core: do not print invalid utf-8 in error messages

src/shared/fileio.c

@@ -539,15 +539,18 @@ static int parse_env_file_push(const char *filename, unsigned line,
         va_list aq, *ap = userdata;
 
         if (!utf8_is_valid(key)) {
-                log_error("%s:%u: invalid UTF-8 for key '%s', ignoring.",
-                          filename, line, key);
+                _cleanup_free_ char *p = utf8_escape_invalid(key);
+
+                log_error("%s:%u: invalid UTF-8 in key '%s', ignoring.",
+                          filename, line, p);
                 return -EINVAL;
         }
 
         if (value && !utf8_is_valid(value)) {
-                /* FIXME: filter UTF-8 */
+                _cleanup_free_ char *p = utf8_escape_invalid(value);
+
                 log_error("%s:%u: invalid UTF-8 value for key %s: '%s', ignoring.",
-                          filename, line, key, value);
+                          filename, line, key, p);
                 return -EINVAL;
         }
 

src/shared/utf8.c

@@ -174,6 +174,32 @@ const char *utf8_is_valid(const char *str) {
         return str;
 }
 
+char *utf8_escape_invalid(const char *str) {
+        char *p, *s;
+
+        assert(str);
+
+        p = s = malloc(strlen(str) * 4 + 1);
+        if (!p)
+                return NULL;
+
+        while (*str) {
+                int len;
+
+                len = utf8_encoded_valid_unichar(str);
+                if (len > 0) {
+                        s = mempcpy(s, str, len);
+                        str += len;
+                } else {
+                        s = mempcpy(s, UTF8_REPLACEMENT_CHARACTER, strlen(UTF8_REPLACEMENT_CHARACTER));
+                        str += 1;
+                }
+        }
+        *s = '\0';
+
+        return p;
+}
+
 char *ascii_is_valid(const char *str) {
         const char *p;
 

src/shared/utf8.h

@@ -25,8 +25,11 @@
 
 #include "macro.h"
 
+#define UTF8_REPLACEMENT_CHARACTER "\xef\xbf\xbd"
+
 const char *utf8_is_valid(const char *s) _pure_;
 char *ascii_is_valid(const char *s) _pure_;
+char *utf8_escape_invalid(const char *s);
 
 bool utf8_is_printable(const char* str, size_t length) _pure_;