handle error response and fail fast

[grexe]

currently, when the login fails for some reason, it seems the MarkDown parser is still called and consequently returns a followup error message because it cannot parse the real error response:

Unable to upload markdown file, test.md:
	Error checking for existing page: invalid character '<' looking for beginning of value

The real error is printed first, but not handled and the program tries to continue:

RRO[0000] Unable to unmarshal API response. Received: '<?xml version="1.0" encoding="UTF-8" standalone="yes"?><status><status-code>401</status-code><message>Request rejected because issuer is either not authorized or not authorized to impersonate</message></status>'

The way the error is handled is a bit misleading and not very intuitive, as it hides the root cause and it looks as if the file contained a markup error.

[bosborne]

Seeing the same error, I think it's because my Confluence wiki won't accept user/pass, it only accepts logins using an API token.

[justmiles]

You can use API tokens. I need to update the README to explain it, but set your username as the email and the password as an API token generated from https://id.atlassian.com/manage/api-tokens

This issue is definitely warranted, though. It should check auth and fail before trying parse the markdown and upload.

related: #3

[bosborne]

Works!

[bosborne]

Minor note: if CONFLUENCE_ENDPOINT has a trailing '/' it does not work, had to remove that.

[finchr]

Minor note: if CONFLUENCE_ENDPOINT has a trailing '/' it does not work, had to remove that.

Thanks! This solved it for me.