re.c: fixed escaped multibyte char

* re.c (unescape_nonascii): escaped multibyte character should be
  copied as-is, just with checking if the encoding matches.
  https://twitter.com/sakuro/status/972014409986883584

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@62718 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

1 files changed, 12 insertions, 0 deletions

diff --git a/re.c b/re.c
index 0508e5f322..84c2e060dd 100644
--- a/re.c
+++ b/re.c
@@ -2537,11 +2537,13 @@ unescape_nonascii(const char *p, const char *end, rb_encoding *enc,
     while (p < end) {
         int chlen = rb_enc_precise_mbclen(p, end, enc);
         if (!MBCLEN_CHARFOUND_P(chlen)) {
+          invalid_multibyte:
             errcpy(err, "invalid multibyte character");
             return -1;
         }
         chlen = MBCLEN_CHARFOUND_LEN(chlen);
         if (1 < chlen || (*p & 0x80)) {
+          multibyte:
             rb_str_buf_cat(buf, p, chlen);
             p += chlen;
             if (*encp == 0)
@@ -2559,6 +2561,16 @@ unescape_nonascii(const char *p, const char *end, rb_encoding *enc,
                 errcpy(err, "too short escape sequence");
                 return -1;
             }
+            chlen = rb_enc_precise_mbclen(p, end, enc);
+            if (!MBCLEN_CHARFOUND_P(chlen)) {
+                goto invalid_multibyte;
+            }
+            if ((chlen = MBCLEN_CHARFOUND_LEN(chlen)) > 1) {
+		/* include the previous backslash */
+                --p;
+                ++chlen;
+                goto multibyte;
+            }
             switch (c = *p++) {
               case '1': case '2': case '3':
               case '4': case '5': case '6': case '7': /* \O, \OO, \OOO or backref */