From 07b3782804272abce9ada300e96fa141e359b359 Mon Sep 17 00:00:00 2001
From: WormHeamer
Date: Fri, 2 Jan 2026 03:00:29 -0500
Subject: convert non-utf-8 files to a readonly hex dump

---
 utf8.h | 11 +++++++++++
 1 file changed, 11 insertions(+)

(limited to 'utf8.h')

diff --git a/utf8.h b/utf8.h
index 164279b..db98e18 100644
--- a/utf8.h
+++ b/utf8.h
@@ -15,6 +15,7 @@ u32 utf8_encode_len(const u32 *src, u32 cp_count);
 void utf8_decode(u32 *dst, const char *src, u32 cp_count);
 void utf8_encode(char *dst, const u32 *src, u32 cp_count);
 u32 utf8_decode_at(const char *s, u32 i, u32 n);
+int utf8_validate(const char *src, u32 n);
 
 #ifdef UTF8_IMPL
 
@@ -85,5 +86,15 @@ u32 utf8_decode_at(const char *s, u32 i, u32 n) {
 	return cp;
 }
 
+int utf8_validate(const char *src, u32 n) {
+	/* TODO: rewrite this to be faster */
+	for (u32 i = 0; i < n; i++) {
+		if (utf8_decode_at(src, i, n) == 0) {
+			return 0;
+		}
+	}
+	return 1;
+}
+
 #endif
 #endif
-- 
cgit v1.2.3