Auth failure with on-prem jira

[rcorre]

My config file looks like this:

jira_api_username: 'me@company.com'
jira_api_token: 'my_token'
jira_api_base_url: 'https://jira.my.company.com'
cloud: False

jiratui fails like so:

❯ jiratui issues search -p XXX
Client error '401 Unauthorized' for url 'https://jira.my.company.com/rest/api/2/search'

If I run curl "https://jira.my.company.com/rest/api/2/search" -H "Authorization: Bearer my_token", it works fine. Any idea what I might be doing wrong?

One odd thing I noticed is that jiratui requires an api username even if I provide a token, whereas with curl it seems I only need the token.

[EoinFarrell]

I was having the same issue, dug in and found JIRA documentation for on premise states:

Using PATs
To use a personal access token for authentication, you have to pass it as a bearer token in the Authorization header of a REST API call.

This can be controlled using the use_bearer_authentication config field in jiratui

jira_api_username: 'xxx@xxx.com'
jira_api_token: 'xxxxxxx'
jira_api_base_url: 'https://jira.company.com/'
cloud: False
use_bearer_authentication: True

As you mentioned the username should not be required as its not used with bearer auth, which is visible in the jiratui auth logic:

class JiraTUIBearerAuth(httpx.Auth):
    def __init__(self, token: str, username: str | None = None):
        self.token = token

    def auth_flow(self, request):
        request.headers['Authorization'] = f'Bearer {self.token.strip()}'
        yield request

but as it's not set with a default value in the config class, the click package considers it a mandatory field.

[rcorre]

use_bearer_authentication worked, thanks!