Update app.py

app.py

@@ -73,17 +73,23 @@ def create_sentence_base_filter(full_text, duration_clip, font_option, font_size
     escaped_text = full_text.replace(':', FFMPEG_ESCAPE_CHAR + ':')
     
     # Filter für den gesamten Satz, sichtbar für die gesamte Clip-Dauer
-    return (
+    drawtext_filter = (
         f"drawtext=text='{escaped_text}':"
         f"fontcolor={base_params['fontcolor']}:"
         f"fontsize={base_params['fontsize']}:"
         f"borderw={base_params['borderw']}:"
         f"bordercolor={base_params['bordercolor']}:"
         + (f"box={base_params['box']}:boxcolor={base_params['boxcolor']}:boxborderw=10:" if base_params["box"] else "") +
-        f"x=(w-text_w)/2:y=(h-text_h)*{y_pos}:{font_option}:"
-        f"enable='between(t, 0, {duration_clip})'"
+        f"x=(w-text_w)/2:y=(h-text_h)*{y_pos}"
     )
 
+    # Korrigierte Logik, um doppelte Doppelpunkte zu vermeiden
+    if font_option:
+        drawtext_filter += f":{font_option}"
+    
+    drawtext_filter += f":enable='between(t, 0, {duration_clip})'"
+    return drawtext_filter
+
 
 def create_highlight_word_filter(word, full_text, start_time, duration, font_option, font_size, y_pos, style):
     """
@@ -149,17 +155,23 @@ def create_highlight_word_filter(word, full_text, start_time, duration, font_opt
     escaped_word = word.replace(':', FFMPEG_ESCAPE_CHAR + ':')
 
     # Filter für das einzelne, hervorgehobene Wort
-    return (
+    drawtext_filter = (
         f"drawtext=text='{escaped_word}':"
         f"fontcolor={params['fontcolor']}:"
         f"fontsize={params['fontsize_override']}:"
         f"borderw={params['borderw']}:"
         f"bordercolor={params['bordercolor']}:"
         + (f"box={params['box']}:boxcolor={params['boxcolor']}:boxborderw=10:" if params["box"] else "") +
-        f"x=(w-text_w)/2:y=(h-text_h)*{y_pos}:{font_option}:"
-        f"alpha='{alpha_expression}'"
+        f"x=(w-text_w)/2:y=(h-text_h)*{y_pos}"
     )
 
+    # Korrigierte Logik, um doppelte Doppelpunkte zu vermeiden
+    if font_option:
+        drawtext_filter += f":{font_option}"
+    
+    drawtext_filter += f":alpha='{alpha_expression}'"
+    return drawtext_filter
+
 
 def generate_slideshow_with_audio(images, input_text, duration_per_word, duration_per_image, fade_duration, font_size, y_pos, audio_file, subtitle_style):
     
@@ -182,7 +194,8 @@ def generate_slideshow_with_audio(images, input_text, duration_per_word, duratio
 
     # Schriftart finden
     font_path = get_font_path()
-    font_option = f":fontfile='{font_path}'" if font_path else ""
+    # font_option enthält jetzt NUR den Parameter-Teil, OHNE führenden Doppelpunkt.
+    font_option = f"fontfile='{font_path}'" if font_path else "" 
 
     # Audio verarbeiten
     audio_temp_dir, temp_audio_file = save_temp_audio(audio_file) if audio_file else (None, None)
@@ -247,6 +260,7 @@ def generate_slideshow_with_audio(images, input_text, duration_per_word, duratio
         ]
         
         try:
+            # Hinzufügen von stdout/stderr Erfassung für bessere Fehlerprotokollierung
             subprocess.run(cmd, check=True, capture_output=True, text=True)
             clips_with_text.append(clip_path)
         except subprocess.CalledProcessError as e:
@@ -348,4 +362,4 @@ with gr.Blocks() as demo:
         outputs=[out_video, status]
     )
 
-demo.launch()
+demo.launch(