Leading and trailing non-`whitespace` `Unicode whitespace` is stripped from paragraphs

The CommonMark dingus implementation currently strips non-whitespace Unicode whitespace from
the start and end of paragraphs. The CommonMark specification and the cmark implementation seem to indicate that these characters should not be stripped.

Example link

Each "space" character in this example is U+1680, OGHAM SPACE MARK, chosen from the list of Unicode whitespace characters that are not in the list of whitespace characters.

Note that the leading and trailing U+1680 characters have been trimmed from the final result.

From CommonMark 0.28, section 4.8:

The paragraph’s raw content is formed by concatenating the lines and removing initial and final [whitespace].

By comparison, in cmark 0.28.3:

Input:

 o o 
 o o 

Output:

<p> o o 
 o o </p>

Ref: https://github.com/commonmark/CommonMark/pull/465

Start by reproducing the linked U+1680 example in the CommonMark dingus implementation and compare its output with the CommonMark specification and cmark 0.28.3. Locate the paragraph whitespace-trimming logic and any related tests. Done means leading and trailing U+1680 characters remain in the rendered paragraph, while ordinary whitespace handling remains correct.