test: establish Playwright E2E UI testing foundation (#426)

[tejassinghbhati]

Linked Issue

Closes #426
Related #58, #14

---

Existing Related Work Reviewed

Issues/PRs reviewed: #426, #58, #14

---

Overlap Assessment

• Classification: Structural Testing Addition / CI Expansion
• Overlapping items: None

Why this is not duplicate/superseded:
This PR introduces an End-to-End (E2E) headless browser testing framework. It builds sequentially on the recent API unit-testing foundations but targets the previously untested Vanilla JS WebAPP frontend.

---

Why this PR Should Proceed

Automated E2E tests are essential for ensuring that future UI modifications—especially complex CSS/HTML updates from upstream OSeMOSYS/MUIO merges—do not silently break critical WebAPP behaviors and workflows.

This establishes a strict frontend quality gate before any PR is merged into main.

---

Summary

What Changed

• Established pyproject.toml to manage pytest-playwright dependencies without bloating production requirements.
• Engineered a robust live_server fixture in conftest.py:
  • Spins up Flask backend on an isolated port (5003)
  • Runs on a background thread
• Replaced arbitrary timeout guesses with a dynamic health-check polling loop:
  • Uses requests
  • Max timeout: 15 seconds
  • Prevents flaky CI failures on slow runners
• Added Playwright-based UI smoke tests (test_ui_smoke.py):
  • Uses getByText and getByRole
  • Verifies:
    • DOM boot
    • Base layout rendering
    • SPA routes (#/Parameters, #/ModelFile)
• Added GitHub Actions workflow (ui-tests.yml):
  • Installs headless Chromium
  • Runs E2E tests automatically on PRs to main

---

Why

The frontend previously relied entirely on manual verification.

E2E tests with Playwright simulate real user workflows, validating:

• UI rendering (DOM)
• Client-side logic
• Flask API integration

This ensures structural issues are caught immediately.

---

Validation

• Tests added/updated (or not applicable)
• Validation steps documented
• Evidence attached (logs/screenshots/output as relevant)

  Note: Tests run directly in the GitHub Actions pipeline

---

Documentation

• Docs updated in this PR (or not applicable)
• Setup/workflow changes reflected in repository documentation

---

Scope Check

• No unrelated refactors
• Implemented from a feature branch
• Change is deliverable without upstream OSeMOSYS/MUIO dependency
• Base repo/branch: EAPD-DRB/MUIOGO:main (not upstream)

[tejassinghbhati]

@m13v again a lot of thanks for your guidance. I just formally opened the PR linked to that issue.

I made sure to tightly integrate the 15-second cold-start polling limits, the strict 5003 isolation mapping, and the exact Text/Role Playwright locators we were discussing. Would love to get your eyes on it when you have a moment to see if the testing structure feels right to you.

Without your suggestions this would have been way more rough around the edges and prone to CI flakiness.