401 with scoped access token

[lobeck]

I have created a scoped PAT with scopes:

read:page:confluence
write:page:confluence

Due to #1199, the client is then initiated as follows:

confluence = Confluence(
    url=CONFLUENCE_BASE_URL,
    username=CONFLUENCE_USER,
    password=CONFLUENCE_API_TOKEN,
    cloud=True
)

Now when doing a call like

page = confluence.get_page_by_id(CONFLUENCE_PAGE_ID)

It results in a requests.exceptions.HTTPError: Unauthorized (401)

It works when using an unscoped access token. This is version 4.0.4.

[dddlr]

curious if you were able to find a workaround for this, apart from using an unscoped access token?

[lobeck]

No, I have't tested further. I'm waiting on some clarifying comments from Atlassian. Also on the other library issues.

[dddlr]

All good, I think atlassian-python-api is an unofficial repo unrelated to atlassian, and i don't have any domain knowledge on confluence REST APIs 😅 hopefully @gonchik has capacity to look into it

[hajimegane]

According to the information provided on this page, the endpoint changes when using a scoped token.

I was able to access it with the following setting:

CONFLUENCE_BASE_URL=https://api.atlassian.com/ex/confluence/{cloudid}

We are able to find the cloudid at https://{your-domain}.atlassian.net/_edge/tenant_info.

[lobeck]

@dddlr you're right, this project is unofficial, but only mentions it in the fine print. From the appearance it seemed official :(

@hajimegane Thank you for the hint. This got it working, but the API Documentation is incomplete, so it's a guesswork to find the right scopes. Created several tickets with Atlassian, so they improve their documentation.

[ksaurabhAparavi]

@lobeck did you get it working with the scope? Can you please provide a small working example? Thanks

[lobeck]

@ksaurabhAparavi no. I also asked Atlassian support about it and they noted, that scopes only work with the v2 API which is not implemented in this library yet.