explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-01-13 02:36:32 +03:00
Code Issues Packages Projects Releases Wiki Activity

fixed extract directory

Browse Source
This commit is contained in:
Henning Peters 2015-10-21 07:59:34 +02:00
parent da4c9cee06
commit ccffd2ef53
1 changed files with 18 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
spacy/en/download.py
View File

@ -1,5 +1,4 @@
from __future__ import print_function from __future__ import print_function
from os import path
import sys import sys
import os import os
import tarfile import tarfile
@ -15,43 +14,44 @@ AWS_STORE = 'https://s3-us-west-1.amazonaws.com/media.spacynlp.com'
ALL_DATA_DIR_URL = '%s/en_data_all-%s.tgz' % (AWS_STORE, VERSION) ALL_DATA_DIR_URL = '%s/en_data_all-%s.tgz' % (AWS_STORE, VERSION)
DEST_DIR = path.join(path.dirname(path.abspath(__file__)), 'data') DEST_DIR = os.path.dirname(os.path.abspath(__file__))
def download_file(url, path): def download_file(url, download_path):
return uget.download(url, path, console=sys.stdout) return uget.download(url, download_path, console=sys.stdout)
def install_data(url, path, filename): def install_data(url, extract_path, download_path):
try: try:
os.makedirs(path) os.makedirs(extract_path)
except FileExistsError: except FileExistsError:
pass pass
download_path = os.path.join(path, filename)
tmp = download_file(url, download_path) tmp = download_file(url, download_path)
assert tmp == download_path assert tmp == download_path
t = tarfile.open(download_path) t = tarfile.open(download_path)
t.extractall(path) t.extractall(extract_path)
@plac.annotations( @plac.annotations(
force=("Force overwrite", "flag", "f", bool), force=("Force overwrite", "flag", "f", bool),
) )
def main(data_size='all', force=False): def main(data_size='all', force=False):
if force and path.exists(DEST_DIR):
shutil.rmtree(DEST_DIR)
filename = ALL_DATA_DIR_URL.rsplit('/', 1)[1] filename = ALL_DATA_DIR_URL.rsplit('/', 1)[1]
download_path = os.path.join(DEST_DIR, filename)
data_path = os.path.join(DEST_DIR, 'data')
if os.path.exists(DEST_DIR): if force and os.path.exists(download_path):
# ugly hack to find out whether something other os.unlink(download_path)
# than the currently wanted file lives there
if len([f for f in os.listdir(DEST_DIR) if f != filename]): if force and os.path.exists(data_path):
shutil.rmtree(data_path)
if os.path.exists(data_path):
print('data already installed at %s, overwrite with --force' % DEST_DIR) print('data already installed at %s, overwrite with --force' % DEST_DIR)
sys.exit(1) sys.exit(1)
install_data(ALL_DATA_DIR_URL, DEST_DIR, filename) install_data(ALL_DATA_DIR_URL, DEST_DIR, download_path)
if __name__ == '__main__': if __name__ == '__main__':