rcx

commit 6d04c0b296ccd15dc0b14aba620df298e4e4c28a
parent cd697484ae8b6cd0a3fcc3f808543df606d7004c
Author: robert <robertrussell.72001@gmail.com>
Date:   Wed, 13 Jul 2022 16:59:43 -0700

Fix utf8decode

utf8decode should, as documented, return 0 iff all of the input string
is a prefix of a valid rune.

Diffstat:
M
src/utf8.c
 | 
7
++++---

1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/src/utf8.c b/src/utf8.c
@@ -61,16 +61,17 @@ utf8decode(rune *c, char *s, usize slen) {
 	}
 	if (len > UTF8_SIZE) /* Invalid leading byte? */
 		return 1;
-	if (len > slen) /* Not enough input? */
-		return 0;
 
 	/* Decode codepoint */
 	rune r = (uchar)s[0] & ~utf8mask[len-1];
-	for (usize i = 1; i < len; i++) {
+	usize l = MIN(len, slen);
+	for (usize i = 1; i < l; i++) {
 		if (((uchar)s[i] & 0xC0) != 0x80) /* Invalid continuation byte? */
 			return i;
 		r = (r << 6) | ((uchar)s[i] & 0x3F);
 	}
+	if (len > slen)
+		return 0; /* Looks good so far, but not enough input */
 
 	if (c && utf8encodable(r) && !utf8overlong(r, len))
 		*c = r;