u'[3PlayMedia Task] unable to get translation status for video=%s -- '
u'lang_code=%s -- process_id=%s -- response=%s'),
translation_process.video.studio_id,
translation_process.lang_code,
translation_process.process_id,
response.text,
)
continue
if translation_status['state'] == 'complete':
translation_download_url = utils.build_url(
THREE_PLAY_TRANSLATION_DOWNLOAD_URL.format(
file_id=translation_process.process_id,
translation_id=translation_process.translation_id,
),
apikey=three_play_secrets.api_key
)
response = requests.get(translation_download_url)
if not response.ok:
LOGGER.error(
u'[3PlayMedia Task] Translation download failed for video=%s -- lang_code=%s -- process_id=%s -- '
u'status=%s',
translation_process.video.studio_id,
translation_process.lang_code,
translation_process.process_id,
response.status_code,
)
continue
# its going to be SRT content and `json.loads` should raise
# ValueError if its a valid response, otherwise it'll be json
# response in result of an error.
try:
json.loads(response.text)
translation_process.status = TranscriptStatus.FAILED
translation_process.save()
LOGGER.error(
u'[3PlayMedia Task] Translation error for video=%s -- lang_code=%s -- process_id=%s -- response=%s',
translation_process.video.studio_id,
translation_process.lang_code,
translation_process.process_id,
response.text,
)
continue
except ValueError:
pass
# We've got the transcript from 3PlayMedia, now update process status accordingly.
translation_process.status = TranscriptStatus.READY
translation_process.save()
try:
sjson_transcript = convert_srt_to_sjson(response.text)
sjson_file = upload_sjson_to_s3(CONFIG, sjson_transcript)
except Exception:
# in case of any exception, log and raise.
LOGGER.exception(
u'[3PlayMedia Task] translation failed for video=%s -- lang_code=%s -- process_id=%s',
translation_process.video.studio_id,
translation_process.lang_code,
translation_process.process_id,
)
raise
# Update edx-val with completed transcript information
val_api = VALAPICall(video_proto=None, val_status=None)
val_api.update_val_transcript(
video_id=translation_process.video.studio_id,
lang_code=translation_process.lang_code,
name=sjson_file,
transcript_format=TRANSCRIPT_SJSON,
provider=TranscriptProvider.THREE_PLAY,
)
# if all the processes for this video are complete, update video status in edx-val
# update transcript status for video in edx-val only if all language transcripts are ready
video_jobs = TranscriptProcessMetadata.objects.filter(video__studio_id=translation_process.video.studio_id)
if all(video_job.status == TranscriptStatus.READY for video_job in video_jobs):
val_api.update_video_status(translation_process.video.studio_id, VideoStatus.TRANSCRIPTION_READY)
from __future__ import absolute_import
import os
import sys
from celery import Celery
import yaml
from VEDA_OS01.transcripts import retrieve_three_play_translations
"""
Start Celery Worker
......@@ -51,6 +51,18 @@ app.conf.update(
CELERY_ACCEPT_CONTENT=['pickle', 'json', 'msgpack', 'yaml']
)
app.conf.beat_schedule = {
'check-3play-translations-every-30-seconds': {
'task': 'tasks.fetch_three_play_translations',
'schedule': 30.0,
},
}
@app.task(name='fetch_three_play_translations')
def fetch_three_play_translations():
retrieve_three_play_translations()
@app.task(name='worker_encode')
def worker_task_fire(veda_id, encode_profile, jobid):
......
import datetime
import ftplib
import logging
import os
import shutil
import sys
from os.path import expanduser
import boto
......@@ -16,6 +13,7 @@ from boto.s3.key import Key
from django.core.urlresolvers import reverse
import veda_deliver_xuetang
from control.veda_deliver_3play import ThreePLayMediaClient
from control_env import *
from veda_deliver_cielo import Cielo24Transcript
from veda_deliver_youtube import DeliverYoutube
......@@ -190,7 +188,6 @@ class VedaDelivery:
self._CLEANUP()
self._THREEPLAY_UPLOAD()
# Transcription Process
# We only want to generate transcripts for `desktop_mp4` profile.
if self.encode_profile == 'desktop_mp4' and self.video_query.process_transcription:
......@@ -203,7 +200,6 @@ class VedaDelivery:
if self.video_query.provider == TranscriptProvider.CIELO24:
self.cielo24_transcription_flow()
def _INFORM_INTAKE(self):
"""
Collect all salient metadata and
......@@ -574,48 +570,55 @@ class VedaDelivery:
)
cielo24.start_transcription_flow()
def _THREEPLAY_UPLOAD(self):
if self.video_query.inst_class.tp_proc is False:
return None
if self.video_query.inst_class.mobile_override is False:
if self.encode_profile != 'desktop_mp4':
return None
ftp1 = ftplib.FTP(
self.auth_dict['threeplay_ftphost']
)
user = self.video_query.inst_class.tp_username.strip()
passwd = self.video_query.inst_class.tp_password.strip()
def start_3play_transcription_process(self):
"""
3PlayMedia Transcription Flow
"""
try:
ftp1.login(user, passwd)
except:
ErrorObject.print_error(
message='3Play Authentication Failure'
# Picks the first course from the list as there may be multiple
# course runs in that list (i.e. all having the same org).
org = utils.extract_course_org(self.video_proto.platform_course_url[0])
transcript_secrets = TranscriptPreferences.objects.get(org=org, provider=self.video_query.provider)
# update transcript status for video in edx-val
VALAPICall(video_proto=None, val_status=None).update_video_status(
self.video_query.studio_id, VideoStatus.TRANSCRIPTION_IN_PROGRESS
)
try:
ftp1.cwd(
self.video_query.inst_class.tp_speed
# Initialize 3playMedia client and start transcription process
s3_video_url = build_url(
self.auth_dict['s3_base_url'],
self.auth_dict['edx_s3_endpoint_bucket'],
self.encoded_file
)
except:
ftp1.mkd(
self.video_query.inst_class.tp_speed
callback_url = build_url(
self.auth_dict['veda_base_url'],
reverse(
'3play_media_callback',
args=[self.auth_dict['transcript_provider_request_token']]
),
# Additional attributes that'll come back with the callback
org=org,
edx_video_id=self.video_query.studio_id,
)
ftp1.cwd(
self.video_query.inst_class.tp_speed
three_play_media = ThreePLayMediaClient(
org=org,
video=self.video_query,
media_url=s3_video_url,
api_key=transcript_secrets.api_key,
api_secret=transcript_secrets.api_secret,
callback_url=callback_url,
turnaround_level=self.video_query.three_play_turnaround,
)
os.chdir(self.node_work_directory)
three_play_media.generate_transcripts()
ftp1.storbinary(
'STOR ' + self.encoded_file,
open(os.path.join(
self.node_work_directory,
self.encoded_file
), 'rb')
except TranscriptPreferences.DoesNotExist:
LOGGER.warning(
'Transcript preference is not found for provider=%s, video=%s',
self.video_query.provider,
self.video_query.studio_id,
)
os.chdir(homedir)
def _XUETANG_ROUTE(self):
if self.video_query.inst_class.xuetang_proc is False:
return None
......
"""
3PlayMedia Transcription Client
"""
import json
import logging
import requests
import sys
from requests.packages.urllib3.exceptions import InsecurePlatformWarning
from VEDA_OS01.models import TranscriptProcessMetadata, TranscriptProvider, TranscriptStatus
from VEDA_OS01.utils import build_url
requests.packages.urllib3.disable_warnings(InsecurePlatformWarning)
LOGGER = logging.getLogger(__name__)
class ThreePlayMediaError(Exception):
"""
An error that occurs during 3PlayMedia actions.
"""
pass
class ThreePlayMediaLanguageNotFoundError(ThreePlayMediaError):
"""
An error when language is not found in available 3playMedia languages.
"""
pass
class ThreePlayMediaPerformTranscriptionError(ThreePlayMediaError):
"""
An error occurred while adding media for transcription.
"""
pass
class ThreePlayMediaUrlError(ThreePlayMediaError):
"""
Occurs when the media url is either inaccessible or of invalid content type.
"""
pass
class ThreePLayMediaClient(object):
def __init__(self, org, video, media_url, api_key, api_secret, callback_url, turnaround_level):
"""
Initialize 3play media client
"""
self.org = org
self.video = video
self.media_url = media_url
self.api_key = api_key
self.api_secret = api_secret
self.callback_url = callback_url
self.turnaround_level = turnaround_level
# default attributes
self.base_url = u'https://api.3playmedia.com/'
self.upload_media_file_url = u'files/'
self.available_languages_url = u'caption_imports/available_languages/'
self.allowed_content_type = u'video/mp4'
def validate_media_url(self):
"""
Validates the media URL
Raises:
3PlayMediaUrlError: on invalid media url or content type
"""
if not self.media_url:
raise ThreePlayMediaUrlError('Invalid media URL "{media_url}".'.format(media_url=self.media_url))
response = requests.head(url=self.media_url)
if not response.ok:
raise ThreePlayMediaUrlError('The URL "{media_url}" is not Accessible.'.format(media_url=self.media_url))
elif response.headers['Content-Type'] != self.allowed_content_type:
raise ThreePlayMediaUrlError(
'Media content-type should be "{allowed_type}". URL was "{media_url}", content-type was "{type}"'.format(
allowed_type=self.allowed_content_type,
media_url=self.media_url,
type=response.headers['Content-Type'],
)
)
def submit_media(self):
"""
Submits the media to perform transcription.
Raises:
ThreePlayMediaPerformTranscriptionError: error while transcription process
"""
self.validate_media_url()
# Prepare requests payload
payload = dict(
# Mandatory attributes required for transcription
link=self.media_url,
apikey=self.api_key,
api_secret_key=self.api_secret,
turnaround_level=self.turnaround_level,
callback_url=self.callback_url,
)
upload_url = build_url(self.base_url, self.upload_media_file_url)
response = requests.post(
url=upload_url,
data=json.dumps(payload),
headers={'Content-Type': 'application/json'}
)
if not response.ok:
raise ThreePlayMediaPerformTranscriptionError(
'Upload file request failed with: {response} -- {status}'.format(
response=response.text, status=response.status_code
)
)
try:
# A normal response should be a text containing file id and if we're getting a deserializable dict, there
# must be an error: http://support.3playmedia.com/hc/en-us/articles/227729828-Files-API-Methods
if isinstance(json.loads(response.text), dict):
raise ThreePlayMediaPerformTranscriptionError(
'Expected file id but got: {response}'.format(response=response.text)
)
except ValueError:
return response.text
def generate_transcripts(self):
"""
Kicks off transcription process for default language.
"""
try:
file_id = self.submit_media()
# Track progress of transcription process
TranscriptProcessMetadata.objects.create(
video=self.video,
process_id=file_id,
lang_code=u'en',
provider=TranscriptProvider.THREE_PLAY,
status=TranscriptStatus.IN_PROGRESS,
)
# Successfully kicked off transcription process for a video with the given language.
LOGGER.info(
'[3PlayMedia] Transcription process has been started for video=%s, language=en.',
self.video.studio_id,
)
except ThreePlayMediaError:
LOGGER.exception(
'[3PlayMedia] Could not process transcripts for video=%s language=en.',
self.video.studio_id,
)
except Exception:
LOGGER.exception(
'[3PlayMedia] Unexpected error while transcription for video=%s language=en .',
self.video.studio_id,
)
raise
def main():
pass
if __name__ == '__main__':
sys.exit(main())
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment