Builds an attribute list from string containing attributes.

Description

Does not modify input. May return "evil" output.
In case of unexpected input, returns false instead of stripping things.

Based on wp_kses_hair() but does not return a multi-dimensional array.

Parameters

$attrstringrequired

Attribute list from HTML element to closing HTML element tag.

Return

Source

function wp_kses_hair_parse( $attr ) {
	if ( '' === $attr ) {
		return array();
	}

	$regex =
		'(?:
				[_a-zA-Z][-_a-zA-Z0-9:.]* # Attribute name.
			|
				\[\[?[^\[\]]+\]\]?        # Shortcode in the name position implies unfiltered_html.
		)
		(?:                               # Attribute value.
			\s*=\s*                       # All values begin with "=".
			(?:
				"[^"]*"                   # Double-quoted.
			|
				\'[^\']*\'                # Single-quoted.
			|
				[^\s"\']+                 # Non-quoted.
				(?:\s|$)                  # Must have a space.
			)
		|
			(?:\s|$)                      # If attribute has no value, space is required.
		)
		\s*                               # Trailing space is optional except as mentioned above.
		';

	/*
	 * Although it is possible to reduce this procedure to a single regexp,
	 * we must run that regexp twice to get exactly the expected result.
	 *
	 * Note: do NOT remove the `x` modifiers as they are essential for the above regex!
	 */

	$validation = "/^($regex)+$/x";
	$extraction = "/$regex/x";

	if ( 1 === preg_match( $validation, $attr ) ) {
		preg_match_all( $extraction, $attr, $attrarr );
		return $attrarr[0];
	} else {
		return false;
	}
}

View all references View on Trac View on GitHub

Changelog