def anthrotranscribe(media_uuid):
"""
Launch transcription process for a given media
"""
register_openers()
media_item = Media.get(client=client, uuid=media_uuid)
media_item.anthrotranscribe(client=client)
def anthrotranscribe(media_uuid):
"""
Launch transcription process for a given media
"""
register_openers()
media_item = Media.get(client=client, uuid=media_uuid)
media_item.anthrotranscribe(client=client)
def batch_download_transcript():
register_openers()
media_to_download_transcript_for = csv_to_json(
csv_filename="%s/%s" %
(settings.get('base', 'path.local.transcripts'),
settings.get('test', 'media.download_transcript.csv')))['data']
for media_item in media_to_download_transcript_for:
download_transcript(media_uuid=media_item[0])
def batch_upload_transcribe():
register_openers()
media_to_upload_transcribe = csv_to_json(
csv_filename="%s/%s" %
(settings.get('base', 'path.local.media'),
settings.get('test', 'media.transcribe.csv')))['data']
for media_item in media_to_upload_transcribe:
upload_transcribe(media_filename=media_item[0])
def download_transcript(media_uuid):
register_openers()
media_item = Media.get(client=client, uuid=media_uuid)
print media_item.current_transcript
transcript = Transcript.get(client=client, uuid=media_item.current_transcript['uuid'], format='json')
if not os.path.exists(settings.get('base','path.local.scripts.output')):
os.makedirs(settings.get('base','path.local.scripts.output'))
f = open("%s/%s.srt" % (settings.get('base','path.local.scripts.output'), media_item.title), 'w')
f.write(transcript.content)
def publish_unpublish(media_uuid):
"""
Publish and unpublish a media item's transcript.
Publishing a transcript will make it available to anyone that knowns its uuid
"""
register_openers()
media_item = Media(accept='text/xml', username=username, password=password, uid=media_uuid)
media_item.get()
#from xml.dom import minidom
#xmldoc = minidom.parseString(media_item.response.read())
# current_transcript_uuid = xmldoc.getElementsByTagName('currentTranscript')[0].getElementsByTagName('id')[0].firstChild.data
media_item.publish()
def batch_upload_transcribe():
register_openers()
media_to_upload_transcribe = csv_to_json(
csv_filename="%s/%s" % (
settings.get('base', 'path.local.media'),
settings.get('test', 'media.transcribe.csv')
)
)['data']
for media_item in media_to_upload_transcribe:
upload_transcribe(
media_filename=media_item[0]
)
def batch_download_transcript():
register_openers()
media_to_download_transcript_for = csv_to_json(
csv_filename="%s/%s" % (
settings.get('base', 'path.local.transcripts'),
settings.get('test', 'media.download_transcript.csv')
)
)['data']
for media_item in media_to_download_transcript_for:
download_transcript(
media_uuid=media_item[0]
)
def batch_align():
register_openers()
# align all the files from the csv
media_to_align = csv_to_json(
csv_filename="%s/%s" %
(settings.get('base', 'path.local.media'),
settings.get('test', 'media.align.csv')))['data']
for media_item in media_to_align:
upload_align(media_filename=media_item[1],
aligndata="%s/%s.txt" % (
settings.get('base', 'path.local.transcripts'),
media_item[0],
))
def publish_unpublish(media_uuid):
"""
Publish and unpublish a media item's transcript.
Publishing a transcript will make it available to anyone that knowns its uuid
"""
register_openers()
media_item = Media(accept='text/xml',
username=username,
password=password,
uid=media_uuid)
media_item.get()
#from xml.dom import minidom
#xmldoc = minidom.parseString(media_item.response.read())
# current_transcript_uuid = xmldoc.getElementsByTagName('currentTranscript')[0].getElementsByTagName('id')[0].firstChild.data
media_item.publish()
def download_transcript(media_uuid):
register_openers()
media_item = Media.get(client=client, uuid=media_uuid)
print media_item.current_transcript
transcript = Transcript.get(client=client,
uuid=media_item.current_transcript['uuid'],
format='json')
if not os.path.exists(settings.get('base', 'path.local.scripts.output')):
os.makedirs(settings.get('base', 'path.local.scripts.output'))
f = open(
"%s/%s.srt" %
(settings.get('base', 'path.local.scripts.output'), media_item.title),
'w')
f.write(transcript.content)
def upload_transcribe(media_filename, title=None):
"""
Upload a media file to Koemei for transcription
"""
register_openers()
try:
log.info("Upload and transcribe file %s ..." % media_filename)
media_item = Media.create(client=client,
media_filename=media_filename,
title=title)
log.info("... OK - media uuid: %s" % media_item.uuid)
except Exception, e:
log.error("... Error creating media %s ..." % media_filename)
log.error(e)
log.error(traceback.format_exc())
raise e
def batch_align():
register_openers()
# align all the files from the csv
media_to_align = csv_to_json(
csv_filename="%s/%s" % (
settings.get('base', 'path.local.media'),
settings.get('test', 'media.align.csv')
)
)['data']
for media_item in media_to_align:
upload_align(
media_filename=media_item[1].strip(),
aligndata="%s/%s.txt" % (
settings.get('base', 'path.local.transcripts'),
media_item[0],
)
)
def upload_transcribe(media_filename, title=None):
"""
Upload a media file to Koemei for transcription
"""
register_openers()
try:
log.info("Upload and transcribe file %s ..." % media_filename)
media_item = Media.create(
client=client,
media_filename=media_filename,
title=title
)
log.info("... OK - media uuid: %s" % media_item.uuid)
except Exception, e:
log.error("... Error creating media %s ..." % media_filename)
log.error(e)
log.error(traceback.format_exc())
raise e
def upload_align(media_filename, aligndata):
"""
Upload a media file to Koemei for alignment
NOTE : you will need your account to be specially setup to use this feature
"""
register_openers()
try:
log.info("Upload and align media %s ..." % media_filename)
media_item = Media.create(client=client,
media_filename=media_filename,
aligndata=aligndata,
transcribe=False)
log.info("... OK - media uuid: %s" % media_item.uuid)
except Exception, e:
log.error("... Error aligning media %s ..." % media_filename)
log.error(e)
log.error(traceback.format_exc())
raise e
def create(
cls,
client,
media_item,
parent_note=None,
**kwargs):
"""
Create a note.
"""
data = {}
# TODO : fill in note params
headers = {}
url = [settings.get('base', 'paths.api.notes')]
register_openers()
response = client.request(url=url, data=data, headers=headers)
response_json = json.loads(response)
note = Note(fields=response_json['note'])
return note
def upload_align(media_filename, aligndata):
"""
Upload a media file to Koemei for alignment
NOTE : you will need your account to be specially setup to use this feature
"""
register_openers()
try:
log.info("Upload and align media %s ..." % media_filename)
media_item = Media.create(
client=client,
media_filename=media_filename,
aligndata=aligndata,
transcribe=False
)
log.info("... OK - media uuid: %s" % media_item.uuid)
except Exception, e:
log.error("... Error aligning media %s ..." % media_filename)
log.error(e)
log.error(traceback.format_exc())
raise e
def create(
cls,
client,
media_filename,
transcribe=True,
aligndata=None,
**kwargs):
"""
Create a media item.
If transcript_filename provided: create media and align the transcript.
@param media_filename: local/remote address of the media file to transcribe
@param metadata_filename: local path to the metadata file containing media info (title, description, ...)
@param transcript_filename: local path to the plain text transcript file to align
@param transcribe: automagically launch transcription
@param kwargs: title, description, tags, ...
"""
data = {}
headers = {}
headers_ = {}
url = [settings.get('base', 'paths.api.media')]
# create the media from a service
if 'service' in kwargs:
data.update({
'service': kwargs.get('service'),
'item_id': kwargs.get('item_id')}
)
if 'title' in kwargs and kwargs["title"] is not None:
data.update({
'title': kwargs.get('title'),
})
if 'description' in kwargs:
data.update({
'description': kwargs.get('description'),
})
# upload from remote url
if 'http' in media_filename:
#url.append("?media=" + urllib.quote(media_filename, safe=''))
#data = "" # should not be empty dict but empty string!
data.update({
'media': media_filename
})
data, headers_ = multipart_encode(data)
# upload from local hard drive
else:
register_openers()
data.update(
{'media': open(media_filename, "rb")}
)
data, headers_ = multipart_encode(data)
headers.update(headers_)
response = client.request(url=url, data=data, headers=headers)
response_json = json.loads(response)
media_item = Media(fields=response_json['media_item'])
if aligndata is not None:
media_item.align(client=client, aligndata=aligndata)
elif transcribe:
media_item.transcribe(client=client)
return media_item
def create(
cls,
client,
media_filename,
transcribe=True,
aligndata=None,
**kwargs):
"""
Create a media item.
If transcript_filename provided: create media and align the transcript.
@param media_filename: local/remote address of the media file to transcribe
@param metadata_filename: local path to the metadata file containing media info (title, description, ...)
@param transcript_filename: local path to the plain text transcript file to align
@param transcribe: automagically launch transcription
@param kwargs: title, description, tags, ...
"""
data = {}
headers = {}
headers_ = {}
url = [settings.get('base', 'paths.api.media')]
# create the media from a service
if 'service' in kwargs:
data.update({
'service': kwargs.get('service'),
'item_id': kwargs.get('item_id')}
)
if 'title' in kwargs:
data.update({
'title': kwargs.get('title'),
})
if 'description' in kwargs:
data.update({
'description': kwargs.get('description'),
})
# upload from remote url
if 'http' in media_filename:
#url.append("?media=" + urllib.quote(media_filename, safe=''))
#data = "" # should not be empty dict but empty string!
data.update({
'media': media_filename
})
data, headers_ = multipart_encode(data)
# upload from local hard drive
else:
register_openers()
data.update(
{'media': open(media_filename, "rb")}
)
data, headers_ = multipart_encode(data)
headers.update(headers_)
response = client.request(url=url, data=data, headers=headers)
response_json = json.loads(response)
media_item = Media(fields=response_json['media_item'])
if aligndata is not None:
media_item.align(client=client, aligndata=aligndata)
elif transcribe:
media_item.transcribe(client=client)
return media_item
