Psuedo-class selector :text producing errors and incorrect results

CHANGELOG.md

@@ -3,6 +3,13 @@ QueryPath Changelog
 
 # Unreleased changes
 
+- Fix fatal error when running a CSS selector against a match set that contains non-element nodes (text, comment, CDATA
+  or processing instruction). Those nodes now simply do not match, instead of calling element-only DOM methods on them.
+- Fix the `:text` pseudo-class so it matches jQuery: it selects `input` elements whose `type` attribute is absent or is
+  `text` (case-insensitively). It never indicated, and still does not indicate, whether a node is a text node.
+- `find('*')` now matches the nodes in the match set as well as their descendants, so that a selector can be tested
+  against an element already in hand. Note that #73 replaces this with jQuery's descendant-only `find()`; this entry
+  is provisional and should be dropped if that lands first.
 - Reorganise, modernise, and repair the `examples/` directory. Each example now lives in its own subdirectory with an `index.php`, and the full set is indexed in `examples/quickstart-guide.md`
 - Convert the remaining legacy examples: `simple_example.php`, `techniques.php`, `svg.php`, `rss.php`, `odt.php`, `parse_php.php`, and `sparql.php`
 - Fix examples that no longer ran: send a `User-Agent` where remote hosts now require one, resolve paths relative to the example rather than the working directory, and stop relying on the removed `qp.php` autoloader and the PHP 8 incompatible `eachLambda()`

src/CSS/DOMTraverser.php

@@ -7,6 +7,7 @@
 
 use DOMDocument;
 use DOMElement;
+use DOMNode;
 use DOMNodeList;
 use DOMXPath;
 use QueryPath\CSS\DOMTraverser\Util;
@@ -177,14 +178,14 @@ public function matches()
 	 * absolutely huge selectors or for versions of PHP tuned to
 	 * strictly limit recursion depth.
 	 *
-	 * @param DOMElement $node
+	 * @param DOMNode $node
 	 *   The DOMNode to check.
 	 * @param             $selector
 	 *
 	 * @return boolean
 	 *   A boolean TRUE if the node matches, false otherwise.
 	 */
-	public function matchesSelector(DOMElement $node, $selector)
+	public function matchesSelector(DOMNode $node, $selector)
 	{
 		return $this->matchesSimpleSelector($node, $selector, 0);
 	}
@@ -196,16 +197,24 @@ public function matchesSelector(DOMElement $node, $selector)
 	 * this checks only a simple selector (plus an optional
 	 * combinator).
 	 *
-	 * @param DOMElement $node
+	 * @param DOMNode $node
 	 * @param             $selectors
 	 * @param             $index
 	 *
 	 * @return boolean
 	 *   A boolean TRUE if the node matches, false otherwise.
 	 * @throws NotImplementedException
 	 */
-	public function matchesSimpleSelector(DOMElement $node, $selectors, $index)
+	public function matchesSimpleSelector(DOMNode $node, $selectors, $index)
 	{
+		// Selectors only ever match elements. A match set may legitimately
+		// contain text, comment, CDATA or processing instruction nodes (e.g.
+		// from contents()), and those simply do not match -- rather than
+		// blowing up on the element-only DOM API used below.
+		if (! $node instanceof DOMElement) {
+			return false;
+		}
+
 		$selector = $selectors[$index];
 		// Note that this will short circuit as soon as one of these
 		// returns FALSE.
@@ -257,7 +266,7 @@ public function matchesSimpleSelector(DOMElement $node, $selectors, $index)
 	 * @return boolean
 	 *   TRUE if the next selector(s) match.
 	 */
-	public function combine(DOMElement $node, $selectors, $index)
+	public function combine(DOMNode $node, $selectors, $index)
 	{
 		$selector = $selectors[$index];
 		//$this->debug(implode(' ', $selectors));
@@ -476,6 +485,11 @@ protected function initialMatchOnID(SimpleSelector $selector, SplObjectStorage $
 		// Now we try to find any matching IDs.
 		/** @var DOMElement $node */
 		foreach ($matches as $node) {
+			// Non-element nodes have neither attributes nor element children.
+			if (! $node instanceof DOMElement) {
+				continue;
+			}
+
 			if ($node->getAttribute('id') === $id) {
 				$found->offsetSet($node);
 			}
@@ -520,6 +534,11 @@ protected function initialMatchOnClasses(SimpleSelector $selector, SplObjectStor
 		// Now we try to find any matching IDs.
 		/** @var DOMElement $node */
 		foreach ($matches as $node) {
+			// Non-element nodes have neither attributes nor element children.
+			if (! $node instanceof DOMElement) {
+				continue;
+			}
+
 			// Refactor me!
 			if ($node->hasAttribute('class')) {
 				$intersect = array_intersect($selector->classes, explode(' ', $node->getAttribute('class')));
@@ -585,11 +604,18 @@ protected function initialMatchOnElement(SimpleSelector $selector, SplObjectStor
 			$element = '*';
 		}
 		$found = $this->newMatches();
-		/** @var DOMDocument $node */
+		/** @var DOMDocument|DOMElement $node */
 		foreach ($matches as $node) {
-			// Capture the case where the initial element is the root element.
-			if ($node->tagName === $element
-				|| ($element === '*' && $node->parentNode instanceof DOMDocument)) {
+			// Only elements and documents can contain elements. Text, comment,
+			// CDATA and processing instruction nodes never match, and do not
+			// support the element-only API used below.
+			if (! $node instanceof DOMElement && ! $node instanceof DOMDocument) {
+				continue;
+			}
+
+			// Capture the case where the node itself matches the element.
+			if ($node instanceof DOMElement
+				&& ($element === '*' || $node->tagName === $element)) {
 				$found->offsetSet($node);
 			}
 			$nl = $node->getElementsByTagName($element);

src/CSS/DOMTraverser/PseudoClass.php

@@ -12,6 +12,7 @@
 
 namespace QueryPath\CSS\DOMTraverser;
 
+use DOMElement;
 use QueryPath\CSS\DOMTraverser;
 use QueryPath\CSS\NotImplementedException;
 use QueryPath\CSS\EventHandler;
@@ -46,6 +47,13 @@ class PseudoClass
 	 */
 	public function elementMatches($pseudoclass, $node, $scope, $value = null)
 	{
+		// Pseudo-classes are only ever satisfied by elements. Text, comment,
+		// CDATA and processing instruction nodes have no tag name, attributes
+		// or element children, so they can never match.
+		if (! $node instanceof DOMElement) {
+			return false;
+		}
+
 		$name = strtolower($pseudoclass);
 		// Need to handle known pseudoclasses.
 		switch ($name) {
@@ -160,6 +168,8 @@ public function elementMatches($pseudoclass, $node, $scope, $value = null)
 			case 'checked':
 				return Util::matchesAttribute($node, $name);
 			case 'text':
+				return $this->isTextInput($node);
+
 			case 'radio':
 			case 'checkbox':
 			case 'file':
@@ -226,6 +236,25 @@ protected function lang($node, $value)
 		return false;
 	}
 
+	/**
+	 * Provides jQuery pseudoclass ':text'.
+	 *
+	 * This mirrors jQuery, where `:text` selects `input` elements of type text
+	 * -- that is, an `input` whose `type` attribute is either absent (`text` is
+	 * the default type of an `input`) or is `text`, matched case-insensitively.
+	 *
+	 * It does NOT indicate whether the node is a text node.
+	 *
+	 * @param DOMElement $node
+	 *
+	 * @return bool
+	 * @see https://api.jquery.com/text-selector/
+	 */
+	protected function isTextInput($node): bool
+	{
+		return Util::isTextInput($node);
+	}
+
 	/**
 	 * Provides jQuery pseudoclass ':header'.
 	 *

src/CSS/DOMTraverser/Util.php

@@ -7,6 +7,7 @@
 
 namespace QueryPath\CSS\DOMTraverser;
 
+use DOMElement;
 use QueryPath\CSS\EventHandler;
 
 /**
@@ -26,6 +27,12 @@ class Util
 	 */
 	public static function matchesAttribute($node, $name, $value = null, $operation = EventHandler::IS_EXACTLY): bool
 	{
+		// Only elements have attributes. Text, comment, CDATA and processing
+		// instruction nodes can never match an attribute selector.
+		if (! $node instanceof DOMElement) {
+			return false;
+		}
+
 		if (! $node->hasAttribute($name)) {
 			return false;
 		}
@@ -47,6 +54,11 @@ public static function matchesAttributeNS(
 		$value = null,
 		$operation = EventHandler::IS_EXACTLY
 	) {
+		// Only elements have attributes.
+		if (! $node instanceof DOMElement) {
+			return false;
+		}
+
 		if (! $node->hasAttributeNS($nsuri, $name)) {
 			return false;
 		}
@@ -165,4 +177,27 @@ public static function parseAnB($rule): array
 
 		return [$aVal, $bVal];
 	}
+
+	/**
+	 * Does this node match jQuery's :text pseudo-class?
+	 *
+	 * jQuery's :text selects input elements whose type attribute is absent, or is "text"
+	 * regardless of case. It says nothing about whether a node is a text node.
+	 *
+	 * Both selector engines ask this question, so they share one answer — they are meant to
+	 * agree, and two copies of the rule would be free to drift apart.
+	 *
+	 * @param mixed $node
+	 *
+	 * @return bool
+	 */
+	public static function isTextInput($node): bool
+	{
+		if (! $node instanceof DOMElement || strtolower($node->localName) !== 'input') {
+			return false;
+		}
+
+		// An input with no type attribute defaults to a text input.
+		return ! $node->hasAttribute('type') || strtolower($node->getAttribute('type')) === 'text';
+	}
 }

src/CSS/QueryPathEventHandler.php

@@ -353,6 +353,30 @@ public function attribute($name, $value = null, $operation = EventHandler::IS_EX
 		$this->findAnyElement = false;
 	}
 
+	/**
+	 * Helper function for the jQuery ':text' pseudo-class.
+	 *
+	 * As in jQuery, ':text' selects `input` elements of type text -- that is, an
+	 * `input` whose `type` attribute is either absent (`text` is the default
+	 * type of an `input`) or is `text`, matched case-insensitively. It does NOT
+	 * indicate whether the node is a text node.
+	 *
+	 * @see https://api.jquery.com/text-selector/
+	 */
+	protected function textInput()
+	{
+		$found   = new SplObjectStorage();
+		$matches = $this->candidateList();
+		foreach ($matches as $item) {
+			if (Util::isTextInput($item)) {
+				$found->offsetSet($item);
+			}
+		}
+
+		$this->matches        = $found;
+		$this->findAnyElement = false;
+	}
+
 	/**
 	 * Helper function to find all elements with exact matches.
 	 *
@@ -557,6 +581,8 @@ public function pseudoClass($name, $value = null)
 				$this->attribute($name);
 				break;
 			case 'text':
+				$this->textInput();
+				break;
 			case 'radio':
 			case 'checkbox':
 			case 'file':