Simple Google Spreadsheets to Pandas DataFrame in IPython Notebook

sample.ipynb

sample.pdf

How exactly does one retrieve valid Client ID and Client Secret? I attempted to do so by parsing the JSON returned by "Create Credentials" > "Service Account Key" at the following link?
https://console.developers.google.com/apis/credentials?project=PROJECT-NAME

The JSON file I've retrieved only has the following keys (and the "client_id" is only an integer, with no domain after a hyphen):
private_key
private_key_id
token_uri
auth_provider_x509_cert_url
auth_uri
client_email
client_id
project_id
type
client_x509_cert_url

The following link might be helpful: https://developers.google.com/identity/protocols/application-default-credentials?hl=en_US

Thanks in advance for any clues. If this is a common stumbling block for people, I could add further instructions to the documentation.

TS

btw, if you are (like us) on python3 with a buggy httplib2 that refuses to connect to a proxy server, the hack below makes oauthlib2 use requests instead of httplib2

class DropInHttplib():
    def __init__(self):
        self.sess = requests.session()

    def request(self, uri, method='GET' , body=None , headers=None):
        resp = None
        if method.lower() == 'get':
            resp = self.sess.get(uri, headers=headers)

        elif method.lower() == 'post':
            resp = self.sess.post(uri,data=body,headers=headers)
        if resp.status_code == 200:
            resp.status = httplib2.http.client.OK
        else:
            resp.status = httplib2.http.client.BAD_REQUEST
        return (resp,resp.text)

then, one can do

httpclient = DropInHttplib()
credentials = oauth2client.tools.run_flow(flow, storage, flags, httpclient)

and we needed to modify the flags so that it would work on a server (where we could not open a browser window)

flags = argparse.ArgumentParser(parents=[oauth2client.tools.argparser]).parse_args(["--noauth_local_webserver"])