[k-takata/Onigmo] Fix out-of-bounds read in parse_char_class()

(Close
https://github.com/k-takata/Onigmo/pull/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.

https://github.com/k-takata/Onigmo/commit/d4cf99d30b

1 files changed, 7 insertions, 2 deletions

diff --git a/regenc.c b/regenc.c
index ff2155aabc..c561cee29d 100644
--- a/regenc.c
+++ b/regenc.c
@@ -640,18 +640,23 @@ onigenc_single_byte_mbc_to_code(const UChar* p, const UChar* end ARG_UNUSED,
 }
 
 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;
 }