From d163da9ac89b2587401cd42e9a21309d1ba8fa3f Mon Sep 17 00:00:00 2001
From: Pierre Ossman <ossman@cendio.se>
Date: Mon, 5 Oct 2020 16:05:15 +0200
Subject: [PATCH] Fix conversion of latin-1 to UTF-8

Signed bug prevented anything not ASCII from being coded correctly.
---
 common/rfb/util.cxx    |  4 ++--
 tests/unit/unicode.cxx | 34 ++++++++++++++++++++++++++++++++++
 2 files changed, 36 insertions(+), 2 deletions(-)

diff --git a/common/rfb/util.cxx b/common/rfb/util.cxx
index 85b668fc..cc5b7690 100644
--- a/common/rfb/util.cxx
+++ b/common/rfb/util.cxx
@@ -381,7 +381,7 @@ namespace rfb {
     in_len = bytes;
     while ((in_len > 0) && (*in != '\0')) {
       char buf[5];
-      sz += ucs4ToUTF8(*in, buf);
+      sz += ucs4ToUTF8(*(const unsigned char*)in, buf);
       in++;
       in_len--;
     }
@@ -395,7 +395,7 @@ namespace rfb {
     in = src;
     in_len = bytes;
     while ((in_len > 0) && (*in != '\0')) {
-      out += ucs4ToUTF8(*in, out);
+      out += ucs4ToUTF8(*(const unsigned char*)in, out);
       in++;
       in_len--;
     }
diff --git a/tests/unit/unicode.cxx b/tests/unit/unicode.cxx
index 4bcb65b3..bb2525de 100644
--- a/tests/unit/unicode.cxx
+++ b/tests/unit/unicode.cxx
@@ -31,6 +31,11 @@ struct _ucs4utf16 {
     const wchar_t *utf16;
 };
 
+struct _latin1utf8 {
+    const char *latin1;
+    const char *utf8;
+};
+
 struct _utf8utf16 {
     const char *utf8;
     const wchar_t *utf16;
@@ -56,6 +61,13 @@ struct _ucs4utf16 ucs4utf16[] = {
     { 0x110200, L"\xfffd" },
 };
 
+struct _latin1utf8 latin1utf8[] = {
+    { "abc",            "abc" },
+    { "\xe5\xe4\xf6",   "\xc3\xa5\xc3\xa4\xc3\xb6" },
+    { "???",            "\xe2\x98\xb9\xe2\x98\xba\xe2\x98\xbb" },
+    { "?",              "\xe5\xe4" },
+};
+
 struct _utf8utf16 utf8utf16[] = {
     { "abc",                                                L"abc" },
     { "\xc3\xa5\xc3\xa4\xc3\xb6",                           L"\xe5\xe4\xf6" },
@@ -133,6 +145,28 @@ int main(int argc, char** argv)
         }
     }
 
+    for (i = 0;i < ARRAY_SIZE(latin1utf8);i++) {
+        /* Expected failure? */
+        if (strchr(latin1utf8[i].latin1, '?') != NULL)
+            continue;
+
+        out = rfb::latin1ToUTF8(latin1utf8[i].latin1);
+        if (strcmp(out, latin1utf8[i].utf8) != 0) {
+            printf("FAILED: latin1ToUTF8() #%d\n", (int)i+1);
+            failures++;
+        }
+        rfb::strFree(out);
+    }
+
+    for (i = 0;i < ARRAY_SIZE(latin1utf8);i++) {
+        out = rfb::utf8ToLatin1(latin1utf8[i].utf8);
+        if (strcmp(out, latin1utf8[i].latin1) != 0) {
+            printf("FAILED: utf8ToLatin1() #%d\n", (int)i+1);
+            failures++;
+        }
+        rfb::strFree(out);
+    }
+
     for (i = 0;i < ARRAY_SIZE(utf8utf16);i++) {
         /* Expected failure? */
         if (wcscmp(utf8utf16[i].utf16, L"\xfffd") == 0)
-- 
2.39.5