typefindfunctions: Fix WebVTT format detection

If WebVTT file consists of "WebVTT" header without body,
the file size can be smaller than 10 bytes.

Part-of: <https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/1359>

diff --git a/subprojects/gst-plugins-base/gst/typefind/gsttypefindfunctions.c b/subprojects/gst-plugins-base/gst/typefind/gsttypefindfunctions.c
index 6731bea..5a4df31 100644 (file)
--- a/subprojects/gst-plugins-base/gst/typefind/gsttypefindfunctions.c
+++ b/subprojects/gst-plugins-base/gst/typefind/gsttypefindfunctions.c
@@ -5018,26 +5018,33 @@ static void
 webvtt_type_find (GstTypeFind * tf, gpointer private)
 {
   const guint8 *data;
+  static const guint8 webvtt_with_bom[] = {
+    0xef, 0xbb, 0xbf, 'W', 'E', 'B', 'V', 'T', 'T'
+  };
 
-  data = gst_type_find_peek (tf, 0, 10);
+  data = gst_type_find_peek (tf, 0, 7);
 
   if (data == NULL)
     return;
 
-  /* there might be a UTF-8 BOM at the beginning */
-  if (memcmp (data, "WEBVTT", 6) != 0 && memcmp (data + 3, "WEBVTT", 6) != 0) {
-    return;
-  }
+  if (memcmp (data, "WEBVTT", 6) == 0) {
+    data += 6;
+  } else {
+    data = gst_type_find_peek (tf, 0, 10);
+
+    if (!data)
+      return;
+
+    /* there might be a UTF-8 BOM at the beginning */
+    if (memcmp (data, webvtt_with_bom, sizeof (webvtt_with_bom)) != 0)
+      return;
 
-  if (data[0] != 'W') {
-    if (data[0] != 0xef || data[1] != 0xbb || data[2] != 0xbf)
-      return;                   /* Not a UTF-8 BOM */
-    data += 3;
+    data += 9;
   }
 
   /* After the WEBVTT magic must be one of these chars:
    *   0x20 (space), 0x9 (tab), 0xa (LF) or 0xd (CR) */
-  if (data[6] != 0x20 && data[6] != 0x9 && data[6] != 0xa && data[6] != 0xd) {
+  if (data[0] != 0x20 && data[0] != 0x9 && data[0] != 0xa && data[0] != 0xd) {
     return;
   }
 

diff --git a/subprojects/gst-plugins-base/tests/check/gst/typefindfunctions.c b/subprojects/gst-plugins-base/tests/check/gst/typefindfunctions.c
index 7f4df8c..0bdef0c 100644 (file)
--- a/subprojects/gst-plugins-base/tests/check/gst/typefindfunctions.c
+++ b/subprojects/gst-plugins-base/tests/check/gst/typefindfunctions.c
@@ -477,6 +477,37 @@ GST_START_TEST (test_manifest_typefinding)
 
 GST_END_TEST;
 
+GST_START_TEST (test_webvtt)
+{
+  GstTypeFindProbability prob;
+  const gchar *media_type;
+  GstCaps *caps;
+  guint8 webvtt[] = {
+    'W', 'E', 'B', 'V', 'T', 'T', '\n', '\n'
+  };
+  guint8 webvtt_with_bom[] = {
+    0xef, 0xbb, 0xbf, 'W', 'E', 'B', 'V', 'T', 'T', '\n', '\n'
+  };
+
+  prob = 0;
+  caps = typefind_data (webvtt, sizeof (webvtt), &prob);
+  fail_unless (caps != NULL);
+  media_type = gst_structure_get_name (gst_caps_get_structure (caps, 0));
+  fail_unless_equals_string (media_type, "application/x-subtitle-vtt");
+  fail_unless_equals_int (prob, GST_TYPE_FIND_MAXIMUM);
+  gst_caps_unref (caps);
+
+  prob = 0;
+  caps = typefind_data (webvtt_with_bom, sizeof (webvtt_with_bom), &prob);
+  fail_unless (caps != NULL);
+  media_type = gst_structure_get_name (gst_caps_get_structure (caps, 0));
+  fail_unless_equals_string (media_type, "application/x-subtitle-vtt");
+  fail_unless_equals_int (prob, GST_TYPE_FIND_MAXIMUM);
+  gst_caps_unref (caps);
+}
+
+GST_END_TEST;
+
 static Suite *
 typefindfunctions_suite (void)
 {
@@ -494,6 +525,7 @@ typefindfunctions_suite (void)
   tcase_add_test (tc_chain, test_random_data);
   tcase_add_test (tc_chain, test_hls_m3u8);
   tcase_add_test (tc_chain, test_manifest_typefinding);
+  tcase_add_test (tc_chain, test_webvtt);
 
   return s;
 }