def main(verbose=False, *speech_corpora):
"""Scan directory for audio files and convert them to wav files
For each speech corpus `speech_corpus`
1. the resulting wav files are written to the directory
`.speechrc.wav16`/<speech_corpus>/
2. the transcripts in data/src/speech/<speech_corpus>/transcripts_*.csv are
updated.
"""
misc.init_app('speech_audio_scan')
config = misc.load_config('.speechrc')
speech_corpora_dir = Path(config.get("speech", "speech_corpora"))
wav16 = Path(config.get("speech", "wav16"))
if len(speech_corpora) < 1:
logging.error("At least one speech corpus must be provided.")
sys.exit(1)
if verbose:
logging.basicConfig(level=logging.DEBUG)
else:
logging.basicConfig(level=logging.INFO)
exit_if_corpus_is_missing(speech_corpora_dir, speech_corpora)
for speech_corpus in speech_corpora:
transcripts = Transcripts(corpus_name=speech_corpus, create_db=True)
out_wav16_subdir = wav16 / speech_corpus
out_wav16_subdir.mkdir(parents=True, exist_ok=True)
in_root_corpus_dir = speech_corpora_dir / speech_corpus
scan_audiodir(str(in_root_corpus_dir), transcripts,
str(out_wav16_subdir))
transcripts.save()
print speech_corpus, "new transcripts saved."
print
t = missing_token
else:
t = tokenize(ts['ts'])[0]
lex_edit(t)
#
# fini
#
curses.nocbreak()
stdscr.keypad(0)
curses.echo()
curses.endwin()
transcripts.save()
print "new transcripts saved."
print
lex.save()
print "new lexicon saved."
print
except:
curses.nocbreak()
stdscr.keypad(0)
curses.echo()
curses.endwin()
print u"*** ERROR: Unexpected error:", sys.exc_info()[0]
traceback.print_exc()