From f79dd027e9f0ff05d949570b657e1e26b941d3fd Mon Sep 17 00:00:00 2001
From: Přemysl Janouch <p@janouch.name>
Date: Wed, 1 Aug 2018 09:06:14 +0200
Subject: kike: add a comment about identifier encoding

---
 kike.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/kike.c b/kike.c
index 7642a11..69384ed 100644
--- a/kike.c
+++ b/kike.c
@@ -233,6 +233,13 @@ irc_is_valid_host (const char *host)
 		|| irc_is_valid_hostaddr (host);
 }
 
+// TODO: currently, we are almost encoding-agnostic (strings just need to be
+// ASCII-compatible).  We should at least have an option to enforce a specific
+// encoding, such as UTF-8.  Note that with Unicode we should not allow all
+// character clasess and exclude the likes of \pM with the goal of enforcing
+// NFC-normalized identifiers--utf8proc is a good candidate library to handle
+// the categorization and validation.
+
 static bool
 irc_is_valid_user (const char *user)
 {
-- 
cgit v1.2.3-70-g09d2