MFH: Fixed bug #35447 (xml_parse_into_struct() chokes on the UTF-8 BOM)

add test

2 files changed, 50 insertions, 5 deletions

diff --git a/ext/xml/compat.c b/ext/xml/compat.c
index 43e116b864..7be85cab22 100644
--- a/ext/xml/compat.c
+++ b/ext/xml/compat.c
@@ -405,15 +405,12 @@ XML_ParserCreate_MM(const XML_Char *encoding, const XML_Memory_Handling_Suite *m
 		efree(parser);
 		return NULL;
 	}
-	if (encoding != NULL) {
-		parser->parser->encoding = xmlStrdup(encoding);
 #if LIBXML_VERSION <= 20617
 	/* for older versions of libxml2, allow correct detection of
 	 * charset in documents with a BOM: */
-	} else {
-		parser->parser->charset = XML_CHAR_ENCODING_NONE;
+	parser->parser->charset = XML_CHAR_ENCODING_NONE;
 #endif
-	}
+
 	parser->parser->replaceEntities = 1;
 	parser->parser->wellFormed = 0;
 	if (sep != NULL) {
diff --git a/ext/xml/tests/bug35447.phpt b/ext/xml/tests/bug35447.phpt
new file mode 100644
index 0000000000..ba8b81e46c
--- /dev/null
+++ b/ext/xml/tests/bug35447.phpt
@@ -0,0 +1,48 @@
+--TEST--
+Bug #35447 (xml_parse_into_struct() chokes on the UTF-8 BOM)
+--SKIPIF--
+<?php
+if (! @xml_parser_create_ns('ISO-8859-1')) { die("skip xml_parser_create_ns is not supported on this plattform");}
+?>
+--FILE--
+<?php
+$data = <<<END_OF_XML
+\xEF\xBB\xBF<?xml version="1.0" encoding="utf-8"?\x3e
+<!DOCTYPE bundle [
+    <!ELEMENT bundle (resource)+>
+    <!ELEMENT resource (#PCDATA)>
+    <!ATTLIST resource
+              key CDATA #REQUIRED
+              type (literal|pattern|sub) "literal"
+              >
+]>
+<resource key="rSeeYou">A bient&amp;244;t</resource>
+END_OF_XML;
+
+$parser = xml_parser_create_ns('UTF-8');
+xml_parser_set_option($parser,XML_OPTION_CASE_FOLDING,0);
+$result = xml_parse_into_struct($parser, $data, $vals, $index);
+xml_parser_free($parser);
+var_dump($vals);
+?>
+--EXPECT--
+array(1) {
+  [0]=>
+  array(5) {
+    ["tag"]=>
+    string(8) "resource"
+    ["type"]=>
+    string(8) "complete"
+    ["level"]=>
+    int(1)
+    ["attributes"]=>
+    array(2) {
+      ["key"]=>
+      string(7) "rSeeYou"
+      ["type"]=>
+      string(7) "literal"
+    }
+    ["value"]=>
+    string(13) "A bient&244;t"
+  }
+}