From 015bc7a0f22c5b462ba9c373ee8079daee8775dd Mon Sep 17 00:00:00 2001
From: mathieugrimault <30295397+mathieugrimault@users.noreply.github.com>
Date: Wed, 19 Dec 2018 17:27:07 +0100
Subject: [PATCH] Improve tolerance of illegal XML characters (#1290)
---
.../src/VuFind/RecordDriver/MarcAdvancedTrait.php | 12 ++++++++----
module/VuFind/src/VuFind/XSLT/Processor.php | 5 ++++-
2 files changed, 12 insertions(+), 5 deletions(-)
diff --git a/module/VuFind/src/VuFind/RecordDriver/MarcAdvancedTrait.php b/module/VuFind/src/VuFind/RecordDriver/MarcAdvancedTrait.php
index 5d0bd86e54e..a470005933d 100644
--- a/module/VuFind/src/VuFind/RecordDriver/MarcAdvancedTrait.php
+++ b/module/VuFind/src/VuFind/RecordDriver/MarcAdvancedTrait.php
@@ -853,11 +853,15 @@ trait MarcAdvancedTrait
{
// Special case for MARC:
if ($format == 'marc21') {
- $xml = $this->getMarcRecord()->toXML();
- $xml = str_replace(
- [chr(27), chr(28), chr(29), chr(30), chr(31), chr(8)], ' ', $xml
+ $sanitizeXmlRegEx
+ = '[^\x{0009}\x{000a}\x{000d}\x{0020}-\x{D7FF}\x{E000}-\x{FFFD}]+';
+ $xml = simplexml_load_string(
+ trim(
+ preg_replace(
+ "/$sanitizeXmlRegEx/u", ' ', $this->getMarcRecord()->toXML()
+ )
+ )
);
- $xml = simplexml_load_string($xml);
if (!$xml || !isset($xml->record)) {
return false;
}
diff --git a/module/VuFind/src/VuFind/XSLT/Processor.php b/module/VuFind/src/VuFind/XSLT/Processor.php
index df99053ad84..38ca7469161 100644
--- a/module/VuFind/src/VuFind/XSLT/Processor.php
+++ b/module/VuFind/src/VuFind/XSLT/Processor.php
@@ -58,7 +58,10 @@ class Processor
$xsl = new XSLTProcessor();
$xsl->importStyleSheet($style);
$doc = new DOMDocument();
- if ($doc->loadXML($xml)) {
+ $sanitizeXmlRegEx
+ = '[^\x{0009}\x{000a}\x{000d}\x{0020}-\x{D7FF}\x{E000}-\x{FFFD}]+';
+ $cleanXml = trim(preg_replace("/$sanitizeXmlRegEx/u", ' ', $xml));
+ if ($doc->loadXML($cleanXml)) {
foreach ($params as $key => $value) {
$xsl->setParameter('', $key, $value);
}
--
GitLab