🐛 Stamp sd-card-text on direct paragraphs only, never replace classes

[chrisjsewell]

Fixes two related bugs in how sd-card-text is applied to paragraphs in cards and dropdowns.

Changes

• Dropdown transform no longer wipes user classes: an inverted conditional ([] if "classes" in para else para["classes"] — always true for docutils elements) meant every paragraph in every dropdown body had its existing classes replaced by ["sd-card-text"]. Any user class (e.g. via rst-class or MyST attrs) was silently destroyed. Now appends.
• Stamping is limited to direct child paragraphs in both the dropdown transform and CardDirective.add_card_child_classes: previously all descendant paragraphs (inside nested admonitions, lists, nested tab-sets, …) were stamped, causing the spacing artefacts in Multi-line content in dropdown has extra spacing on first line #40. Nested content now renders with its natural spacing, consistent with the same content outside a card.

Verification

• New tests/test_card_text.py (rst + myst): user-class preservation, direct-vs-nested stamping for card, dropdown, and grid-item-card (the delegated path), and the Multi-line content in dropdown has extra spacing on first line #40 two-paragraph reproduction. The regression cases were verified failing against the unfixed code.
• Zero existing regression fixtures change: audited — existing fixtures only ever contained direct-child sd-card-text paragraphs (which is why Multi-line content in dropdown has extra spacing on first line #40 went undetected).
• Shipped-docs impact audit: the only rendering delta across the whole docs build is two tab-panel paragraphs inside the "tabs in dropdown" example losing the class — provably pixel-identical, since .sd-tab-content>:last-child{margin-bottom:0 !important} already forces the same margin either way.
• No CSS changes. Follow-up recorded for the card redesign: consider moving stamping to a CSS child selector entirely.

Closes #40

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 90.42%. Comparing base (6e06c98) to head (525cda2).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #278      +/-   ##
==========================================
+ Coverage   90.40%   90.42%   +0.01%     
==========================================
  Files          13       13              
  Lines        1147     1149       +2     
==========================================
+ Hits         1037     1039       +2     
  Misses        110      110              

Flag	Coverage Δ
pytests	90.42% <100.00%> (+0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ 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.