From 6d03e5e5540a5196db48e1e4350de45f4e9bc619 Mon Sep 17 00:00:00 2001
From: Zig Tan <zig@moodle.com>
Date: Fri, 29 Jun 2018 12:25:41 +0800
Subject: [PATCH 1/2] MDL-61052 assign: Improve strip_images() to avoid HTML5
 tags errors

---
 .../editpdf/classes/document_services.php     | 27 +++++++++----------
 1 file changed, 13 insertions(+), 14 deletions(-)

diff --git a/mod/assign/feedback/editpdf/classes/document_services.php b/mod/assign/feedback/editpdf/classes/document_services.php
index fa6dbfbb598..71477b7c435 100644
--- a/mod/assign/feedback/editpdf/classes/document_services.php
+++ b/mod/assign/feedback/editpdf/classes/document_services.php
@@ -120,25 +120,24 @@ EOD;
      * @return string New html with no image tags.
      */
     protected static function strip_images($html) {
+        // Load HTML and suppress any parsing errors (DOMDocument->loadHTML() does not current support HTML5 tags).
         $dom = new DOMDocument();
-        $dom->loadHTML("<?xml version=\"1.0\" encoding=\"UTF-8\" ?>" . $html);
-        $images = $dom->getElementsByTagName('img');
-        $i = 0;
+        libxml_use_internal_errors(true);
+        $dom->loadHTML('<?xml version="1.0" encoding="UTF-8" ?>' . $html);
+        libxml_clear_errors();
 
-        for ($i = ($images->length - 1); $i >= 0; $i--) {
-            $node = $images->item($i);
+        // Find all img tags.
+        if ($imgnodes = $dom->getElementsByTagName('img')) {
+            // Replace img nodes with the img alt text without overriding DOM elements.
+            for ($i = ($imgnodes->length - 1); $i >= 0; $i--) {
+                $imgnode = $imgnodes->item($i);
+                $alt = ($imgnode->hasAttribute('alt')) ? ' [ ' . $imgnode->getAttribute('alt') . ' ] ' : ' ';
+                $textnode = $dom->createTextNode($alt);
 
-            if ($node->hasAttribute('alt')) {
-                $replacement = ' [ ' . $node->getAttribute('alt') . ' ] ';
-            } else {
-                $replacement = ' ';
+                $imgnode->parentNode->replaceChild($textnode, $imgnode);
             }
-
-            $text = $dom->createTextNode($replacement);
-            $node->parentNode->replaceChild($text, $node);
         }
-        $count = 1;
-        return str_replace("<?xml version=\"1.0\" encoding=\"UTF-8\" ?>", "", $dom->saveHTML(), $count);
+        return preg_replace('/^<\?xml[^>[]*(\[[^]]*\])?>/', '', $dom->saveHTML());
     }
 
     /**

From 3bea0fa387ba507165028450d658473645647853 Mon Sep 17 00:00:00 2001
From: Damyon Wiese <damyon@moodle.com>
Date: Mon, 2 Jul 2018 14:35:39 +0800
Subject: [PATCH 2/2] MDL-61052 assign: Do not queue conversions

Identical submissions do not require individual conversions from the document converter. Text passing through
format_text cannot be trusted and is likely to be unique each time, even for the same text.
---
 mod/assign/feedback/editpdf/classes/document_services.php | 3 ++-
 mod/assign/submission/onlinetext/locallib.php             | 7 +++----
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/mod/assign/feedback/editpdf/classes/document_services.php b/mod/assign/feedback/editpdf/classes/document_services.php
index 71477b7c435..384ccdd2bdd 100644
--- a/mod/assign/feedback/editpdf/classes/document_services.php
+++ b/mod/assign/feedback/editpdf/classes/document_services.php
@@ -137,7 +137,8 @@ EOD;
                 $imgnode->parentNode->replaceChild($textnode, $imgnode);
             }
         }
-        return preg_replace('/^<\?xml[^>[]*(\[[^]]*\])?>/', '', $dom->saveHTML());
+        $count = 1;
+        return str_replace("<?xml version=\"1.0\" encoding=\"UTF-8\" ?>", "", $dom->saveHTML(), $count);
     }
 
     /**
diff --git a/mod/assign/submission/onlinetext/locallib.php b/mod/assign/submission/onlinetext/locallib.php
index b2365b31192..e5e8e1933cb 100644
--- a/mod/assign/submission/onlinetext/locallib.php
+++ b/mod/assign/submission/onlinetext/locallib.php
@@ -398,10 +398,9 @@ class assign_submission_onlinetext extends assign_submission_plugin {
         // Note that this check is the same logic as the result from the is_empty function but we do
         // not call it directly because we already have the submission record.
         if ($onlinetextsubmission && !empty($onlinetextsubmission->onlinetext)) {
-            $finaltext = $this->assignment->download_rewrite_pluginfile_urls($onlinetextsubmission->onlinetext, $user, $this);
-            $formattedtext = format_text($finaltext,
-                                         $onlinetextsubmission->onlineformat,
-                                         array('context'=>$this->assignment->get_context()));
+            // Do not pass the text through format_text. The result may not be displayed in Moodle and
+            // may be passed to external services such as document conversion or portfolios.
+            $formattedtext = $this->assignment->download_rewrite_pluginfile_urls($onlinetextsubmission->onlinetext, $user, $this);
             $head = '<head><meta charset="UTF-8"></head>';
             $submissioncontent = '<!DOCTYPE html><html>' . $head . '<body>'. $formattedtext . '</body></html>';