Improve `utf8len()` performance with UTF-8

7b260d9c · dearblue · a7be9f1c · 7b260d9c
Commit 7b260d9c authored Jul 14, 2019 by dearblue
Hide whitespace changes
Inline Side-by-side

Showing with 2 additions and 0 deletions

src/string.c src/string.c +2 -0

No files found.
--- a/src/string.c
+++ b/src/string.c
@@ -233,7 +233,9 @@ utf8len(const char* p, const char* e)
  mrb_int len;
  mrb_int i;

+  if ((unsigned char)*p < 0x80) return 1;
  len = utf8len_codepage[(unsigned char)*p];
+  if (len == 1) return 1;
  if (len > e - p) return 1;
  for (i = 1; i < len; ++i)
    if ((p[i] & 0xc0) != 0x80)