From ab972a838c308eb00655ae74d3eaed4aad32ce1d Mon Sep 17 00:00:00 2001
From: "J. King" <jking@jkingweb.ca>
Date: Mon, 15 Feb 2021 11:34:45 -0500
Subject: [PATCH] Fix DOCTYPE serialization

Also patch all top-level comments out
---
 lib/DOM/Document.php                |  1 -
 lib/DOM/DocumentType.php            | 16 ----------------
 tests/cases/TestTreeConstructor.php | 26 ++++++++++++++++++--------
 3 files changed, 18 insertions(+), 25 deletions(-)
 delete mode 100644 lib/DOM/DocumentType.php

diff --git a/lib/DOM/Document.php b/lib/DOM/Document.php
index 4126f28..7ed8448 100644
--- a/lib/DOM/Document.php
+++ b/lib/DOM/Document.php
@@ -10,7 +10,6 @@ class Document extends \DOMDocument {
 
         $this->registerNodeClass('DOMComment', '\dW\HTML5\Comment');
         $this->registerNodeClass('DOMDocumentFragment', '\dW\HTML5\DocumentFragment');
-        $this->registerNodeClass('DOMDocumentType', '\dW\HTML5\DocumentType');
         $this->registerNodeClass('DOMElement', '\dW\HTML5\Element');
         $this->registerNodeClass('DOMProcessingInstruction', '\dW\HTML5\ProcessingInstruction');
         $this->registerNodeClass('DOMText', '\dW\HTML5\Text');
diff --git a/lib/DOM/DocumentType.php b/lib/DOM/DocumentType.php
deleted file mode 100644
index 0d7552f..0000000
--- a/lib/DOM/DocumentType.php
+++ /dev/null
@@ -1,16 +0,0 @@
-<?php
-declare(strict_types=1);
-namespace dW\HTML5;
-
-class DocumentType extends \DOMDocumentType {
-    function __toString(): string {
-        # Append the literal string "<!DOCTYPE" (U+003C LESS-THAN SIGN, U+0021
-        # EXCLAMATION MARK, U+0044 LATIN CAPITAL LETTER D, U+004F LATIN CAPITAL LETTER
-        # O, U+0043 LATIN CAPITAL LETTER C, U+0054 LATIN CAPITAL LETTER T, U+0059 LATIN
-        # CAPITAL LETTER Y, U+0050 LATIN CAPITAL LETTER P, U+0045 LATIN CAPITAL LETTER
-        # E), followed by a space (U+0020 SPACE), followed by the value of current
-        # node’s name IDL attribute, followed by the literal string ">" (U+003E
-        # GREATER-THAN SIGN).
-        return "<!DOCTYPE {$this->name}>";
-    }
-}
diff --git a/tests/cases/TestTreeConstructor.php b/tests/cases/TestTreeConstructor.php
index ad0dad9..c61bd39 100644
--- a/tests/cases/TestTreeConstructor.php
+++ b/tests/cases/TestTreeConstructor.php
@@ -33,7 +33,10 @@ class TestTreeConstructor extends \PHPUnit\Framework\TestCase {
             $this->markTestIncomplete("Fragment tests still to be implemented");
         }
         // certain tests need to be patched to ignore unavoidable limitations of PHP's DOM
-        $exp = $this->patchTest($data, $fragment, $exp);
+        [$exp, $patched] = $this->patchTest($data, $fragment, $exp);
+        if ($patched) {
+            $this->markAsRisky();
+        }
         // convert parse error constants into standard symbols in specification
         $errorMap = array_map(function($str) {
             return strtolower(str_replace("_", "-", $str));
@@ -72,11 +75,15 @@ class TestTreeConstructor extends \PHPUnit\Framework\TestCase {
     }
 
     protected function patchTest(string $data, $fragment, array $exp): array {
-        if (strpos($exp[0], "| <!--") === 0) {
-            // comments before the DOCTYPE are silently dropped by the PHP DOM
-            array_shift($exp);
+        $patched = false;
+        // comments outside the root element are silently dropped by the PHP DOM
+        for ($a = 0; $a < sizeof($exp); $a++) {
+            if (strpos($exp[$a], "| <!--") === 0) {
+                array_splice($exp, $a--, 1);
+                $patched = true;
+            }
         }
-        return $exp;
+        return [$exp, $patched];
     }
 
     protected function push(string $data): void {
@@ -87,9 +94,12 @@ class TestTreeConstructor extends \PHPUnit\Framework\TestCase {
         $this->out = [];
         $this->depth = 0;
         if ($d->doctype) {
-            $dt = "<!DOCTYPE ".$d->doctype->name;
-            $dt .= strlen($d->doctype->publicId) ? ' "'.$d->doctype->publicId.'"' : "";
-            $dt .= strlen($d->doctype->systemId) ? ' "'.$d->doctype->systemId.'"' : "";
+            $dt = "<!DOCTYPE ";
+            $dt .= ($d->doctype->name !== " ") ? $d->doctype->name : "";
+            if (strlen($d->doctype->publicId) || strlen($d->doctype->systemId)) {
+                $dt .= ' "'.$d->doctype->publicId.'"';
+                $dt .= ' "'.$d->doctype->systemId.'"';
+            }
             $dt .= ">";
             $this->push($dt);
         }