From d4e93269f280a57863f55f30a910afabaf424c24 Mon Sep 17 00:00:00 2001
From: Martin Lambers <marlam@marlam.de>
Date: Sat, 28 Sep 2024 13:13:37 +0200
Subject: [PATCH] Properly encode non-ASCII strings in the From header

This fixes Github issue #162.
---
 src/msmtp.c |  4 +++-
 src/tools.c | 56 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 src/tools.h |  8 ++++++++
 3 files changed, 67 insertions(+), 1 deletion(-)

diff --git a/src/msmtp.c b/src/msmtp.c
index 95297a6..75a4a29 100644
--- a/src/msmtp.c
+++ b/src/msmtp.c
@@ -4187,8 +4187,10 @@ int main(int argc, char *argv[])
         {
             if (account->from_full_name)
             {
+                char *enc_name = encode_for_header(account->from_full_name);
                 fprintf(prepend_header_tmpfile, "From: %s <%s>\n",
-                        account->from_full_name, account->from);
+                        enc_name, account->from);
+                free(enc_name);
             }
             else
             {
diff --git a/src/tools.c b/src/tools.c
index d029f24..11718d3 100644
--- a/src/tools.c
+++ b/src/tools.c
@@ -50,9 +50,11 @@
 #include <unistd.h>
 #ifdef ENABLE_NLS
 # include <locale.h>
+# include <langinfo.h>
 #endif
 
 #include "xalloc.h"
+#include "base64.h"
 #include "tools.h"
 
 
@@ -907,3 +909,57 @@ int check_hostname_matches_domain(const char *hostname, const char *domain)
     return (hostname[hostname_len - 1 - domain_len] == '.'
             && strcasecmp(hostname + (hostname_len - domain_len), domain) == 0) ? 1 : 0;
 }
+
+
+/*
+ * encode_for_header()
+ *
+ * see tools.h
+ */
+
+char *encode_for_header(const char *s)
+{
+    int needsEncoding = 0;
+    for (int i = 0; s[i]; i++)
+    {
+        if (s[i] < 32 || s[i] >= 127)
+        {
+            needsEncoding = 1;
+            break;
+        }
+    }
+    if (needsEncoding)
+    {
+        /* create a string of the form "=?ENCODING?B?BASE64STRING?=" */
+        size_t s_len = strlen(s);
+        size_t b64_s_len = BASE64_LENGTH(s_len);
+        char* encoding = xstrdup(
+#ifdef ENABLE_NLS
+                nl_langinfo(CODESET)
+#else
+                "UTF-8";
+#endif
+                );
+        size_t e_len = strlen(encoding);
+        size_t enc_len = 2 + e_len + 3 + b64_s_len + 3;
+        char *enc = xmalloc(enc_len + 1);
+        size_t i = 0;
+        enc[i++] = '=';
+        enc[i++] = '?';
+        for (size_t j = 0; j < e_len; j++)
+        {
+            enc[i++] = tolower(encoding[j]);
+        }
+        free(encoding);
+        enc[i++] = '?';
+        enc[i++] = 'B';
+        enc[i++] = '?';
+        base64_encode(s, s_len, enc + i, enc_len - i + 1);
+        strcat(enc, "?=");
+        return enc;
+    }
+    else
+    {
+        return xstrdup(s);
+    }
+}
diff --git a/src/tools.h b/src/tools.h
index 86ba9ff..ebdeb73 100644
--- a/src/tools.h
+++ b/src/tools.h
@@ -259,4 +259,12 @@ void split_mail_address(const char *address, char **local_part, char **domain_pa
  */
 int check_hostname_matches_domain(const char *hostname, const char *domain);
 
+/*
+ * encode_for_header()
+ *
+ * Encode a string so that it can be used in a mail header.
+ * The result is allocated and needs to be freed.
+ */
+char *encode_for_header(const char *s);
+
 #endif