From 579829eef233ec927d9c71692fd880ddef66c2e7 Mon Sep 17 00:00:00 2001
From: nicm <nicm>
Date: Fri, 1 Sep 2023 18:43:54 +0000
Subject: [PATCH] Only compare the actual size of the UTF-8 character, not all
 of it.

---
 utf8-combined.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/utf8-combined.c b/utf8-combined.c
index e1b7934f..49f7a648 100644
--- a/utf8-combined.c
+++ b/utf8-combined.c
@@ -962,7 +962,7 @@ utf8_combined_first_cmp(struct utf8_combined_first *uf1,
 		return (-1);
 	if (ud1->size > ud2->size)
 		return (1);
-	return (memcmp(ud1->data, ud2->data, sizeof *ud1->data));
+	return (memcmp(ud1->data, ud2->data, ud1->size));
 }
 RB_HEAD(utf8_combined_tree, utf8_combined_first);
 RB_GENERATE_STATIC(utf8_combined_tree, utf8_combined_first, entry,
@@ -979,7 +979,7 @@ utf8_combined_second_cmp(const void *vp1, const void *vp2)
 		return (-1);
 	if (ud1->size > ud2->size)
 		return (1);
-	return (memcmp(ud1->data, ud2->data, sizeof *ud1->data));
+	return (memcmp(ud1->data, ud2->data, ud1->size));
 }
 
 static int
@@ -1035,7 +1035,7 @@ utf8_build_combined(void)
 		wc = utf8_combined_table[i].first;
 		mlen = wctomb(first.data, wc);
 		if (mlen <= 0 || mlen > UTF8_SIZE) {
-			log_debug("invalid combined character %08X", wc);
+			log_debug("invalid combined character %05X", wc);
 			continue;
 		}
 		first.size = mlen;
-- 
2.11.4.GIT