Fix nested list handling in Javadoc to AsciiDoc converter

[kyooosukedn]

Description

Fixes #50474

The JavadocToAsciidocTransformer previously flattened all nested HTML lists (<ul>/<ol>) to the same level, using the same marker for all items. This resulted in incorrect AsciiDoc output where nested bullet lists were indistinguishable from top-level items.

Example

This Javadoc:

/**
 * Some javadoc with list of lists:
 * <ul>
 * <li>First sentence. Second sentence.</li>
 * <li>
 * And some nested bullet list:
 * <ul>
 * <li>Element 1 with one sentence.</li>
 * <li>Element 2 with sentence 2.</li>
 * </ul>
 * </li>
 * </ul>
 */

Before (flattened):

Some javadoc with list of lists:

 - First sentence. Second sentence.
 -  And some nested bullet list:

 - Element 1 with one sentence.
 - Element 2 with sentence 2.

After (proper nesting):

Some javadoc with list of lists:

 * First sentence. Second sentence.
 * And some nested bullet list:
 ** Element 1 with one sentence.
 ** Element 2 with sentence 2.

Changes

• Track list nesting depth in the Context class
• Use AsciiDoc-standard nested list markers (* for unordered, . for ordered) repeated according to depth (e.g., ** for level 2, *** for level 3)
• Strip leading whitespace from text nodes inside list items to handle HTML formatting artifacts
• Avoid extra blank lines for nested lists (only top-level lists get trailing blank lines)
• Updated existing test to match new marker style
• Added 4 new tests:
  • parseJavaDocWithNestedUnorderedList - nested <ul> inside <ul>
  • parseJavaDocWithNestedOrderedList - nested <ol> inside <ol>
  • parseJavaDocWithDeeplyNestedLists - 3 levels of nesting
  • parseJavaDocWithMixedNestedLists - <ol> inside <ul>