def perform(self, input_file_path, speech_length, nonspeech_length):
audiofile = AudioFileMFCC(gf.absolute_path(input_file_path, __file__))
audiofile.run_vad()
self.assertEqual(len(audiofile.intervals(speech=True)), speech_length)
self.assertEqual(len(audiofile.intervals(speech=False)),
nonspeech_length)
def perform_command(self):
"""
Perform command and return the appropriate exit code.
:rtype: int
"""
if len(self.actual_arguments) < 2:
return self.print_help()
audio_file_path = self.actual_arguments[0]
mode = self.actual_arguments[1]
if mode not in [u"speech", u"nonspeech", u"both"]:
return self.print_help()
output_file_path = None
if len(self.actual_arguments) >= 3:
output_file_path = self.actual_arguments[2]
output_time = not self.has_option([u"-i", u"--index"])
self.check_c_extensions("cmfcc")
if not self.check_input_file(audio_file_path):
return self.ERROR_EXIT_CODE
if (output_file_path is not None) and (
not self.check_output_file(output_file_path)):
return self.ERROR_EXIT_CODE
self.print_info(u"Reading audio...")
try:
audio_file_mfcc = AudioFileMFCC(audio_file_path,
rconf=self.rconf,
logger=self.logger)
except AudioFileConverterError:
self.print_error(u"Unable to call the ffmpeg executable '%s'" %
(self.rconf[RuntimeConfiguration.FFMPEG_PATH]))
self.print_error(u"Make sure the path to ffmpeg is correct")
return self.ERROR_EXIT_CODE
except (AudioFileUnsupportedFormatError, AudioFileNotInitializedError):
self.print_error(u"Cannot read file '%s'" % (audio_file_path))
self.print_error(u"Check that its format is supported by ffmpeg")
return self.ERROR_EXIT_CODE
except Exception as exc:
self.print_error(
u"An unexpected error occurred while reading the audio file:")
self.print_error(u"%s" % exc)
return self.ERROR_EXIT_CODE
self.print_info(u"Reading audio... done")
self.print_info(u"Executing VAD...")
audio_file_mfcc.run_vad()
self.print_info(u"Executing VAD... done")
speech = audio_file_mfcc.intervals(speech=True, time=output_time)
nonspeech = audio_file_mfcc.intervals(speech=False, time=output_time)
if mode == u"speech":
intervals = speech
elif mode == u"nonspeech":
intervals = nonspeech
elif mode == u"both":
speech = [[x[0], x[1], u"speech"] for x in speech]
nonspeech = [[x[0], x[1], u"nonspeech"] for x in nonspeech]
intervals = sorted(speech + nonspeech)
intervals = [tuple(interval) for interval in intervals]
self.write_to_file(output_file_path, intervals, output_time)
return self.NO_ERROR_EXIT_CODE
