VTT reader: fix empty line detection

src/main/python/ttconv/vtt/reader.py

@@ -451,6 +451,7 @@ class _State(Enum):
 
   state = _State.START
   current_p = None
+  subtitle_text = None
 
   for line_index, line in enumerate(_none_terminated(lines)):
 
@@ -524,11 +525,17 @@ class _State(Enum):
     if state in (_State.TEXT, _State.TEXT_MORE):
 
       if line is None or _EMPTY_RE.fullmatch(line):
-        subtitle_text = subtitle_text.strip('\r\n').replace(r"\n\r", "\n")
-
-        _parse_cue_text(subtitle_text, current_p, line_index)
+        if subtitle_text is not None:
+          _parse_cue_text(
+            subtitle_text.strip('\r\n').replace(r"\n\r", "\n"),
+            current_p,
+            line_index
+          )
+        else:
+          LOGGER.warning("Ignoring cue due to a spurious blank line at line %s", line_index)
 
         state = _State.LOOKING
+        subtitle_text = None
         continue
 
       if state is _State.TEXT:

src/test/python/test_vtt_reader.py

@@ -97,6 +97,28 @@ def test_blank_lines(self):
     f = io.StringIO(SAMPLE)
     self.assertIsNotNone(to_model(f))
 
+  def test_malformed_blank_lines(self):
+    # from https://github.com/sandflow/ttconv/issues/439
+    # the first cue should be ignored since it is malformed
+    SAMPLE = """WEBVTT
+Kind: captions
+Language: en
+
+00:00:00.799 --> 00:00:02.869 align:start position:0%
+
+hi<00:00:01.040><c> everyone</c><00:00:01.920><c> today</c><00:00:02.240><c> we're</c><00:00:02.399><c> going</c><00:00:02.639><c> to</c><00:00:02.720><c> be</c>
+
+00:00:02.869 --> 00:00:02.879 align:start position:0%
+hi everyone today we're going to be
+"""
+
+    doc = to_model(io.StringIO(SAMPLE))
+    self.assertIsNotNone(doc)
+    body = list(doc.get_body())
+    self.assertEqual(len(body), 1)
+    div = list(body[0])
+    self.assertEqual(len(div), 1)
+
   def test_italic(self):
     f = io.StringIO(r"""WEBVTT