Charset: Add explanatory note about what consitutes “valid” UTF-8.

This patch adds a clarifying note about what constitutes a valid UTF-8 byte stream. This was brought up in review as a potentially ambiguous term, so a link to the spec has been provided to fix the behavior to the standard.

Developed in #9716
Discussed in https://core.trac.wordpress.org/ticket/38044

Follow-up to [60630].

Props dmsnell, agulbra.
See #38044.


git-svn-id: https://develop.svn.wordpress.org/trunk@60702 602fd350-edb4-49c9-b593-d223f7449a82

Author: dmsnell

src/wp-includes/formatting.php

@@ -940,6 +940,13 @@ function seems_utf8( $str ) {
  *                                                     // E.g. The “ü” in ISO-8859-1 is a single byte 0xFC,
  *                                                     // but in UTF-8 is the two-byte sequence 0xC3 0xBC.
  *
+ * A “valid” string consists of “well-formed UTF-8 code unit sequence[s],” meaning
+ * that the bytes conform to the UTF-8 encoding scheme, all characters use the minimal
+ * byte sequence required by UTF-8, and that no sequence encodes a UTF-16 surrogate
+ * code point or any character above the representable range.
+ *
+ * @see https://www.unicode.org/versions/Unicode16.0.0/core-spec/chapter-3/#G32860
+ *
  * @see _wp_is_valid_utf8_fallback
  *
  * @since 6.9.0