init
Browse files- download_audio.py +38 -3
- requirements.txt +2 -1
- util.py +14 -19
download_audio.py
CHANGED
@@ -1,8 +1,43 @@
|
|
|
|
|
|
|
|
|
|
1 |
from util import wget
|
2 |
|
|
|
3 |
url_metadata_s2s = "https://dl.fbaipublicfiles.com/seamless/data/seamless_align_nov2023_extension/seamless.dataset.metadata.public.enA-jaA.tsv.gz"
|
4 |
url_metadata_s2t = "https://dl.fbaipublicfiles.com/seamless/data/seamless.dataset.metadata.public.enA-jpn.withduration.tsv.gz"
|
5 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
6 |
|
7 |
-
wget(url_metadata_s2s, cache_dir=cache_dir)
|
8 |
-
wget(url_metadata_s2t, cache_dir=cache_dir)
|
|
|
1 |
+
import os
|
2 |
+
from os.path import join as p_join
|
3 |
+
import pandas as pd
|
4 |
+
from tqdm import tqdm
|
5 |
from util import wget
|
6 |
|
7 |
+
|
8 |
url_metadata_s2s = "https://dl.fbaipublicfiles.com/seamless/data/seamless_align_nov2023_extension/seamless.dataset.metadata.public.enA-jaA.tsv.gz"
|
9 |
url_metadata_s2t = "https://dl.fbaipublicfiles.com/seamless/data/seamless.dataset.metadata.public.enA-jpn.withduration.tsv.gz"
|
10 |
+
cache_dir_root = "./download"
|
11 |
+
|
12 |
+
|
13 |
+
def get_metadata(url: str):
|
14 |
+
cache_dir = p_join(cache_dir_root, "meta")
|
15 |
+
filename = os.path.basename(url).replace(".gz", "")
|
16 |
+
if not os.path.exists(filename):
|
17 |
+
assert wget(url, cache_dir=cache_dir)
|
18 |
+
df = pd.read_csv(p_join(cache_dir, filename), sep=r'[\t\s]', header=None)[[0, 2, 6, 9, 10, 11, 12]]
|
19 |
+
df.columns = ["id", "url", "text_lid_score", "laser_score", "direction", "side", "line_no"]
|
20 |
+
print(f"load metadata: {filename}, ({len(df)} rows)")
|
21 |
+
return df
|
22 |
+
|
23 |
+
|
24 |
+
def get_audio(url: str, filename: str):
|
25 |
+
cache_dir = p_join(cache_dir_root, "audio")
|
26 |
+
if not os.path.exists(p_join(cache_dir, filename)):
|
27 |
+
return wget(url, filename=filename, cache_dir=cache_dir)
|
28 |
+
return False
|
29 |
+
|
30 |
+
|
31 |
+
def process_dataset(url_metadata):
|
32 |
+
df_metadata = get_metadata(url_metadata)
|
33 |
+
num_missing_files = 0
|
34 |
+
for _, row in tqdm(df_metadata.iterrows(), total=len(df_metadata)):
|
35 |
+
filename = f"{row['direction']}.{row['side']}.{os.path.basename(row['url'])}"
|
36 |
+
num_missing_files += not get_audio(row['url'], filename)
|
37 |
+
print(f"missing files: {num_missing_files}/{len(df_metadata)}")
|
38 |
+
|
39 |
+
|
40 |
+
if __name__ == '__main__':
|
41 |
+
process_dataset(url_metadata_s2s)
|
42 |
+
process_dataset(url_metadata_s2t)
|
43 |
|
|
|
|
requirements.txt
CHANGED
@@ -1 +1,2 @@
|
|
1 |
-
requests
|
|
|
|
1 |
+
requests
|
2 |
+
pandas
|
util.py
CHANGED
@@ -1,24 +1,26 @@
|
|
1 |
import os
|
2 |
-
from os.path import join as p_join
|
3 |
-
from typing import Optional
|
4 |
import tarfile
|
5 |
import zipfile
|
6 |
import gzip
|
|
|
7 |
import requests
|
|
|
|
|
|
|
8 |
|
9 |
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
def wget(url: str, cache_dir: str, filename: Optional[str] = None, uncompress_file: bool = True):
|
14 |
os.makedirs(cache_dir, exist_ok=True)
|
15 |
filename = os.path.basename(url) if not filename else filename
|
16 |
output_file = p_join(cache_dir, filename)
|
17 |
-
|
18 |
-
|
19 |
-
|
20 |
-
|
21 |
-
|
|
|
|
|
|
|
22 |
|
23 |
if output_file.endswith('.tar.gz') or output_file.endswith('.tgz') or output_file.endswith('.tar'):
|
24 |
if output_file.endswith('.tar'):
|
@@ -28,20 +30,13 @@ def wget(url: str, cache_dir: str, filename: Optional[str] = None, uncompress_fi
|
|
28 |
tar.extractall(cache_dir)
|
29 |
tar.close()
|
30 |
os.remove(output_file)
|
31 |
-
return output_file.replace('.tar.gz', '').replace('.tgz', '').replace('.tar', '')
|
32 |
elif output_file.endswith('.gz'):
|
33 |
with gzip.open(output_file, 'rb') as f:
|
34 |
with open(output_file.replace('.gz', ''), 'wb') as f_write:
|
35 |
f_write.write(f.read())
|
36 |
os.remove(output_file)
|
37 |
-
return output_file.replace('.gz', '')
|
38 |
elif output_file.endswith('.zip'):
|
39 |
with zipfile.ZipFile(output_file, 'r') as zip_ref:
|
40 |
zip_ref.extractall(cache_dir)
|
41 |
os.remove(output_file)
|
42 |
-
|
43 |
-
return output_file
|
44 |
-
|
45 |
-
|
46 |
-
if __name__ == '__main__':
|
47 |
-
wget()
|
|
|
1 |
import os
|
|
|
|
|
2 |
import tarfile
|
3 |
import zipfile
|
4 |
import gzip
|
5 |
+
import traceback
|
6 |
import requests
|
7 |
+
from os.path import join as p_join
|
8 |
+
from typing import Optional
|
9 |
+
from urllib3.connection import ConnectionError
|
10 |
|
11 |
|
12 |
+
def wget(url: str, cache_dir: str, filename: Optional[str] = None):
|
|
|
|
|
|
|
13 |
os.makedirs(cache_dir, exist_ok=True)
|
14 |
filename = os.path.basename(url) if not filename else filename
|
15 |
output_file = p_join(cache_dir, filename)
|
16 |
+
try:
|
17 |
+
with open(output_file, "wb") as f:
|
18 |
+
r = requests.get(url)
|
19 |
+
f.write(r.content)
|
20 |
+
except ConnectionError or KeyboardInterrupt:
|
21 |
+
traceback.print_exc()
|
22 |
+
os.remove(output_file)
|
23 |
+
return False
|
24 |
|
25 |
if output_file.endswith('.tar.gz') or output_file.endswith('.tgz') or output_file.endswith('.tar'):
|
26 |
if output_file.endswith('.tar'):
|
|
|
30 |
tar.extractall(cache_dir)
|
31 |
tar.close()
|
32 |
os.remove(output_file)
|
|
|
33 |
elif output_file.endswith('.gz'):
|
34 |
with gzip.open(output_file, 'rb') as f:
|
35 |
with open(output_file.replace('.gz', ''), 'wb') as f_write:
|
36 |
f_write.write(f.read())
|
37 |
os.remove(output_file)
|
|
|
38 |
elif output_file.endswith('.zip'):
|
39 |
with zipfile.ZipFile(output_file, 'r') as zip_ref:
|
40 |
zip_ref.extractall(cache_dir)
|
41 |
os.remove(output_file)
|
42 |
+
return True
|
|
|
|
|
|
|
|
|
|