[libsoup/websockets-fixes-2.66: 7/19] WebSockets: allow null characters in text messages data

[Claudio Saavedra <csaavedra src gnome org>]

commit 732f1d7f60b8708c167dc935b4bcb34e4389664f
Author: Carlos Garcia Campos <cgarcia igalia com>
Date:   Thu Jun 27 16:03:21 2019 +0200

    WebSockets: allow null characters in text messages data
    
    RFC 6455 says that text messages should contains valid UTF-8, and null
    characters valid according to RFC 3629. However, we are using
    g_utf8_validate(), which considers null characters as errors, to
    validate WebSockets text messages. This patch adds an internal
    utf8_validate() function based on g_utf8_validate() but allowing null
    characters and just returning a gboolean since we are always ignoring
    the end parameter in case of errors.
    soup_websocket_connection_send_text() assumes the given text is null
    terminated, so we need a new public function to allow sending text
    messages containing null characters. This patch adds
    soup_websocket_connection_send_message() that receives a
    SoupWebsocketDataType and GBytes, which is consistent with
    SoupWebsocketConnection::message signal.

 libsoup/soup-websocket-connection.c | 89 ++++++++++++++++++++++++++++++++++---
 1 file changed, 83 insertions(+), 6 deletions(-)
---
diff --git a/libsoup/soup-websocket-connection.c b/libsoup/soup-websocket-connection.c
index 66bd6871..963d539d 100644
--- a/libsoup/soup-websocket-connection.c
+++ b/libsoup/soup-websocket-connection.c
@@ -156,6 +156,82 @@ static void queue_frame (SoupWebsocketConnection *self, SoupWebsocketQueueFlags
 
 static void protocol_error_and_close (SoupWebsocketConnection *self);
 
+/* Code below is based on g_utf8_validate() implementation,
+ * but handling NULL characters as valid, as expected by
+ * WebSockets and compliant with RFC 3629.
+ */
+#define VALIDATE_BYTE(mask, expect)                             \
+        G_STMT_START {                                          \
+          if (G_UNLIKELY((*(guchar *)p & (mask)) != (expect)))  \
+                  return FALSE;                                 \
+        } G_STMT_END
+
+/* see IETF RFC 3629 Section 4 */
+static gboolean
+utf8_validate (const char *str,
+               gsize max_len)
+
+{
+        const gchar *p;
+
+        for (p = str; ((p - str) < max_len); p++) {
+                if (*(guchar *)p < 128)
+                        /* done */;
+                else {
+                        if (*(guchar *)p < 0xe0) { /* 110xxxxx */
+                                if (G_UNLIKELY (max_len - (p - str) < 2))
+                                        return FALSE;
+
+                                if (G_UNLIKELY (*(guchar *)p < 0xc2))
+                                        return FALSE;
+                        } else {
+                                if (*(guchar *)p < 0xf0) { /* 1110xxxx */
+                                        if (G_UNLIKELY (max_len - (p - str) < 3))
+                                                return FALSE;
+
+                                        switch (*(guchar *)p++ & 0x0f) {
+                                        case 0:
+                                                VALIDATE_BYTE(0xe0, 0xa0); /* 0xa0 ... 0xbf */
+                                                break;
+                                        case 0x0d:
+                                                VALIDATE_BYTE(0xe0, 0x80); /* 0x80 ... 0x9f */
+                                                break;
+                                        default:
+                                                VALIDATE_BYTE(0xc0, 0x80); /* 10xxxxxx */
+                                        }
+                                } else if (*(guchar *)p < 0xf5) { /* 11110xxx excluding out-of-range */
+                                        if (G_UNLIKELY (max_len - (p - str) < 4))
+                                                return FALSE;
+
+                                        switch (*(guchar *)p++ & 0x07) {
+                                        case 0:
+                                                VALIDATE_BYTE(0xc0, 0x80); /* 10xxxxxx */
+                                                if (G_UNLIKELY((*(guchar *)p & 0x30) == 0))
+                                                        return FALSE;
+                                                break;
+                                        case 4:
+                                                VALIDATE_BYTE(0xf0, 0x80); /* 0x80 ... 0x8f */
+                                                break;
+                                        default:
+                                                VALIDATE_BYTE(0xc0, 0x80); /* 10xxxxxx */
+                                        }
+                                        p++;
+                                        VALIDATE_BYTE(0xc0, 0x80); /* 10xxxxxx */
+                                } else {
+                                        return FALSE;
+                                }
+                        }
+
+                        p++;
+                        VALIDATE_BYTE(0xc0, 0x80); /* 10xxxxxx */
+                }
+        }
+
+        return TRUE;
+}
+
+#undef VALIDATE_BYTE
+
 static void
 frame_free (gpointer data)
 {
@@ -635,7 +711,7 @@ receive_close (SoupWebsocketConnection *self,
                data += 2;
                len -= 2;
                
-               if (!g_utf8_validate ((char *)data, len, NULL)) {
+               if (!utf8_validate ((const char *)data, len)) {
                        g_debug ("received non-UTF8 close data: %d '%.*s' %d", (int)len, (int)len, (char 
*)data, (int)data[0]);
                        protocol_error_and_close (self);
                        return;
@@ -780,9 +856,8 @@ process_contents (SoupWebsocketConnection *self,
                /* Actually deliver the message? */
                if (fin) {
                        if (pv->message_opcode == 0x01 &&
-                           !g_utf8_validate((char *)pv->message_data->data,
-                                            pv->message_data->len,
-                                            NULL)) {
+                           !utf8_validate((const char *)pv->message_data->data,
+                                          pv->message_data->len)) {
 
                                g_debug ("received invalid non-UTF8 text data");
 
@@ -1722,7 +1797,9 @@ soup_websocket_connection_get_close_data (SoupWebsocketConnection *self)
  * @self: the WebSocket
  * @text: the message contents
  *
- * Send a text (UTF-8) message to the peer.
+ * Send a %NULL-terminated text (UTF-8) message to the peer. If you need
+ * to send text messages containing %NULL characters use
+ * soup_websocket_connection_send_message() instead.
  *
  * The message is queued to be sent and will be sent when the main loop
  * is run.
@@ -1740,7 +1817,7 @@ soup_websocket_connection_send_text (SoupWebsocketConnection *self,
        g_return_if_fail (text != NULL);
 
        length = strlen (text);
-       g_return_if_fail (g_utf8_validate (text, length, NULL));
+        g_return_if_fail (utf8_validate (text, length));
 
        send_message (self, SOUP_WEBSOCKET_QUEUE_NORMAL, 0x01, (const guint8 *) text, length);
 }