def _output_smil(self, output_file, parameters=None):
"""
Output to SMIL
"""
text_ref = parameters[gc.PPN_TASK_OS_FILE_SMIL_PAGE_REF]
audio_ref = parameters[gc.PPN_TASK_OS_FILE_SMIL_AUDIO_REF]
output_file.write("<smil xmlns=\"http://www.w3.org/ns/SMIL\" xmlns:epub=\"http://www.idpf.org/2007/ops\" version=\"3.0\">\n")
output_file.write(" <body>\n")
output_file.write(" <seq id=\"s%s\" epub:textref=\"%s\">\n" % (str(1).zfill(6), text_ref))
i = 1
for fragment in self.fragments:
text = fragment.text_fragment
output_file.write(" <par id=\"p%s\">\n" % (str(i).zfill(6)))
output_file.write(" <text src=\"%s#%s\"/>\n" % (
text_ref,
text.identifier
))
output_file.write(" <audio clipBegin=\"%s\" clipEnd=\"%s\" src=\"%s\"/>\n" % (
gf.time_to_hhmmssmmm(fragment.begin),
gf.time_to_hhmmssmmm(fragment.end),
audio_ref
))
output_file.write(" </par>\n")
i += 1
output_file.write(" </seq>\n")
output_file.write(" </body>\n")
output_file.write("</smil>\n")
def _output_vtt(self, output_file):
"""
Output to WebVTT
"""
output_file.write("WEBVTT\n\n")
i = 1
for fragment in self.fragments:
text = fragment.text_fragment
output_file.write("%d\n" % i)
output_file.write("%s --> %s\n" % (
gf.time_to_hhmmssmmm(fragment.begin),
gf.time_to_hhmmssmmm(fragment.end)
))
output_file.write("%s\n" % text.text)
output_file.write("\n")
i += 1
def createSyncedLyricsFile(lyrics, file):
global lyricsSynced, errors
f = open("tempSync.txt", "w+")
f.write(lyrics)
f.close()
config = TaskConfiguration()
config[gc.PPN_TASK_LANGUAGE] = Language.FRA
config[gc.PPN_TASK_IS_TEXT_FILE_FORMAT] = TextFileFormat.PLAIN
config[gc.PPN_TASK_OS_FILE_FORMAT] = SyncMapFormat.AUDH
task = Task()
task.configuration = config
try:
task.audio_file_path_absolute = file
task.text_file_path_absolute = "tempSync.txt"
ExecuteTask(task).execute()
syncedLyricsFile = open(file[:-4] + ".lrc", "w+")
for fragment in task.sync_map_leaves():
syncedLyricsFile.write(
str('[' + gf.time_to_hhmmssmmm(fragment.interval.begin, '.')[3:-1] + ']' + fragment.text + '\n'))
syncedLyricsFile.close()
print(" Sync Added", sep=' ', end='', flush=True)
lyricsSynced += 1
except Exception as e :
errors += 1
print(" Sync error", sep=' ', end='',flush=True)
def _write_vtt(self, output_file):
"""
Write to WebVTT file
"""
output_file.write("WEBVTT\n\n")
i = 1
for fragment in self.fragments:
text = fragment.text_fragment
output_file.write("%d\n" % i)
output_file.write("%s --> %s\n" % (
gf.time_to_hhmmssmmm(fragment.begin),
gf.time_to_hhmmssmmm(fragment.end)
))
for line in text.lines:
output_file.write("%s\n" % line)
output_file.write("\n")
i += 1
def print_result(self, audio_len, start, end):
"""
Print result of SD.
:param audio_len: the length of the entire audio file, in seconds
:type audio_len: float
:param start: the start position of the spoken text
:type start: float
:param end: the end position of the spoken text
:type end: float
"""
msg = []
zero = 0
head_len = start
text_len = end - start
tail_len = audio_len - end
msg.append(u"")
msg.append(u"Head: %.3f %.3f (%.3f)" % (zero, start, head_len))
msg.append(u"Text: %.3f %.3f (%.3f)" % (start, end, text_len))
msg.append(u"Tail: %.3f %.3f (%.3f)" % (end, audio_len, tail_len))
msg.append(u"")
zero_h = gf.time_to_hhmmssmmm(0)
start_h = gf.time_to_hhmmssmmm(start)
end_h = gf.time_to_hhmmssmmm(end)
audio_len_h = gf.time_to_hhmmssmmm(audio_len)
head_len_h = gf.time_to_hhmmssmmm(head_len)
text_len_h = gf.time_to_hhmmssmmm(text_len)
tail_len_h = gf.time_to_hhmmssmmm(tail_len)
msg.append("Head: %s %s (%s)" % (zero_h, start_h, head_len_h))
msg.append("Text: %s %s (%s)" % (start_h, end_h, text_len_h))
msg.append("Tail: %s %s (%s)" % (end_h, audio_len_h, tail_len_h))
msg.append(u"")
self.print_info(u"\n".join(msg))
def print_result(self, audio_len, start, end):
"""
Print result of SD.
:param audio_len: the length of the entire audio file, in seconds
:type audio_len: float
:param start: the start position of the spoken text
:type start: float
:param end: the end position of the spoken text
:type end: float
"""
msg = []
zero = 0
head_len = start
text_len = end - start
tail_len = audio_len - end
msg.append(u"")
msg.append(u"Head: %.3f %.3f (%.3f)" % (zero, start, head_len))
msg.append(u"Text: %.3f %.3f (%.3f)" % (start, end, text_len))
msg.append(u"Tail: %.3f %.3f (%.3f)" % (end, audio_len, tail_len))
msg.append(u"")
zero_h = gf.time_to_hhmmssmmm(0)
start_h = gf.time_to_hhmmssmmm(start)
end_h = gf.time_to_hhmmssmmm(end)
audio_len_h = gf.time_to_hhmmssmmm(audio_len)
head_len_h = gf.time_to_hhmmssmmm(head_len)
text_len_h = gf.time_to_hhmmssmmm(text_len)
tail_len_h = gf.time_to_hhmmssmmm(tail_len)
msg.append("Head: %s %s (%s)" % (zero_h, start_h, head_len_h))
msg.append("Text: %s %s (%s)" % (start_h, end_h, text_len_h))
msg.append("Tail: %s %s (%s)" % (end_h, audio_len_h, tail_len_h))
msg.append(u"")
self.print_info(u"\n".join(msg))
def test_time_to_hhmmssmmm(self):
tests = [
[None, "00:00:00.000"],
[0.000, "00:00:00.000"],
[12.000, "00:00:12.000"],
[12.345, "00:00:12.345"],
[60, "00:01:00.000"],
[83.000, "00:01:23.000"],
[83.456, "00:01:23.456"],
[3600.000, "01:00:00.000"],
[3612.000, "01:00:12.000"],
[3612.340, "01:00:12.340"], # numerical issues
[4980.000, "01:23:00.000"],
[5025.000, "01:23:45.000"],
[5025.670, "01:23:45.670"], # numerical issues
]
for test in tests:
self.assertEqual(gf.time_to_hhmmssmmm(test[0]), test[1])
def test_time_to_hhmmssmmm(self):
tests = [
(None, "00:00:00.000"),
(0.000, "00:00:00.000"),
(12.000, "00:00:12.000"),
(12.345, "00:00:12.345"),
(60, "00:01:00.000"),
(83.000, "00:01:23.000"),
(83.456, "00:01:23.456"),
(3600.000, "01:00:00.000"),
(3612.000, "01:00:12.000"),
(3612.340, "01:00:12.340"), # numerical issues
(4980.000, "01:23:00.000"),
(5025.000, "01:23:45.000"),
(5025.670, "01:23:45.670"), # numerical issues
]
for test in tests:
self.assertEqual(gf.time_to_hhmmssmmm(test[0]), test[1])
def test_time_to_hhmmssmmm(self):
tests = [
[None, "00:00:00.000"],
[0.000, "00:00:00.000"],
[12.000, "00:00:12.000"],
[12.345, "00:00:12.345"],
[60, "00:01:00.000"],
[83.000, "00:01:23.000"],
[83.456, "00:01:23.456"],
[3600.000, "01:00:00.000"],
[3612.000, "01:00:12.000"],
[3612.340, "01:00:12.340"], # numerical issues
[4980.000, "01:23:00.000"],
[5025.000, "01:23:45.000"],
[5025.670, "01:23:45.670"], # numerical issues
]
for test in tests:
self.assertEqual(gf.time_to_hhmmssmmm(test[0]), test[1])
def test_time_to_hhmmssmmm(self):
tests = [
(None, "00:00:00.000"),
(0.000, "00:00:00.000"),
(12.000, "00:00:12.000"),
(12.345, "00:00:12.345"),
(60, "00:01:00.000"),
(83.000, "00:01:23.000"),
(83.456, "00:01:23.456"),
(3600.000, "01:00:00.000"),
(3612.000, "01:00:12.000"),
(3612.340, "01:00:12.340"), # numerical issues
(4980.000, "01:23:00.000"),
(5025.000, "01:23:45.000"),
(5025.670, "01:23:45.670"), # numerical issues
]
for test in tests:
self.assertEqual(gf.time_to_hhmmssmmm(test[0]), test[1])
def main():
""" Entry point """
if len(sys.argv) < 5:
usage()
return
language = sys.argv[1]
text_file_path = sys.argv[2]
text_format = sys.argv[3]
audio_file_path = sys.argv[-1]
verbose = False
parameters = {}
for i in range(4, len(sys.argv)-1):
args = sys.argv[i].split("=")
if len(args) == 1:
verbose = (args[0] in ["v", "-v", "verbose", "--verbose"])
if len(args) == 2:
key, value = args
if key == "id_regex":
parameters[gc.PPN_JOB_IS_TEXT_UNPARSED_ID_REGEX] = value
if key == "class_regex":
parameters[gc.PPN_JOB_IS_TEXT_UNPARSED_CLASS_REGEX] = value
if key == "sort":
parameters[gc.PPN_JOB_IS_TEXT_UNPARSED_ID_SORT] = value
if key == "min_head_length":
parameters["min_head_length"] = float(value)
if key == "max_head_length":
parameters["max_head_length"] = float(value)
if key == "min_tail_length":
parameters["min_head_length"] = float(value)
if key == "max_tail_length":
parameters["max_tail_length"] = float(value)
if not gf.can_run_c_extension():
print "[WARN] Unable to load Python C Extensions"
print "[WARN] Running the slower pure Python code"
print "[WARN] See the README file for directions to compile the Python C Extensions"
logger = Logger(tee=verbose)
print "[INFO] Reading audio..."
tmp_handler, tmp_file_path = tempfile.mkstemp(
suffix=".wav",
dir=gf.custom_tmp_dir()
)
converter = FFMPEGWrapper(logger=logger)
converter.convert(audio_file_path, tmp_file_path)
audio_file = AudioFile(tmp_file_path)
print "[INFO] Reading audio... done"
print "[INFO] Reading text..."
if text_format == "list":
text_file = TextFile()
text_file.read_from_list(text_file_path.split("|"))
else:
text_file = TextFile(text_file_path, text_format, parameters)
text_file.set_language(language)
print "[INFO] Reading text... done"
print "[INFO] Detecting audio interval..."
sd = SD(audio_file, text_file, logger=logger)
min_head_length = gc.SD_MIN_HEAD_LENGTH
if "min_head_length" in parameters:
min_head_length = parameters["min_head_length"]
max_head_length = gc.SD_MAX_HEAD_LENGTH
if "max_head_length" in parameters:
max_head_length = parameters["max_head_length"]
min_tail_length = gc.SD_MIN_TAIL_LENGTH
if "min_tail_length" in parameters:
min_tail_length = parameters["min_tail_length"]
max_tail_length = gc.SD_MAX_TAIL_LENGTH
if "max_tail_length" in parameters:
max_tail_length = parameters["max_tail_length"]
start, end = sd.detect_interval(
min_head_length,
max_head_length,
min_tail_length,
max_tail_length
)
zero = 0
audio_len = audio_file.audio_length
head_len = start
text_len = end - start
tail_len = audio_len - end
print "[INFO] Detecting audio interval... done"
print "[INFO] "
print "[INFO] Head: %.3f %.3f (%.3f)" % (zero, start, head_len)
print "[INFO] Text: %.3f %.3f (%.3f)" % (start, end, text_len)
print "[INFO] Tail: %.3f %.3f (%.3f)" % (end, audio_len, tail_len)
print "[INFO] "
zero_h = gf.time_to_hhmmssmmm(0)
start_h = gf.time_to_hhmmssmmm(start)
end_h = gf.time_to_hhmmssmmm(end)
audio_len_h = gf.time_to_hhmmssmmm(audio_len)
head_len_h = gf.time_to_hhmmssmmm(head_len)
text_len_h = gf.time_to_hhmmssmmm(text_len)
tail_len_h = gf.time_to_hhmmssmmm(tail_len)
print "[INFO] Head: %s %s (%s)" % (zero_h, start_h, head_len_h)
print "[INFO] Text: %s %s (%s)" % (start_h, end_h, text_len_h)
print "[INFO] Tail: %s %s (%s)" % (end_h, audio_len_h, tail_len_h)
#print "[INFO] Cleaning up..."
cleanup(tmp_handler, tmp_file_path)
def main():
""" Entry point """
if len(sys.argv) < 5:
usage()
return
language = sys.argv[1]
text_file_path = sys.argv[2]
text_format = sys.argv[3]
audio_file_path = sys.argv[-1]
verbose = False
parameters = {}
for i in range(4, len(sys.argv) - 1):
args = sys.argv[i].split("=")
if len(args) == 1:
verbose = (args[0] in ["v", "-v", "verbose", "--verbose"])
if len(args) == 2:
key, value = args
if key == "id_regex":
parameters[gc.PPN_JOB_IS_TEXT_UNPARSED_ID_REGEX] = value
if key == "class_regex":
parameters[gc.PPN_JOB_IS_TEXT_UNPARSED_CLASS_REGEX] = value
if key == "sort":
parameters[gc.PPN_JOB_IS_TEXT_UNPARSED_ID_SORT] = value
if key == "min_head_length":
parameters["min_head_length"] = float(value)
if key == "max_head_length":
parameters["max_head_length"] = float(value)
if key == "min_tail_length":
parameters["min_head_length"] = float(value)
if key == "max_tail_length":
parameters["max_tail_length"] = float(value)
if not gf.can_run_c_extension():
print "[WARN] Unable to load Python C Extensions"
print "[WARN] Running the slower pure Python code"
print "[WARN] See the README file for directions to compile the Python C Extensions"
logger = Logger(tee=verbose)
print "[INFO] Reading audio..."
tmp_handler, tmp_file_path = tempfile.mkstemp(suffix=".wav",
dir=gf.custom_tmp_dir())
converter = FFMPEGWrapper(logger=logger)
converter.convert(audio_file_path, tmp_file_path)
audio_file = AudioFile(tmp_file_path)
print "[INFO] Reading audio... done"
print "[INFO] Reading text..."
if text_format == "list":
text_file = TextFile()
text_file.read_from_list(text_file_path.split("|"))
else:
text_file = TextFile(text_file_path, text_format, parameters)
text_file.set_language(language)
print "[INFO] Reading text... done"
print "[INFO] Detecting audio interval..."
sd = SD(audio_file, text_file, logger=logger)
min_head_length = gc.SD_MIN_HEAD_LENGTH
if "min_head_length" in parameters:
min_head_length = parameters["min_head_length"]
max_head_length = gc.SD_MAX_HEAD_LENGTH
if "max_head_length" in parameters:
max_head_length = parameters["max_head_length"]
min_tail_length = gc.SD_MIN_TAIL_LENGTH
if "min_tail_length" in parameters:
min_tail_length = parameters["min_tail_length"]
max_tail_length = gc.SD_MAX_TAIL_LENGTH
if "max_tail_length" in parameters:
max_tail_length = parameters["max_tail_length"]
start, end = sd.detect_interval(min_head_length, max_head_length,
min_tail_length, max_tail_length)
zero = 0
audio_len = audio_file.audio_length
head_len = start
text_len = end - start
tail_len = audio_len - end
print "[INFO] Detecting audio interval... done"
print "[INFO] "
print "[INFO] Head: %.3f %.3f (%.3f)" % (zero, start, head_len)
print "[INFO] Text: %.3f %.3f (%.3f)" % (start, end, text_len)
print "[INFO] Tail: %.3f %.3f (%.3f)" % (end, audio_len, tail_len)
print "[INFO] "
zero_h = gf.time_to_hhmmssmmm(0)
start_h = gf.time_to_hhmmssmmm(start)
end_h = gf.time_to_hhmmssmmm(end)
audio_len_h = gf.time_to_hhmmssmmm(audio_len)
head_len_h = gf.time_to_hhmmssmmm(head_len)
text_len_h = gf.time_to_hhmmssmmm(text_len)
tail_len_h = gf.time_to_hhmmssmmm(tail_len)
print "[INFO] Head: %s %s (%s)" % (zero_h, start_h, head_len_h)
print "[INFO] Text: %s %s (%s)" % (start_h, end_h, text_len_h)
print "[INFO] Tail: %s %s (%s)" % (end_h, audio_len_h, tail_len_h)
#print "[INFO] Cleaning up..."
cleanup(tmp_handler, tmp_file_path)
