setup.py

from six.moves.urllib import request
from os.path import dirname, abspath, exists, join
from os import mkdir
import gzip
import shutil
import nltk

PROJECT_ROOT = dirname(abspath(__file__))

data_path = join(PROJECT_ROOT, 'data')
if not exists(data_path):
    mkdir(data_path)

for i in range(1, 928):
    num = str(i).zfill(4)
    print('Downloading pubmed18n{}.xml.gz ...'.format(num))
    request.urlretrieve(
        'ftp://ftp.ncbi.nlm.nih.gov/pubmed/baseline/pubmed18n{}.xml.gz'.format(num),
        join(data_path, 'pubmed18n{}.xml.gz'.format(num)))
    print('Extracting pubmed18n{}.xml.gz ...'.format(num))
    with gzip.open(join(data_path, 'pubmed18n{}.xml.gz'.format(num)), 'rb') as f_in:
        with open(join(data_path, 'pubmed18n{}.xml'.format(num)), 'wb') as f_out:
            shutil.copyfileobj(f_in, f_out)

print('Downloading snp_pubmed_cited.gz ...')
request.urlretrieve(
        'ftp://ftp.ncbi.nlm.nih.gov/snp/Entrez/eLinks/snp_pubmed_cited.gz',
        join(data_path, 'snp_pubmed_cited.gz'))
print('Extracting snp_pubmed_cited.gz ...')
with gzip.open(join(data_path, 'snp_pubmed_cited.gz'), 'rb') as f_in:
    with open(join(data_path, 'snp_pubmed_cited'), 'wb') as f_out:
            shutil.copyfileobj(f_in, f_out)

nltk.download('punkt')
nltk.download('stopwords')
nltk.download('wordnet')