[SCC reader] Ensure a caption exists when receiving a BS or TOx code

#468
#467

Author: palemieux

src/main/python/ttconv/scc/context.py

@@ -103,10 +103,12 @@ def new_buffered_caption(self):
     """Resets buffered caption"""
     self.buffered_caption = SccCaptionParagraph(self.safe_area_x_offset, self.safe_area_y_offset, SccCaptionStyle.PopOn)
 
-  def get_caption_to_process(self) -> Optional[SccCaptionParagraph]:
+  def get_caption_to_process(self, time_code: SmpteTimeCode) -> Optional[SccCaptionParagraph]:
     """Returns the caption currently being processed"""
     if self.current_style in (SccCaptionStyle.PaintOn, SccCaptionStyle.RollUp):
       # If the Paint-On or Roll-Up style is activated, write directly on active caption
+      if not self.has_active_caption():
+        self.new_active_caption(time_code, self.current_style)
       return self.active_caption
     if self.current_style is SccCaptionStyle.PopOn:
       # For Pop-On style, write first on a buffered caption
@@ -159,11 +161,12 @@ def push_active_caption_to_model(self, time_code: SmpteTimeCode, clear_active_ca
       if not previous_caption.is_empty():
         self.div.push_child(previous_caption.to_paragraph(self.div.get_doc()))
 
-  def backspace(self):
+  def backspace(self, time_code: SmpteTimeCode):
     """Move the cursors in a column to the left"""
-    self.get_caption_to_process().get_current_text().backspace()
-    (row, indent) = self.get_caption_to_process().get_cursor()
-    self.get_caption_to_process().set_cursor_at(row, max(indent - 1, 0))
+    caption = self.get_caption_to_process(time_code)
+    caption.get_current_text().backspace()
+    (row, indent) = caption.get_cursor()
+    caption.set_cursor_at(row, max(indent - 1, 0))
 
   def paint_on_active_caption(self, time_code: SmpteTimeCode):
     """Initialize active caption for paint-on style"""
@@ -200,11 +203,10 @@ def process_preamble_address_code(self, pac: SccPreambleAddressCode, time_code:
 
       self.paint_on_active_caption(time_code)
 
-      if self.active_caption.get_caption_style() is SccCaptionStyle.PaintOn:
-        # Clear target row on Paint-On style
-        target_row = self.active_caption.get_lines().get(pac_row)
-        if target_row is not None:
-          target_row.clear()
+      # Clear target row on Paint-On style
+      target_row = self.active_caption.get_lines().get(pac_row)
+      if target_row is not None:
+        target_row.clear()
 
       self.active_caption.set_cursor_at(pac_row, pac_indent)
 
@@ -245,7 +247,7 @@ def process_preamble_address_code(self, pac: SccPreambleAddressCode, time_code:
   def process_mid_row_code(self, mid_row_code: SccMidRowCode, time_code: SmpteTimeCode):
     """Processes SCC Mid-Row Code to map it to the model"""
 
-    processed_caption = self.get_caption_to_process()
+    processed_caption = self.get_caption_to_process(time_code)
 
     color = mid_row_code.get_color()
     font_style = mid_row_code.get_font_style()
@@ -295,10 +297,10 @@ def process_mid_row_code(self, mid_row_code: SccMidRowCode, time_code: SmpteTime
         and processed_caption.get_caption_style() is SccCaptionStyle.PaintOn:
       processed_caption.get_current_text().set_begin(time_code)
 
-  def process_attribute_code(self, attribute_code: SccAttributeCode):
+  def process_attribute_code(self, attribute_code: SccAttributeCode, time_code: SmpteTimeCode):
     """Processes SCC Attribute Code to map it to the model"""
 
-    processed_caption = self.get_caption_to_process()
+    processed_caption = self.get_caption_to_process(time_code)
 
     if processed_caption is None:
       LOGGER.warning("No current SCC caption nor content initialized")
@@ -375,13 +377,13 @@ def process_control_code(self, control_code: SccControlCode, time_code: SmpteTim
       self.new_buffered_caption()
 
     elif control_code is SccControlCode.TO1:
-      self.get_caption_to_process().indent_cursor(1)
+      self.get_caption_to_process(time_code).indent_cursor(1)
 
     elif control_code is SccControlCode.TO2:
-      self.get_caption_to_process().indent_cursor(2)
+      self.get_caption_to_process(time_code).indent_cursor(2)
 
     elif control_code is SccControlCode.TO3:
-      self.get_caption_to_process().indent_cursor(3)
+      self.get_caption_to_process(time_code).indent_cursor(3)
 
     elif control_code is SccControlCode.CR:
       # Roll the displayed caption up one row (Roll-Up)
@@ -421,7 +423,7 @@ def process_control_code(self, control_code: SccControlCode, time_code: SmpteTim
       # Backspace
       # When a Backspace is received, the cursor moves to the left one column position erasing
       # the character or Mid-Row Code occupying that location, unless the cursor is in Column 1
-      self.backspace()
+      self.backspace(time_code)
 
   def process_text(self, word: str, time_code: SmpteTimeCode):
     """Processes SCC text words"""

src/main/python/ttconv/scc/line.py

@@ -150,7 +150,7 @@ def process(self, context: SccContext) -> SmpteTimeCode:
 
         elif isinstance(scc_code, SccAttributeCode):
           debug += scc_code.debug(scc_word.value)
-          context.process_attribute_code(scc_code)
+          context.process_attribute_code(scc_code, self.time_code)
           context.previous_word_type = type(scc_code)
 
         elif isinstance(scc_code, SccMidRowCode):
@@ -170,7 +170,7 @@ def process(self, context: SccContext) -> SmpteTimeCode:
           context.previous_word_type = type(scc_code)
 
         elif isinstance(scc_code, SccExtendedCharacter):
-          context.backspace()
+          context.backspace(self.time_code)
 
           word = scc_code.get_unicode_value()
           debug += word

src/test/python/test_scc_reader.py

@@ -1271,13 +1271,26 @@ def test_scc_with_negative_cursor(self):
     self.check_caption(p_list[0], "caption1", "00:00:01:16", "00:00:02:00", "Scary!  Eeech!")
     self.assertEqual(region_1, p_list[0].get_region())
 
+  def test_scc_content_starting_with_tab(self):
+    scc_content = """Scenarist_SCC V1.0
+00:42:41;20	942C 9429 97A2 5BF7
+"""
+    print(to_disassembly(scc_content))
+    self.assertIsNotNone(to_model(scc_content))
+
+  def test_scc_content_starting_with_bs(self):
+    scc_content = """Scenarist_SCC V1.0
+00:42:41;20	942C 9429 94A1 5BF7
+"""
+    self.assertIsNotNone(to_model(scc_content))
+
   def test_scc_content_starting_with_backspace(self):
     scc_content = """Scenarist_SCC V1.0
-10:01:44;17	94AE 9420 9470 9723 946E 80C1 92B0 20ec 6120 e6e9 6e20 64e5 7320 616e 6edc e573 2031 38b0 b02c 942C 8080 8080 942F
+10:01:44;17	94AE 9420 9470 9723 946E 94A1 92B0 20ec 6120 e6e9 6e20 64e5 7320 616e 6edc e573 2031 38b0 b02c 942C 8080 8080 942F
 """
 
     expected_scc_disassembly = """\
-10:01:44;17	{ENM}{RCL}{1500}{TO3}{15WhI}{??}À la fin des années 1800,{EDM}{}{}{EOC}
+10:01:44;17	{ENM}{RCL}{1500}{TO3}{15WhI}{BS}À la fin des années 1800,{EDM}{}{}{EOC}
 """
 
     scc_disassembly = to_disassembly(scc_content)