charcount_papers.py

from apiclient.discovery import build
from httplib2 import Http
from oauth2client import file, client, tools
from time import time
from datetime import datetime


class Document(object):
    def __init__(self, title, id):
        """Title is free, id is google docs identifier"""
        self.title = title
        self.id = id

    def char_count(self, service):
        data = service.files().export(fileId=self.id, mimeType='text/plain').execute()
        if data:
            decoded = data.decode('utf-8')
            return len([i for i in decoded if i not in ['\n', '\r', '_']])
        else:
            return False


def main():
    service = setup_api(version='v3')
    documents = [
        Document(title="Paper1", id="ID1"),
        Document(title="Paper2", id="ID2"),
        Document(title="Paper3", id="ID3"),
        Document(title="Paper4", id="ID4"),
    ]
    for paper in documents:
        wc = paper.char_count(service)
        if wc:
            now = datetime.fromtimestamp(time()).strftime('%Y-%m-%d %H:%M:%S')
            print(f"{paper.title}\t{now}\t{wc}")


def setup_api(version='v2'):
    """
    Load api, or open browser for authorization
    If you adapt the scopes you have to delete the credentials.json file
    to enable reauthorization
    """
    SCOPES = (
        'https://www.googleapis.com/auth/drive.metadata.readonly',
        'https://www.googleapis.com/auth/drive.file',
        'https://www.googleapis.com/auth/drive',
        'https://www.googleapis.com/auth/spreadsheets.readonly',
    )
    store = file.Storage('credentials.json')
    client_secret = 'client_secret.json'
    creds = store.get()
    if not creds or creds.invalid:
        flow = client.flow_from_clientsecrets(client_secret, SCOPES)
        creds = tools.run_flow(flow, store)
    return build('drive', version, http=creds.authorize(Http()))


if __name__ == '__main__':
    main()