def main():
parser = argparse.ArgumentParser(description='Auto translate jp CCs in MPEG TS file.')
parser.add_argument('infile', help='Input filename (MPEG2 Transport Stream File)', type=str)
parser.add_argument('pid', help='Pid of closed caption ES to extract from stream.', type=int)
#parser.add_argument('-k', '--secret_key', help='Windows secret key for bing translate API.', type=str, default='')
args = parser.parse_args()
pid = args.pid
infilename = args.infile
if not os.path.exists(infilename):
print 'Please provide input Transport Stream file.'
os.exit(-1)
#open an Ass file and formatter
ass_file = ASSFile(infilename+'_ENG.ass')
ass = ASSFormatter(ass_file)
#CC data is not, in itself timestamped, so we've got to use packet info
#to reconstruct the timing of the closed captions (i.e. how many seconds into
#the file are they shown?)
initial_timestamp = 0
pes_packet = None
pes = []
elapsed_time_s = 0
for packet in next_ts_packet(infilename):
#always process timestamp info, regardless of PID
if packet.adapatation_field() and packet.adapatation_field().PCR():
current_timestamp = packet.adapatation_field().PCR()
initial_timestamp = initial_timestamp or current_timestamp
delta = current_timestamp - initial_timestamp
elapsed_time_s = float(delta)/90000.0
#if this is the stream PID we're interestd in, reconstruct the ES
if packet.pid() == pid:
if packet.payload_start():
pes = copy.deepcopy(packet.payload())
else:
pes.extend(packet.payload())
pes_packet = PESPacket(pes)
#if our packet is fully formed (payload all present) we can parse its contents
if pes_packet.length() == (pes_packet.header_size() + pes_packet.payload_size()):
data_group = DataGroup(pes_packet.payload())
if not data_group.is_management_data():
#We now have a Data Group that contains caption data.
#We take out its payload, but this is further divided into 'Data Unit' structures
caption = data_group.payload()
#iterate through the Data Units in this payload via another generator.
for data_unit in next_data_unit(caption):
#we're only interested in those Data Units which are "statement body" to get CC data.
if not isinstance(data_unit.payload(), StatementBody):
continue
ass.format(data_unit.payload().payload(), elapsed_time_s)
def OnESPacket(current_pid, packet, header_size):
"""
Callback invoked on the successful extraction of an Elementary Stream packet from the
Transport Stream file packets.
:param current_pid: The TS Program ID for the TS packets this info originated from
:param packet: The ENTIRE ES packet, header and payload-- which may have been assembled
from multiple TS packet payloads.
:param header_size: Size of the header in bytes (characters in the string). Provided to more
easily separate the packet into header and payload.
:return: None
"""
global pid
global VERBOSE
global SILENT
global elapsed_time_s
if pid >= 0 and current_pid != pid:
return
try:
payload = ES.get_pes_payload(packet)
f = list(payload)
#f = bytearray(payload)
data_group = DataGroup(f)
if not data_group.is_management_data():
#We now have a Data Group that contains caption data.
#We take out its payload, but this is further divided into 'Data Unit' structures
caption = data_group.payload()
#iterate through the Data Units in this payload via another generator.
for data_unit in next_data_unit(caption):
#we're only interested in those Data Units which are "statement body" to get CC data.
if not isinstance(data_unit.payload(), StatementBody):
continue
#okay. Finally we've got a data unit with CC data. Feed its payload to the custom
if pid < 0 and VERBOSE and not SILENT:
pid = current_pid
print("Found Closed Caption data in PID: " + str(pid))
print("Will now only process this PID to improve performance.")
#formatter function above. This dumps the basic text to stdout.
cc = formatter(data_unit.payload().payload(), elapsed_time_s)
if cc and VERBOSE:
#according to best practice, always deal internally with UNICODE, and encode to
#your encoding of choice as late as possible. Here, i'm encoding as UTF-8 for
#my command line.
#DECODE EARLY, ENCODE LATE
print(cc.encode('utf-8'))
except EOFError:
pass
except Exception, err:
if VERBOSE and not SILENT and pid >= 0:
print("Exception thrown while handling DataGroup in ES. This may be due to many factors"
+ "such as file corruption or the .ts file using as yet unsupported features.")
traceback.print_exc(file=sys.stdout)
def OnESPacket(current_pid, packet, header_size):
"""
Callback invoked on the successful extraction of an Elementary Stream packet from the
Transport Stream file packets.
:param current_pid: The TS Program ID for the TS packets this info originated from
:param packet: The ENTIRE ES packet, header and payload-- which may have been assembled
from multiple TS packet payloads.
:param header_size: Size of the header in bytes (characters in the string). Provided to more
easily separate the packet into header and payload.
:return: None
"""
global pid
global VERBOSE
global SILENT
global elapsed_time_s
global ass
global infilename
global outfilename
global tmax
global time_offset
if pid >= 0 and current_pid != pid:
return
try:
payload = ES.get_pes_payload(packet)
f = list(payload)
data_group = DataGroup(f)
if not data_group.is_management_data():
#We now have a Data Group that contains caption data.
#We take out its payload, but this is further divided into 'Data Unit' structures
caption = data_group.payload()
#iterate through the Data Units in this payload via another generator.
for data_unit in next_data_unit(caption):
#we're only interested in those Data Units which are "statement body" to get CC data.
if not isinstance(data_unit.payload(), StatementBody):
continue
if not ass:
v = not SILENT
ass = ASSFormatter(tmax=tmax, video_filename=outfilename, verbose=v)
ass.format(data_unit.payload().payload(), elapsed_time_s)
# this code used to sed the PID we're scanning via first successful ARIB decode
# but i've changed it below to draw present CC language info form ARIB
# management data. Leaving this here for reference.
#if pid < 0 and not SILENT:
# pid = current_pid
# print("Found Closed Caption data in PID: " + str(pid))
# print("Will now only process this PID to improve performance.")
else:
# management data
management_data = data_group.payload()
numlang = management_data.num_languages()
if pid < 0 and numlang > 0:
for language in range(numlang):
if not SILENT:
print("Closed caption management data for language: "
+ management_data.language_code(language)
+ " available in PID: " + str(current_pid))
print("Will now only process this PID to improve performance.")
pid = current_pid
except EOFError:
pass
except FileOpenError as ex:
# allow IOErrors to kill application
raise ex
except Exception, err:
if not SILENT and pid >= 0:
print("Exception thrown while handling DataGroup in ES. This may be due to many factors"
+ "such as file corruption or the .ts file using as yet unsupported features.")
traceback.print_exc(file=sys.stdout)
def main():
parser = argparse.ArgumentParser(description='Draw CC Packets from MPG2 Transport Stream file.')
parser.add_argument('infile', help='Input filename (MPEG2 Transport Stream File)', type=str)
parser.add_argument('pid', help='Pid of closed caption ES to extract from stream.', type=int)
args = parser.parse_args()
pid = args.pid
infilename = args.infile
if not os.path.exists(infilename):
print 'Please provide input Transport Stream file.'
os.exit(-1)
#open an Ass file and formatter
ass_file = ASSFile(infilename+'.ass')
ass = ASSFormatter(ass_file)
#CC data is not, in itself timestamped, so we've got to use packet info
#to reconstruct the timing of the closed captions (i.e. how many seconds into
#the file are they shown?)
initial_timestamp = 0
pes_packet = None
pes = []
elapsed_time_s = 0
for packet in next_ts_packet(infilename):
#always process timestamp info, regardless of PID
if packet.adapatation_field() and packet.adapatation_field().PCR():
current_timestamp = packet.adapatation_field().PCR()
initial_timestamp = initial_timestamp or current_timestamp
delta = current_timestamp - initial_timestamp
elapsed_time_s = float(delta)/90000.0
#if this is the stream PID we're interestd in, reconstruct the ES
if packet.pid() == pid:
try:
if packet.payload_start():
pes = copy.deepcopy(packet.payload())
else:
pes.extend(packet.payload())
pes_packet = PESPacket(pes)
#if our packet is fully formed (payload all present) we can parse its contents
if pes_packet.length() == (pes_packet.header_size() + pes_packet.payload_size()):
data_group = DataGroup(pes_packet.payload())
if not data_group.is_management_data():
#We now have a Data Group that contains caption data.
#We take out its payload, but this is further divided into 'Data Unit' structures
caption = data_group.payload()
#iterate through the Data Units in this payload via another generator.
for data_unit in next_data_unit(caption):
#we're only interested in those Data Units which are "statement body" to get CC data.
if not isinstance(data_unit.payload(), StatementBody):
continue
ass.format(data_unit.payload().payload(), elapsed_time_s)
#okay. Finally we've got a data unit with CC data. Feed its payload to the custom
#formatter function above. This dumps the basic text to stdout.
#cc = formatter(data_unit.payload().payload(), elapsed_time_s)
#if cc:
#according to best practice, always deal internally with UNICODE, and encode to
#your encoding of choice as late as possible. Here, i'm encoding as UTF-8 for
#my command line.
#DECODE EARLY, ENCODE LATE
#print(cc.encode('utf-8'))
except:
pass
def main():
parser = argparse.ArgumentParser(
description=
'Remove ARIB formatted Closed Caption information from an MPEG TS file and format the results as a standard .ass subtitle file.'
)
parser.add_argument('infile',
help='Input filename (MPEG2 Transport Stream File)',
type=str)
parser.add_argument(
'-p',
'--pid',
help=
'Specify a PID of a PES known to contain closed caption info (tool will attempt to find the proper PID if not specified.).',
type=int,
default=-1)
parser.add_argument('-v',
'--verbose',
help='Verbose output.',
action='store_true')
parser.add_argument('-q',
'--quiet',
help='Does not write to stdout.',
action='store_true')
parser.add_argument('-t',
'--tmax',
help='Subtitle display time limit (seconds).',
type=int,
default=5)
parser.add_argument(
'-o',
'--timeoffset',
help=
'Shift all time values in generated .ass file by indicated floating point offset in seconds.',
type=float,
default=0.0)
args = parser.parse_args()
pid = args.pid
infilename = args.infile
quiet = args.quiet
verbose = args.verbose
tmax = args.tmax
time_offset = args.timeoffset
if not os.path.exists(infilename):
print 'Input filename :' + infilename + " does not exist."
os.exit(-1)
#open an Ass file and formatter
ass_file = None #ASSFile(infilename+'.ass')
ass = None #ASSFormatter(ass_file, tmax=tmax)
#CC data is not, in itself timestamped, so we've got to use packet info
#to reconstruct the timing of the closed captions (i.e. how many seconds into
#the file are they shown?)
initial_timestamp = 0
pes_packet = None
pes = []
elapsed_time_s = 0
# get filesize for progress meter
total_filesize = os.path.getsize(infilename)
read_size = 0
percent_read = 0
prev_percent_read = percent_read
if not quiet and not verbose:
#show initial progress information
sys.stdout.write("progress: %d%% \r" % (percent_read))
sys.stdout.flush()
for packet in next_ts_packet(infilename):
read_size += TSPacket.PACKET_SIZE_BYTES
percent_read = ((read_size / float(total_filesize)) * 100)
new_percent_read = int(percent_read * 100)
if not quiet and not verbose and new_percent_read != prev_percent_read:
prev_percent_read = new_percent_read
#print("totalsize:"+str(total_filesize)+" read_size "+str(read_size) + " percent: " + str(new_percent_read))
sys.stdout.write("progress: %.2f%% \r" % (percent_read))
sys.stdout.flush()
#always process timestamp info, regardless of PID
if packet.adapatation_field() and packet.adapatation_field().PCR():
current_timestamp = packet.adapatation_field().PCR()
initial_timestamp = initial_timestamp or current_timestamp
delta = current_timestamp - initial_timestamp
elapsed_time_s = float(delta) / 90000.0 + time_offset
#if this is the stream PID we're interestd in, reconstruct the ES
if pid < 0 or (pid == packet.pid()):
try:
if packet.payload_start():
pes = copy.deepcopy(packet.payload())
else:
pes.extend(packet.payload())
pes_packet = PESPacket(pes)
#if our packet is fully formed (payload all present) we can parse its contents
if pes_packet.length() == (pes_packet.header_size() +
pes_packet.payload_size()):
data_group = DataGroup(pes_packet.payload())
if not data_group.is_management_data():
#We now have a Data Group that contains caption data.
#We take out its payload, but this is further divided into 'Data Unit' structures
caption = data_group.payload()
#iterate through the Data Units in this payload via another generator.
for data_unit in next_data_unit(caption):
#we're only interested in those Data Units which are "statement body" to get CC data.
if not isinstance(data_unit.payload(),
StatementBody):
continue
# only write the file if we've actually found some Closed Captions
if not ass_file:
ass_file = ASSFile(infilename + '.ass')
if not ass:
ass = ASSFormatter(ass_file,
tmax=tmax,
video_filename=infilename)
ass.format(data_unit.payload().payload(),
elapsed_time_s)
if pid < 0:
pid = packet.pid()
print("Found Closed Caption data in PID: " +
str(pid))
print(
"Will now only process this PID to improve performance."
)
#print("properly formed packet with pid: "+ str(packet.pid()))
#okay. Finally we've got a data unit with CC data. Feed its payload to the custom
#formatter function above. This dumps the basic text to stdout.
#cc = formatter(data_unit.payload().payload(), elapsed_time_s)
#if cc:
#according to best practice, always deal internally with UNICODE, and encode to
#your encoding of choice as late as possible. Here, i'm encoding as UTF-8 for
#my command line.
#DECODE EARLY, ENCODE LATE
#print(cc.encode('utf-8'))
except:
#print("exception thrown on packet with PID: " + str(packet.pid()))
pass
if pid < 0 or not ass:
print("Did not find any Closed Caption data in the file " + infilename)
def OnESPacket(current_pid, packet, header_size):
"""
Callback invoked on the successful extraction of an Elementary Stream packet from the
Transport Stream file packets.
:param current_pid: The TS Program ID for the TS packets this info originated from
:param packet: The ENTIRE ES packet, header and payload-- which may have been assembled
from multiple TS packet payloads.
:param header_size: Size of the header in bytes (characters in the string). Provided to more
easily separate the packet into header and payload.
:return: None
"""
global pid
global VERBOSE
global SILENT
global elapsed_time_s
global ass
global infilename
global outfilename
global tmax
global time_offset
if pid >= 0 and current_pid != pid:
return
try:
payload = ES.get_pes_payload(packet)
f = list(payload)
data_group = DataGroup(f)
if not data_group.is_management_data():
#We now have a Data Group that contains caption data.
#We take out its payload, but this is further divided into 'Data Unit' structures
caption = data_group.payload()
#iterate through the Data Units in this payload via another generator.
for data_unit in next_data_unit(caption):
#we're only interested in those Data Units which are "statement body" to get CC data.
if not isinstance(data_unit.payload(), StatementBody):
continue
if not ass:
v = not SILENT
ass = ASSFormatter(tmax=tmax, video_filename=outfilename, verbose=v)
ass.format(data_unit.payload().payload(), elapsed_time_s)
# this code used to sed the PID we're scanning via first successful ARIB decode
# but i've changed it below to draw present CC language info form ARIB
# management data. Leaving this here for reference.
#if pid < 0 and not SILENT:
# pid = current_pid
# print("Found Closed Caption data in PID: " + str(pid))
# print("Will now only process this PID to improve performance.")
else:
# management data
management_data = data_group.payload()
numlang = management_data.num_languages()
if pid < 0 and numlang > 0:
for language in range(numlang):
if not SILENT:
print("Closed caption management data for language: "
+ management_data.language_code(language)
+ " available in PID: " + str(current_pid))
print("Will now only process this PID to improve performance.")
pid = current_pid
except EOFError:
pass
except FileOpenError as ex:
# allow IOErrors to kill application
raise ex
except Exception, err:
if not SILENT and pid >= 0:
print("Exception thrown while handling DataGroup in ES. This may be due to many factors"
+ "such as file corruption or the .ts file using as yet unsupported features.")
traceback.print_exc(file=sys.stdout)
def main():
parser = argparse.ArgumentParser(
description='Auto translate jp CCs in MPEG TS file.')
parser.add_argument('infile',
help='Input filename (MPEG2 Transport Stream File)',
type=str)
parser.add_argument(
'pid',
help='Pid of closed caption ES to extract from stream.',
type=int)
#parser.add_argument('-k', '--secret_key', help='Windows secret key for bing translate API.', type=str, default='')
args = parser.parse_args()
pid = args.pid
infilename = args.infile
if not os.path.exists(infilename):
print 'Please provide input Transport Stream file.'
os.exit(-1)
#open an Ass file and formatter
ass_file = ASSFile(infilename + '_ENG.ass')
ass = ASSFormatter(ass_file)
#CC data is not, in itself timestamped, so we've got to use packet info
#to reconstruct the timing of the closed captions (i.e. how many seconds into
#the file are they shown?)
initial_timestamp = 0
pes_packet = None
pes = []
elapsed_time_s = 0
for packet in next_ts_packet(infilename):
#always process timestamp info, regardless of PID
if packet.adapatation_field() and packet.adapatation_field().PCR():
current_timestamp = packet.adapatation_field().PCR()
initial_timestamp = initial_timestamp or current_timestamp
delta = current_timestamp - initial_timestamp
elapsed_time_s = float(delta) / 90000.0
#if this is the stream PID we're interestd in, reconstruct the ES
if packet.pid() == pid:
if packet.payload_start():
pes = copy.deepcopy(packet.payload())
else:
pes.extend(packet.payload())
pes_packet = PESPacket(pes)
#if our packet is fully formed (payload all present) we can parse its contents
if pes_packet.length() == (pes_packet.header_size() +
pes_packet.payload_size()):
data_group = DataGroup(pes_packet.payload())
if not data_group.is_management_data():
#We now have a Data Group that contains caption data.
#We take out its payload, but this is further divided into 'Data Unit' structures
caption = data_group.payload()
#iterate through the Data Units in this payload via another generator.
for data_unit in next_data_unit(caption):
#we're only interested in those Data Units which are "statement body" to get CC data.
if not isinstance(data_unit.payload(), StatementBody):
continue
ass.format(data_unit.payload().payload(),
elapsed_time_s)