jplu/stanfordnlprestapi

Stanford NLP framework wrapped with a REST API.

stanfordnlprestapi's Issues

Reduce fat jar size

Remove the Stanford CoreNLP models from the fat jar.

Fix README

The README file needs to be cleaned up (typos, links for collaborators, better phrasing)

Better Dockerfile handling

The current version of the Dockerfile need to be modified everytime the version changes. Find a way to compile the project directly when building the Docker image.

setupenv.py is Python2 only

Here is a possible solution that should work with both Python2 and 3 :

try:
    from urllib2 import urlopen
except ImportError: # Python 3
    from urllib.request import urlopen
import zipfile
import progressbar

url = "http://adel.eurecom.fr/data/models-gazetteers-properties.zip"
bar = progressbar.ProgressBar()
file_name = url.split('/')[-1]
u = urlopen(url)

with open(file_name, 'wb') as f:
    file_size =  int(u.headers['Content-Length'])

    print("Downloading: %s Bytes: %s" % (file_name, file_size))

    for i in bar(range(0, file_size, 8192)):
        buffer = u.read(8192)

        if not buffer:
            break

        f.write(buffer)

print("Unzip " + file_name)

zip_ref = zipfile.ZipFile("models-gazetteers-properties.zip")
number_files = len(zip_ref.infolist())

bar.update(0)

for i in bar(range(number_files)):
    zip_ref.extract(zip_ref.infolist()[i])

zip_ref.close()