def make_speech_corpus(top, dest, txtdest, snddest, srcfolder):
'''This function tests whether the information in an spl file is sufficient to
extract the recording and text. It also creates a directory name based on the
speaker id and the sessions id for the processed files.'''
spls = os.listdir(srcfolder)
for splfile in sorted(spls):
if os.path.splitext(splfile)[1] != ".spl":
continue
# Parse the spl file and check whether key information has been found.
# This is necessary because not all files are complete, some contain errors
# from maual editing and some spl files point to recordings that do not
# exit in the corpus
session = Session(os.path.abspath(srcfolder), splfile)
if session.speaker_id == "": # ignore if there is no speaker
continue
if not session.wavdir: # ignore if there is no matching directory
continue
if len(session.record_states) < 2: # unsure whether this has an effect
continue
session.sessiondir = os.path.join(dest, session.filestem) + "." + session.speaker_id
#
create_parallel_file_list(session, snddest, txtdest)
def make_speech_corpus(top, dest, txtdest, snddest, srcfolder):
'''This function tests whether the information in an spl file is sufficient to
extract the recording and text. It also creates a directory name based on the
speaker id and the sessions id for the processed files.'''
spls = os.listdir(srcfolder)
for splfile in sorted(spls):
if os.path.splitext(splfile)[1] != ".spl":
continue
# Parse the spl file and check whether key information has been found.
# This is necessary because not all files are complete, some contain errors
# from maual editing and some spl files point to recordings that do not
# exit in the corpus
session = Session(os.path.abspath(srcfolder), splfile)
if session.speaker_id == "": # ignore if there is no speaker
continue
if not session.wavdir: # ignore if there is no matching directory
continue
if len(session.record_states) < 2: # unsure whether this has an effect
continue
session.sessiondir = os.path.join(
dest, session.filestem) + "." + session.speaker_id
#
create_parallel_file_list(session, snddest, txtdest)
def make_speech_corpus(top, dest, srcfolder):
global n
spls = os.listdir(srcfolder)
for splfile in sorted(spls):
if os.path.splitext(splfile)[1] != ".spl":
continue
session = Session(os.path.abspath(srcfolder), splfile)
if session.speaker_id == "": # ignore if there is no speaker
continue
if not session.wavdir: # ignore if there is no matching directory
continue
if len(session.record_states) < 2: # unsure whether this has an effect
continue
session.sessiondir = os.path.join(dest, session.filestem) +"."+ session.speaker_id
if os.path.exists(session.sessiondir):
n += 1
session.sessiondir = session.sessiondir+ "_" +str(n)
session.speaker_id+ "_" +str(n)
os.mkdir(session.sessiondir)
create_parallel_files(session)
def make_speech_corpus(top, dest, srcfolder):
global n
spls = os.listdir(srcfolder)
for splfile in sorted(spls):
if os.path.splitext(splfile)[1] != ".spl":
continue
session = Session(os.path.abspath(srcfolder), splfile)
if session.speaker_id == "": # ignore if there is no speaker
continue
if not session.wavdir: # ignore if there is no matching directory
continue
if len(session.record_states) < 2: # unsure whether this has an effect
continue
session.sessiondir = os.path.join(dest, session.filestem) +"."+ session.speaker_id
if os.path.exists(session.sessiondir):
n += 1
session.sessiondir = session.sessiondir+ "_" +str(n)
session.speaker_id+ "_" +str(n)
os.mkdir(session.sessiondir)
create_parallel_files(session)
