Rewrite normalizeDocument

Author: tvdijen

src/XML/DOMDocumentFactory.php

@@ -16,7 +16,6 @@
 use function restore_error_handler;
 use function set_error_handler;
 use function sprintf;
-use function strpos;
 
 /**
  * @package simplesamlphp/xml-common
@@ -168,53 +167,47 @@ public static function create(string $encoding = 'UTF-8'): Dom\XMLDocument
      * @return \Dom\XMLDocument The same document instance, potentially modified. If the document has no root element
      *                          or no namespace declarations to normalize, it is returned unchanged.
      */
-    public static function normalizeDocument(Dom\XMLDocument $doc): Dom\XMLDocument
+    public static function normalizeDocument(Dom\Document $doc): Dom\Document
     {
-        // Get the root element
         $root = $doc->documentElement;
         if ($root === null) {
             return $doc;
         }
 
         $xpath = XPath::getXPath($doc);
+
         $xmlnsAttributes = [];
 
-        // Collect namespace declarations needed for prefixed elements in the document
-        foreach ($xpath->query('//*[namespace::*]') as $node) {
-            if ($node instanceof Dom\Element) {
-                $name = 'xmlns:' . $node->prefix;
-                // Both prefix and namespaceURI NULL equals the default xmlns:xml namespace
-                if ($node->prefix !== null && $node->namespaceURI !== null) {
-                    $xmlnsAttributes[$name] = $node->namespaceURI;
+        // More reliable collection: look for attributes starting with xmlns on every element
+        foreach ($xpath->query('//*') as $element) {
+            if ($element instanceof Dom\Element) {
+                foreach ($element->attributes as $attr) {
+                    if (str_starts_with($attr->nodeName, 'xmlns')) {
+                        $name = $attr->nodeName;
+                        $value = $attr->nodeValue;
+                        $xmlnsAttributes[$name] = $value;
+                    }
                 }
             }
         }
 
-        // If no xmlns attributes found, return early with debug info
-        if (empty($xmlnsAttributes)) {
-            return $doc;
-        }
-
-        // Remove xmlns attributes from all elements (proper XMLNS namespace removal)
-        foreach ($xpath->query('//*[namespace::*]') as $node) {
-            if (!$node instanceof Dom\Element) {
-                continue;
-            }
-
-            foreach ($node->attributes as $attr) {
-                if ($attr->namespaceURI === C::NS_XMLNS) {
-                    $node->removeAttributeNS(C::NS_XMLNS, $attr->localName);
-                    continue;
+        // Remove from all elements
+        foreach ($xpath->query('//*') as $element) {
+            if ($element instanceof Dom\Element) {
+                $toRemove = [];
+                foreach ($element->attributes as $attr) {
+                    if (str_starts_with($attr->nodeName, 'xmlns')) {
+                        $toRemove[] = $attr->nodeName;
+                    }
                 }
 
-                if (strpos($attr->nodeName, 'xmlns') === 0 || $attr->nodeName === 'xmlns') {
-                    // Fallback for implementations that still expose xmlns attrs without namespaceURI
-                    $node->removeAttribute($attr->nodeName);
+                foreach ($toRemove as $name) {
+                    $element->removeAttribute($name);
                 }
             }
         }
 
-        // Add all collected xmlns attributes to the root element
+        // Re-attach to root
         foreach ($xmlnsAttributes as $name => $value) {
             $root->setAttributeNS(C::NS_XMLNS, $name, $value);
         }

src/XML/SerializableElementTrait.php

@@ -31,23 +31,18 @@ trait SerializableElementTrait
     public function __toString(): string
     {
         $xml = $this->toXML();
-
-        /** @var \Dom\XMLDocument $ownerDocument */
-        $ownerDocument = $xml->ownerDocument;
-        $xmlString = $ownerDocument->saveXml($ownerDocument->documentElement);
-
-        $doc = DOMDocumentFactory::fromString($xmlString);
-        $doc->formatOutput = $this->formatOutput;
+        $doc = $xml->ownerDocument;
 
         if (static::getNormalization() === true) {
-            /** @var \Dom\XMLDocument $normalized */
             $normalized = DOMDocumentFactory::normalizeDocument($doc);
-            $normalized->normalize();
-            return $normalized->saveXml($normalized->firstChild);
+            $normalized->formatOutput = $this->formatOutput;
+            return $normalized->saveXml($normalized->documentElement);
         }
 
-        $doc->normalize();
-        return $doc->saveXml($doc->firstChild);
+        // Non-normalized path
+        $doc->formatOutput = $this->formatOutput;
+        $doc->normalizeDocument();
+        return $doc->saveXml($doc->documentElement);
     }
 
 

tests/resources/schemas/simplesamlphp.xsd

@@ -1,13 +1,4 @@
 <?xml version="1.0" encoding="utf-8"?>
-<!DOCTYPE schema
-  PUBLIC "-//W3C//DTD XMLSchema 200102//EN" "http://www.w3.org/2001/XMLSchema.dtd"
- [
-   <!ATTLIST schema 
-     xmlns:ds CDATA #FIXED "urn:x-simplesamlphp:namespace">
-   <!ENTITY ssp 'urn:x-simplesamlphp:namespace'>
-   <!ENTITY % p ''>
-   <!ENTITY % s ''>
-  ]>
 
 <!-- Schema for SimpleSAMLphp dummy classes -->
 

tests/resources/xml/ssp_ExtendableElement.xml

@@ -1,4 +1,4 @@
-<ssp:ExtendableElement xmlns:ssp="urn:x-simplesamlphp:namespace" xmlns:dummy="urn:custom:dummy">
-  <ssp:Chunk>some</ssp:Chunk>
-  <dummy:Chunk>some</dummy:Chunk>
+<ssp:ExtendableElement xmlns:ssp="urn:x-simplesamlphp:namespace">
+  <ssp:Chunk xmlns:ssp="urn:x-simplesamlphp:namespace">some</ssp:Chunk>
+  <dummy:Chunk xmlns:dummy="urn:custom:dummy">some</dummy:Chunk>
 </ssp:ExtendableElement>