MFH: fix bug #46335 (DOMText::splitText doesn't handle multibyte characters)

add test

2 files changed, 39 insertions, 4 deletions

diff --git a/ext/dom/tests/bug46335.phpt b/ext/dom/tests/bug46335.phpt
new file mode 100644
index 0000000000..bea4ae9e7d
--- /dev/null
+++ b/ext/dom/tests/bug46335.phpt
@@ -0,0 +1,35 @@
+--TEST--
+Bug #46335 (DOMText::splitText doesn't handle multibyte characters).
+--SKIPIF--
+<?php require_once('skipif.inc'); ?>
+--FILE--
+<?php 
+$textascii = 'This is an "example" of using DOM splitText';
+$text = 'This is an ‘example’ of using DOM splitText';
+$start = 30;
+$length = 3;
+
+$dom = new DOMDocument('1.0', 'UTF-8');
+$node = $dom->createTextNode($textascii);
+$dom->appendChild($node);
+
+print "Text: $node->textContent\n";
+
+$matched = $node->splitText($start);
+$matched->splitText($length);
+print "splitText (ASCII): $matched->textContent\n";
+
+$node = $dom->createTextNode($text);
+$dom->appendChild($node);
+
+print "Text: $node->textContent\n";
+
+$matched = $node->splitText($start);
+$matched->splitText($length);
+print "splitText (UTF-8): $matched->textContent\n";
+?>
+--EXPECT--
+Text: This is an "example" of using DOM splitText
+splitText (ASCII): DOM
+Text: This is an ‘example’ of using DOM splitText
+splitText (UTF-8): DOM
diff --git a/ext/dom/text.c b/ext/dom/text.c
index 26479f3186..3234b4cc25 100644
--- a/ext/dom/text.c
+++ b/ext/dom/text.c
@@ -172,19 +172,19 @@ PHP_FUNCTION(dom_text_split_text)
 	if (cur == NULL) {
 		RETURN_FALSE;
 	}
-	length = xmlStrlen(cur);
+	length = xmlUTF8Strlen(cur);
 
 	if (offset > length || offset < 0) {
 		xmlFree(cur);
 		RETURN_FALSE;
 	}
 
-	first = xmlStrndup(cur, offset);
-	second = xmlStrdup(cur + offset);
+	first = xmlUTF8Strndup(cur, offset);
+	second = xmlUTF8Strsub(cur, offset, length - offset);
 	
 	xmlFree(cur);
 
-	xmlNodeSetContentLen(node, first, offset);
+	xmlNodeSetContent(node, first);
 	nnode = xmlNewDocText(node->doc, second);
 	
 	xmlFree(first);