Make auth callback URL relative to base API_URL

[jake-low]

What type of PR is this? (check all applicable)

• 🍕 Feature
• 🐛 Bug Fix
• 📝 Documentation
• 🧑‍💻 Refactor
• ✅ Test
• 🤖 Build or CI
• ❓ Other (please specify)

Describe this PR

This is a tweak to the changes made in #6933. That PR works correctly AFAIK for tasks.hotosm.org's use case, but I discovered when backporting the fix to osmus/tasking-manager (which powers tasks.openstreetmap.us and tasks.teachosm.org) that it does not work if the backend is deployed to a nonstandard URL path.

This change makes the callback URL string a relative path (removing the hardcoded /api/v2 prefix). The appropriate prefix will then be prepended in fetchLocalJSONAPI() when the string is converted to a URL object. This prefix may be /api/v2 or it may be another value depending on the configuration environment variables provided at runtime.

This change is running in production on tasks.openstreetmap.us and tasks.teachosm.org already, so I am fairly confident that it works.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[ramyaragupathy]

@suzit-10 can you please review this PR?

[suzit-10]

Thank you @jake-low for raising this 🙏
Removing the hardcoded /api/v2 is definitely a good improvement, and it’s very helpful for instances with non-standard backend URL (e.g., without /api/v2).
Besides the authorization URL, there are a few other places with the same prefix — could you update those too? I think it will also need a small tweak to fetchLocalJSONAPI(), since it’s used across the codebase, and simply prepending may not work in all cases (e.g., where api/v2 isn’t expected).

Really appreciate you taking the time to look into this 🙌

[jake-low]

Hi @suzit-10, thanks for reviewing this. I unfortunately don't have time to revisit this PR and modify or retest it. I believe it fixes a genuine bug, although it may not fix all occurrences of that bug since I was focused only on fixing the code introduced in #6933.

Feel free to either merge this PR or close it at your discretion.

[suzit-10]

Hi @jake-low, thanks for the update! I appreciate you taking the time to address this. I'll proceed with the required changes across the code and merge the PR.