This module will handle authentication for your SharePoint Online/O365 site, allowing you to make straightforward HTTP requests from Python. It extends the commonly used Requests module, meaning that returned objects are familliar, easy to work with and well documented.

SharePy can be installed from the Python Package Index, PyPI.

pip install sharepy

import sharepy
s = sharepy.connect("example.sharepoint.com")

You will be prompted to enter your username and password, which are used to request a security token from Microsoft. An access cookie and request digest token are then retrieved and saved to properties for later use. The digest token will be refreshed automatically as it expires.

A username and password can also be provided as arguments of the connect function, if prompts are not desirable.

r = s.get("https://example.sharepoint.com/_api/web/lists/GetByTitle('Test Library')")

This will return a Requests response object. See the requests documentation for details. By default, the headers Accept: application/json; odata=verbose and Content-type: application/json; odata=verbose are sent with all requests, so API responses will be formatted as JSON where available.

Headers can be added or overridden by supplying a dictionary to the relevant method:

r = s.get("https://example.sharepoint.com/_api/...", headers={"Accept": "application/atom+xml"})

The request will send a digest header, allowing modifications to be made to SharePoint objects.

r = s.getfile("https://example.sharepoint.com/Library/Test%20File.pdf")

This will download the file to the current directory and return a response object. Alternatively you can specify a location to save the file to:

r = s.getfile("https://example.sharepoint.com/Library/Test%20File.pdf", filename="downloads/file.pdf")

Successfully uploading a file to SharePoint is a complex task which is described in detail in issue #4.

The actual file upload can be achieved with the following code, where filepath is the path to the file to upload, folder is the library on the SharePoint server and filename is the name to give to the file on upload.

with open(filepath, "rb") as f:
    data = f.read()

url = "https://example.sharepoint.com/_api/web/GetFolderByServerRelativeUrl('{}')/Files/add(url='{}',overwrite=true)"
r = s.post(url.format(folder, filename), data=data, headers={"content-length": len(data)})

Properties of the authentication session can be saved to a file using the save method, so the session can be used without having to re-authenticate each time a program is run:

s.save()

Later, the load function can be used to restore the session:

s = sharepy.load()

The default file name for saving and loading sessions is sp-session.pkl, however an alternative location can be provided as an argument to save() and load().

SharePy implements Requests authentication classes that can also be used directly with Requests itself:

import requests
import sharepy

auth = sharepy.auth.SharePointOnline(username="[email protected]")
auth.login(site="example.sharepoint.com")
r = requests.get("https://example.sharepoint.com", auth=auth)

Available authentication classes are:

• SharepointOnline - For normal SharePoint Online sites
• SharepointADFS - For ADFS-enabled sites

The authentication URL is detected automatically when using sharepy.connect(). If a different URL is required for a region-specific account, it can be specified by manually creating an auth object and setting its login_url property:

import sharepy

auth = sharepy.auth.SharePointOnline(username="[email protected]")
auth.login_url = "https://login.microsoftonline.de/extSES.srf"
s = sharepy.SharePointSession("example.sharepoint.com", auth)

• Constructing SharePoint API calls: SharePoint REST API documentation
• Handling JSON objects in Python: Python JSON module documentation

This software is distributed under the GNU General Public License v3. Copyright 2016-2021 Jonathan Holvey.

1. The authentication method used here is based on this post by Luc Stakenborg.
2. Additional help regarding request digests from sadegh's comment on this post by Paul Ryan.
3. Contributed code from @joemeneses for ADFS authentication.