O365 Toolkit : Maintain multiple user logins

[ironjunk]

Objective: I am trying to develop a small python app (Streamlit) which utilizes the Office365 Toolkit to schedule events on the outlook calendar for logged in outlook users using prompts.

What I have implemented so far:

1. I have referred to Office365 | Langchain documentation for the basic implementation.
2. I have successfully completed all the pre-requisites i.e. registered an App on Azure my-app ⇾ set up necessary API permissions ⇾ save the Application (Client) ID and Client Secret.
3. I have successfully implemented and authorized my-app by following the auth flow and giving consent using my outlook user [email protected].
4. A file named o365_token.txt is generated in the project root folder.
5. I have successfully implemented and tested prompt input and can see the appropriate outcome on my outlook calendar.

What I am struggling with: As next step in development, I wish to now enable my app to handle multiple user auth requests i.e. trigger the auth flow for a different user [email protected] while maintaining my original user [email protected]'s session.

My Question: Is this something I can achieve with the O365 Toolkit directly? If not, is there some other way of implementing this requirement that you could suggest?

P.S.: I am new to using Langchain and O365 Toolkit. Apologies if this query comes off as basic or if this is something that is already discussed or solved.

[dosubot[bot]]

🤖

Hello,

Thank you for reaching out and for your detailed explanation of your issue. Your question is not basic at all, and it's great to see you're making progress with your application.

To handle multiple user authentication requests in your Python app using the Office365 Toolkit, you can indeed modify the _auth method in the O365BaseLoader class. Currently, the method uses a single token file to store the authentication token. This approach will not work for multiple users because it will overwrite the existing token each time a new user authenticates.

To handle multiple users, you need to store the tokens in separate files for each user. You can achieve this by modifying the FileSystemTokenBackend to use a different token file for each user. You can use the user's unique identifier (like user ID or username) to create a unique token file for each user.

Here's how you can modify the _auth method:

def _auth(self, user_id: str) -> Account:
    """Authenticates the OneDrive API client
    Args:
        user_id: The unique identifier for the user.

    Returns:
        The authenticated Account object.
    """
    try:
        from O365 import Account, FileSystemTokenBackend
    except ImportError:
        raise ImportError(
            "O365 package not found, please install it with `pip install o365`"
        )
    token_backend = FileSystemTokenBackend(
        token_path=Path.home() / ".credentials" / user_id
    )
    account = Account(
        credentials=(
            self.settings.client_id,
            self.settings.client_secret.get_secret_value(),
        ),
        scopes=self._scopes,
        token_backend=token_backend,
        **{"raise_http_errors": False},
    )
    # make the auth
    account.authenticate()
    return account

In this modified _auth method, the user_id parameter is added. This user_id is used to create a unique token file for each user in the .credentials directory. Now, when a new user authenticates, a new token file will be created for that user, and it will not overwrite the existing user's token. This allows for different users to trigger the auth flow while maintaining the original user's session.

Please note that this modification is not directly supported by the LangChain framework, and you will need to make this modification in your own code. Also, please ensure that you handle the user IDs and token files securely to prevent unauthorized access.

I hope this helps! If you have any other questions, feel free to ask.

Best,
Dosu

Sources

• libs/langchain/langchain/document_loaders/base_o365.py

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.

[kevinstyle]

@ironjunk Hi, I want to try O365 Tool in langchain too, could you please tell me how to apply Application (Client) ID and Client Secret ? grateful