From aa355cd4fea2e38442a5c563e5d5a1878f61a4d1 Mon Sep 17 00:00:00 2001
From: Matthias Klumpp <matthias@tenstral.net>
Date: Fri, 8 Dec 2023 23:05:00 +0100
Subject: [PATCH] Sanitize UTF-8 before adding it to the database

---
 src/asgen/datastore.d | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/src/asgen/datastore.d b/src/asgen/datastore.d
index 19118ab..bef936f 100644
--- a/src/asgen/datastore.d
+++ b/src/asgen/datastore.d
@@ -192,9 +192,12 @@ public:
     private MDB_val makeDbValue (string data)
     {
         import core.stdc.string : strlen;
+        import std.encoding : sanitize, EncodingSchemeUtf8;
+
+        auto dataSane = sanitize(data);
+        auto cdata = dataSane.toStringz();
 
         MDB_val mval;
-        auto cdata = data.toStringz();
         mval.mv_size = char.sizeof * strlen(cdata) + 1;
         mval.mv_data = cast(void*) cdata;
         return mval;