def convert(input_file, output_file):
"""
Convert between text file formats (supported formats are stm, json, srt, vtt, txt, and html)
"""
check_input_file_validity(input_file)
input_file = assign_if_valid(input_file)
input_file.write(output_file)
def main():
parser = argparse.ArgumentParser(
description=
"Compares a reference and transcript file and calculates word error rate (WER) between these two files"
)
parser.add_argument(
"reference_file",
metavar="reference_file",
type=str,
help='reference "truth" file',
)
parser.add_argument(
"transcript_file",
metavar="transcript_file",
type=str,
help="transcript possibly containing errors",
)
parser.add_argument(
"--char-level",
help="calculate character error rate instead of word error rate",
action="store_true",
)
parser.add_argument(
"--ignore-nsns",
help="ignore non silence noises like um, uh, etc.",
action="store_true",
)
# parse arguments
args = parser.parse_args()
# read files from arguments
ref = assign_if_valid(args.reference_file)
hyp = assign_if_valid(args.transcript_file)
if ref is None or hyp is None:
print(
"Error with an input file. Please check all files exist and are accepted by ASRToolkit"
)
elif args.char_level:
print("CER: {:5.3f}%".format(cer(ref, hyp, args.ignore_nsns)))
else:
print("WER: {:5.3f}%".format(wer(ref, hyp, args.ignore_nsns)))
def convert(input_file, output_file):
"""
Convert between text file formats (supported formats are stm, json, srt, vtt, txt, and html)
Validates lines of transcript before writing new file.
STM files are unformatted (eg 10 -> ten)
"""
check_input_file_validity(input_file)
input_file = assign_if_valid(input_file)
input_file.write(output_file)
def compute_wer(reference_file,
transcript_file,
char_level=False,
ignore_nsns=False):
"""
Compares a reference and transcript file and calculates word error rate (WER) between these two files
If --char-level is given, compute CER instead
If --ignore-nsns is given, ignore non silence noises
"""
# read files from arguments
ref = assign_if_valid(reference_file)
hyp = assign_if_valid(transcript_file)
if ref is None or hyp is None:
print(
"Error with an input file. Please check all files exist and are accepted by ASRToolkit"
)
elif char_level:
print("CER: {:5.3f}%".format(cer(ref, hyp, ignore_nsns)))
else:
print("WER: {:5.3f}%".format(wer(ref, hyp, ignore_nsns)))
def convert(input_file, output_file):
input_file = assign_if_valid(input_file)
input_file.write(output_file)