HTML API: Backport updates from Core (WordPress#63723)

Updates the HTML API files in the 6.7 compatability layer, bringing
in code that has been merged into Core.

Co-authored-by: Jon Surrell <[email protected]>

lib/compat/wordpress-6.7/html-api/class-gutenberg-html-active-formatting-elements-6-7.php

@@ -0,0 +1,187 @@
+<?php
+/**
+ * HTML API: WP_HTML_Active_Formatting_Elements class
+ *
+ * @package WordPress
+ * @subpackage HTML-API
+ * @since 6.4.0
+ */
+
+/**
+ * Core class used by the HTML processor during HTML parsing
+ * for managing the stack of active formatting elements.
+ *
+ * This class is designed for internal use by the HTML processor.
+ *
+ * > Initially, the list of active formatting elements is empty.
+ * > It is used to handle mis-nested formatting element tags.
+ * >
+ * > The list contains elements in the formatting category, and markers.
+ * > The markers are inserted when entering applet, object, marquee,
+ * > template, td, th, and caption elements, and are used to prevent
+ * > formatting from "leaking" into applet, object, marquee, template,
+ * > td, th, and caption elements.
+ * >
+ * > In addition, each element in the list of active formatting elements
+ * > is associated with the token for which it was created, so that
+ * > further elements can be created for that token if necessary.
+ *
+ * @since 6.4.0
+ *
+ * @access private
+ *
+ * @see https://html.spec.whatwg.org/#list-of-active-formatting-elements
+ * @see WP_HTML_Processor
+ */
+class Gutenberg_HTML_Active_Formatting_Elements_6_7 {
+	/**
+	 * Holds the stack of active formatting element references.
+	 *
+	 * @since 6.4.0
+	 *
+	 * @var WP_HTML_Token[]
+	 */
+	private $stack = array();
+
+	/**
+	 * Reports if a specific node is in the stack of active formatting elements.
+	 *
+	 * @since 6.4.0
+	 *
+	 * @param WP_HTML_Token $token Look for this node in the stack.
+	 * @return bool Whether the referenced node is in the stack of active formatting elements.
+	 */
+	public function contains_node( Gutenberg_HTML_Token_6_7 $token ) {
+		foreach ( $this->walk_up() as $item ) {
+			if ( $token->bookmark_name === $item->bookmark_name ) {
+				return true;
+			}
+		}
+
+		return false;
+	}
+
+	/**
+	 * Returns how many nodes are currently in the stack of active formatting elements.
+	 *
+	 * @since 6.4.0
+	 *
+	 * @return int How many node are in the stack of active formatting elements.
+	 */
+	public function count() {
+		return count( $this->stack );
+	}
+
+	/**
+	 * Returns the node at the end of the stack of active formatting elements,
+	 * if one exists. If the stack is empty, returns null.
+	 *
+	 * @since 6.4.0
+	 *
+	 * @return WP_HTML_Token|null Last node in the stack of active formatting elements, if one exists, otherwise null.
+	 */
+	public function current_node() {
+		$current_node = end( $this->stack );
+
+		return $current_node ? $current_node : null;
+	}
+
+	/**
+	 * Pushes a node onto the stack of active formatting elements.
+	 *
+	 * @since 6.4.0
+	 *
+	 * @see https://html.spec.whatwg.org/#push-onto-the-list-of-active-formatting-elements
+	 *
+	 * @param WP_HTML_Token $token Push this node onto the stack.
+	 */
+	public function push( Gutenberg_HTML_Token_6_7 $token ) {
+		/*
+		 * > If there are already three elements in the list of active formatting elements after the last marker,
+		 * > if any, or anywhere in the list if there are no markers, that have the same tag name, namespace, and
+		 * > attributes as element, then remove the earliest such element from the list of active formatting
+		 * > elements. For these purposes, the attributes must be compared as they were when the elements were
+		 * > created by the parser; two elements have the same attributes if all their parsed attributes can be
+		 * > paired such that the two attributes in each pair have identical names, namespaces, and values
+		 * > (the order of the attributes does not matter).
+		 *
+		 * @todo Implement the "Noah's Ark clause" to only add up to three of any given kind of formatting elements to the stack.
+		 */
+		// > Add element to the list of active formatting elements.
+		$this->stack[] = $token;
+	}
+
+	/**
+	 * Removes a node from the stack of active formatting elements.
+	 *
+	 * @since 6.4.0
+	 *
+	 * @param WP_HTML_Token $token Remove this node from the stack, if it's there already.
+	 * @return bool Whether the node was found and removed from the stack of active formatting elements.
+	 */
+	public function remove_node( Gutenberg_HTML_Token_6_7 $token ) {
+		foreach ( $this->walk_up() as $position_from_end => $item ) {
+			if ( $token->bookmark_name !== $item->bookmark_name ) {
+				continue;
+			}
+
+			$position_from_start = $this->count() - $position_from_end - 1;
+			array_splice( $this->stack, $position_from_start, 1 );
+			return true;
+		}
+
+		return false;
+	}
+
+	/**
+	 * Steps through the stack of active formatting elements, starting with the
+	 * top element (added first) and walking downwards to the one added last.
+	 *
+	 * This generator function is designed to be used inside a "foreach" loop.
+	 *
+	 * Example:
+	 *
+	 *     $html = '<em><strong><a>We are here';
+	 *     foreach ( $stack->walk_down() as $node ) {
+	 *         echo "{$node->node_name} -> ";
+	 *     }
+	 *     > EM -> STRONG -> A ->
+	 *
+	 * To start with the most-recently added element and walk towards the top,
+	 * see WP_HTML_Active_Formatting_Elements::walk_up().
+	 *
+	 * @since 6.4.0
+	 */
+	public function walk_down() {
+		$count = count( $this->stack );
+
+		for ( $i = 0; $i < $count; $i++ ) {
+			yield $this->stack[ $i ];
+		}
+	}
+
+	/**
+	 * Steps through the stack of active formatting elements, starting with the
+	 * bottom element (added last) and walking upwards to the one added first.
+	 *
+	 * This generator function is designed to be used inside a "foreach" loop.
+	 *
+	 * Example:
+	 *
+	 *     $html = '<em><strong><a>We are here';
+	 *     foreach ( $stack->walk_up() as $node ) {
+	 *         echo "{$node->node_name} -> ";
+	 *     }
+	 *     > A -> STRONG -> EM ->
+	 *
+	 * To start with the first added element and walk towards the bottom,
+	 * see WP_HTML_Active_Formatting_Elements::walk_down().
+	 *
+	 * @since 6.4.0
+	 */
+	public function walk_up() {
+		for ( $i = count( $this->stack ) - 1; $i >= 0; $i-- ) {
+			yield $this->stack[ $i ];
+		}
+	}
+}

lib/compat/wordpress-6.7/html-api/class-gutenberg-html-attribute-token-6-7.php

@@ -0,0 +1,116 @@
+<?php
+/**
+ * HTML API: WP_HTML_Attribute_Token class
+ *
+ * @package WordPress
+ * @subpackage HTML-API
+ * @since 6.2.0
+ */
+
+/**
+ * Core class used by the HTML tag processor as a data structure for the attribute token,
+ * allowing to drastically improve performance.
+ *
+ * This class is for internal usage of the WP_HTML_Tag_Processor class.
+ *
+ * @access private
+ * @since 6.2.0
+ * @since 6.5.0 Replaced `end` with `length` to more closely match `substr()`.
+ *
+ * @see WP_HTML_Tag_Processor
+ */
+class Gutenberg_HTML_Attribute_Token_6_7 {
+	/**
+	 * Attribute name.
+	 *
+	 * @since 6.2.0
+	 *
+	 * @var string
+	 */
+	public $name;
+
+	/**
+	 * Attribute value.
+	 *
+	 * @since 6.2.0
+	 *
+	 * @var int
+	 */
+	public $value_starts_at;
+
+	/**
+	 * How many bytes the value occupies in the input HTML.
+	 *
+	 * @since 6.2.0
+	 *
+	 * @var int
+	 */
+	public $value_length;
+
+	/**
+	 * The string offset where the attribute name starts.
+	 *
+	 * @since 6.2.0
+	 *
+	 * @var int
+	 */
+	public $start;
+
+	/**
+	 * Byte length of text spanning the attribute inside a tag.
+	 *
+	 * This span starts at the first character of the attribute name
+	 * and it ends after one of three cases:
+	 *
+	 *  - at the end of the attribute name for boolean attributes.
+	 *  - at the end of the value for unquoted attributes.
+	 *  - at the final single or double quote for quoted attributes.
+	 *
+	 * Example:
+	 *
+	 *     <div class="post">
+	 *          ------------ length is 12, including quotes
+	 *
+	 *     <input type="checked" checked id="selector">
+	 *                           ------- length is 6
+	 *
+	 *     <a rel=noopener>
+	 *        ------------ length is 11
+	 *
+	 * @since 6.5.0 Replaced `end` with `length` to more closely match `substr()`.
+	 *
+	 * @var int
+	 */
+	public $length;
+
+	/**
+	 * Whether the attribute is a boolean attribute with value `true`.
+	 *
+	 * @since 6.2.0
+	 *
+	 * @var bool
+	 */
+	public $is_true;
+
+	/**
+	 * Constructor.
+	 *
+	 * @since 6.2.0
+	 * @since 6.5.0 Replaced `end` with `length` to more closely match `substr()`.
+	 *
+	 * @param string $name         Attribute name.
+	 * @param int    $value_start  Attribute value.
+	 * @param int    $value_length Number of bytes attribute value spans.
+	 * @param int    $start        The string offset where the attribute name starts.
+	 * @param int    $length       Byte length of the entire attribute name or name and value pair expression.
+	 * @param bool   $is_true      Whether the attribute is a boolean attribute with true value.
+	 */
+	public function __construct( $name, $value_start, $value_length, $start, $length, $is_true ) {
+		$this->name            = $name;
+		$this->value_starts_at = $value_start;
+		$this->value_length    = $value_length;
+		$this->start           = $start;
+		$this->length          = $length;
+		$this->is_true         = $is_true;
+	}
+}