From d4cf99d30bd5f6a8a4ababd0b9d7b06f3a479a24 Mon Sep 17 00:00:00 2001
From: "K.Takata" <kentkt@csc.jp>
Date: Thu, 1 Aug 2019 21:27:51 +0900
Subject: [PATCH] Fix out-of-bounds read in parse_char_class() (Close #139)
 
/[\x{111111}]/ causes out-of-bounds read when encoding is a single byte
encoding. \x{111111} is an invalid codepoint for a single byte encoding.
Check if it is a valid codepoint.
---
 regenc.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
 
diff -Naur a/regenc.c b/regenc.c
--- a/regenc.c  2020-10-20 14:19:32.284000000 +0800
+++ b/regenc.c  2020-10-20 14:22:55.412000000 +0800
@@ -625,18 +625,23 @@
 }
 
 extern int
-onigenc_single_byte_code_to_mbclen(OnigCodePoint code ARG_UNUSED, OnigEncoding enc ARG_UNUSED)
+onigenc_single_byte_code_to_mbclen(OnigCodePoint code, OnigEncoding enc ARG_UNUSED)
 {
+  if (code > 0xff)
+    return ONIGERR_INVALID_CODE_POINT_VALUE;
   return 1;
 }
 
 extern int
 onigenc_single_byte_code_to_mbc(OnigCodePoint code, UChar *buf, OnigEncoding enc ARG_UNUSED)
 {
+  if (code > 0xff) {
 #ifdef RUBY
-  if (code > 0xff)
     rb_raise(rb_eRangeError, "%u out of char range", code);
+#else
+    return ONIGERR_INVALID_CODE_POINT_VALUE;
 #endif
+    }
   *buf = (UChar )(code & 0xff);
   return 1;
 }