The fix for unintended markdown is too greedy

[coffee2code]

Both ​wp_transition_comment_status() and ​wp_transition_post_status() (and likely others) have consecutive hook names that contain underscores. Unfortunately, as is known, underscores in function descriptions are treated as Markdown by the parser (and thus become <em> and </em> tags).

fix_unintended_markdown() exists to fix these. However, its regex is too greedy, fixing only the unintended markup bookending a series of unintended markup. It should capture <em> and </em> pairs ungreedily.

[coffee2code]

In 5579:

developer.wordpress.org: Change regex in fix_unintended_markdown() to be ungreedy.

Examples:

• ​https://developer.wordpress.org/reference/functions/wp_transition_comment_status/
• ​https://developer.wordpress.org/reference/functions/wp_transition_post_status/

Fixes #2881.