jonny/scrape-dandi
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Merge branch 'main' of https://git.jon-e.net/jonny/scrape-dandi

Browse source
This commit is contained in:
sneakers-the-rat 2023-10-30 12:18:16 -07:00
commit e89e8cf913
1 changed files with 25 additions and 15 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

40
scrape_dandi/__init__.py
View file

@ -1,7 +1,8 @@
from pathlib import Path
from tqdm import trange
from datetime import datetime
import sys
import click
from dandi.consts import ZARR_EXTENSIONS, metadata_all_fields
from dandi.dandiarchive import DandisetURL, _dandi_url_parser, parse_dandi_url
from dandi.cli.cmd_download import download
@ -11,11 +12,11 @@ OUT_DIR = Path('/mnt/seedbank/p2p/dandi/')
#OUT_DIR = '.'
LOG_TXT = 'log.txt'
SKIP_DANDISETS = [
'000108' # humongous 372 human light sheet imaging
108 # humongous 372 human light sheet imaging
]
DANDI_ID = 'DANDI:{:06d}'
MAX_DANDISET = 683
JOBS = 64
def check_nwb(dandiset:int) -> bool:
@ -36,22 +37,31 @@ def check_nwb(dandiset:int) -> bool:
def main():
for i in trange(MAX_DANDISET):
if not check_nwb(i):
if not check_nwb(i) or i in SKIP_DANDISETS:
with open(LOG_TXT, 'a') as lfile:
lfile.write(f"{datetime.now().isoformat()} - {i:03d} - SKIP\n")
continue
id = DANDI_ID.format(i)
download(
[
id,
'-o', str(OUT_DIR),
'--existing', 'refresh',
'--jobs', '24'
]
)
with open(LOG_TXT, 'a') as lfile:
lfile.write(f"{datetime.now().isoformat()} - {i:03d} - GET\n")
try:
download.main([
id,
'-o', str(OUT_DIR),
'--existing', 'refresh',
'--jobs', str(JOBS)
], standalone_mode=False
)
print('\nafterdl')
with open(LOG_TXT, 'a') as lfile:
lfile.write(f"{datetime.now().isoformat()} - {i:03d} - GET\n")
except KeyboardInterrupt:
sys.exit(1)
except Exception as e:
print('\nexception')
with open(LOG_TXT, 'a') as lfile:
lfile.write(f"{datetime.now().isoformat()} - {i:03d} - ERROR\n")
lfile.write(str(e))
# continue