chore: Migrate PropertyFilter tokens to internal Token

[ernst-dev]

Description

Replace PropertyFilter's inline token implementation with the shared internal Token component (src/token/internal.tsx), following the same pattern established in b24a0ee (file-token migration).

What changed:

• src/property-filter/filtering-token/index.tsx: The local TokenGroup component's token-box rendering (the inner <div class="token"> + content div + action button) is now delegated to InternalToken with role="presentation" (the outer div[role="group"] still provides the accessibility grouping). __customContent carries the popover-trigger content and the dismiss/edit button; __tokenBoxClassName applies the filtering-token's own border/background/radius classes on top of the token-box mixin; disableInnerPadding removes the token-box default padding so the content div's own padding takes effect.
• src/property-filter/filtering-token/styles.scss: Removed the border-inline-start divider from .dismiss-button, .inner-dismiss-button, and .edit-button. This is the one intentional visual change: the divider between the token label and the dismiss/edit button is gone.

What is NOT changed:

• Public API surface of PropertyFilter (no prop changes, no interface changes)
• Accessibility attributes: role="group", aria-label, aria-disabled on the outer group wrapper are all preserved exactly
• The internal Token component (src/token/internal.tsx) — not touched
• Grouped-token / and-or operation rendering and dismiss behavior
• Test-util selectors (filtering-token-dismiss-button, filtering-token-inner-dismiss-button, etc.) — preserved by keeping TokenDismissButton with its original class names

Related links, issue #, if available: n/a

How has this been tested?

Unit tests: All 345 existing PropertyFilter tests pass (17 test suites including filtering-token.test.tsx with 14 tests). No tests weakened. No snapshot tests affected (no snapshot tests in this suite).

Visual parity: Captured before/after screenshots of property-filter/property-filter-tokens-permutations dev page (single token, token with operation, grouped tokens, long labels, disabled states). Pixel-by-pixel diff: zero differences (PIL ImageChops.difference bounding box = None). The only intentional visual change — divider removal — is confirmed by code inspection of the removed border-inline-start line in styles.scss.

Evidence screenshots:

• /home/ernstka/workplace/chat-bubble-reference/pf-tokens-before-permutations.png
• /home/ernstka/workplace/chat-bubble-reference/pf-tokens-after-permutations.png

Build: gulp build succeeds. TypeScript compilation clean (tsc --noEmit no errors). ESLint clean on changed files.

Review checklist

The following items are to be evaluated by the author(s) and the reviewer(s).

Correctness

• Changes include appropriate documentation updates. — No public-facing documentation changes needed; this is an internal refactor.
• Changes are backward-compatible if not indicated, see CONTRIBUTING.md. — ✅ No public API changes. The one intentional visual change (divider removal) is by design.
• Changes do not include unsupported browser features, see CONTRIBUTING.md. — ✅ No new browser features used.
• Changes were manually tested for accessibility, see accessibility guidelines. — ✅ role="group", aria-label, aria-disabled on group wrappers preserved exactly. role="presentation" on the inner InternalToken prevents a redundant nested group from being exposed to assistive tech.

Security

• If the code handles URLs: all URLs are validated through the checkSafeUrl function. — No URL handling in this change.

Testing

• Changes are covered with new/existing unit tests? — ✅ All 14 filtering-token unit tests pass. All 345 PropertyFilter tests pass.
• Changes are covered with new/existing integration tests? — Existing integration tests cover PropertyFilter token behavior.

---

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.60%. Comparing base (bcde582) to head (ee831c8).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #4754      +/-   ##
==========================================
- Coverage   97.60%   97.60%   -0.01%     
==========================================
  Files         951      951              
  Lines       30744    30739       -5     
  Branches    11281    11276       -5     
==========================================
- Hits        30007    30002       -5     
  Misses        690      690              
  Partials       47       47              

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.