From cf71104424ca3234da06abf4ce6e5794fda594c5 Mon Sep 17 00:00:00 2001 From: Hiroshiba Kazuyuki Date: Fri, 24 Nov 2017 06:43:24 +0900 Subject: atr subset scirpts --- scripts/ln_jnas_subset.py | 30 ++++++++++++++++++++++++++++++ 1 file changed, 30 insertions(+) create mode 100644 scripts/ln_jnas_subset.py (limited to 'scripts/ln_jnas_subset.py') diff --git a/scripts/ln_jnas_subset.py b/scripts/ln_jnas_subset.py new file mode 100644 index 0000000..e5aba5c --- /dev/null +++ b/scripts/ln_jnas_subset.py @@ -0,0 +1,30 @@ +import argparse +import multiprocessing +from pathlib import Path + +from jnas_metadata_loader import load_from_directory +from jnas_metadata_loader.jnas_metadata import JnasMetadata + +parser = argparse.ArgumentParser() +parser.add_argument('jnas', type=Path) +parser.add_argument('output', type=Path) +parser.add_argument('--format', default='{sex}{text_id}_{mic}_atr_{subset}{sen_id}.wav') +argument = parser.parse_args() + +jnas = argument.jnas # type: Path +output = argument.output # type: Path + +jnas_list = load_from_directory(str(jnas)) +atr_list = jnas_list.subset_news_or_atr('B') + +output.mkdir(exist_ok=True) + + +def process(d: JnasMetadata): + p = d.path + out = output / argument.format.format(**d._asdict()) + out.symlink_to(p) + + +pool = multiprocessing.Pool() +pool.map(process, atr_list) -- cgit v1.2.3-70-g09d2