Update app.py

app.py

@@ -17,22 +17,9 @@ import re
 
 # Download NLTK data during initialization
 try:
+    nltk.data.find('tokenizers/punkt')
+except LookupError:
     nltk.download('punkt', quiet=True)
-except Exception as e:
-    print(f"Warning: Failed to download NLTK data: {str(e)}")
-    print("Downloading from alternative source...")
-    try:
-        import ssl
-        try:
-            _create_unverified_https_context = ssl._create_unverified_context
-        except AttributeError:
-            pass
-        else:
-            ssl._create_default_https_context = _create_unverified_https_context
-        nltk.download('punkt', quiet=True)
-    except Exception as e:
-        print(f"Critical error downloading NLTK data: {str(e)}")
-        raise
 
 # Configure logging
 logging.basicConfig(
@@ -160,7 +147,7 @@ class TTSDatasetCollector:
 
             try:
                 # Try NLTK first
-                self.sentences = nltk.sent_tokenize(text.strip(), language='english')
+                self.sentences = nltk.sent_tokenize(text.strip())
             except Exception as e:
                 logger.warning(f"NLTK tokenization failed, falling back to simple splitting: {str(e)}")
                 # Fallback to simple splitting
@@ -207,7 +194,7 @@ class TTSDatasetCollector:
 
     def get_styled_text(self, text: str) -> str:
         """Get text with current font styling"""
-        font_css = FONT_STYLES[self.current_font]['css']
+        font_css = FONT_STYLES.get(self.current_font, {}).get('css', '')
         return f"<div style='{font_css}'>{text}</div>"
 
     def set_font(self, font_style: str) -> Tuple[bool, str]:
@@ -532,18 +519,28 @@ def create_interface():
     # Add font-face declarations
     font_face_css = ""
     for font_style, font_info in FONT_STYLES.items():
-        if font_style in ['nastaliq', 'naskh'] or font_style in collector.custom_fonts:
-            font_file_name = font_info['family'] + '.ttf' if font_style not in collector.custom_fonts else font_info['family'] + '.ttf'
+        font_file_name = font_info['family'] + '.ttf'
+        font_path = collector.fonts_path / font_file_name
+        if os.path.exists(font_path):
             font_face_css += f"""
             @font-face {{
                 font-family: '{font_info["family"]}';
-                src: url('fonts/{font_file_name}') format('truetype');
+                src: url('file/{font_path}') format('truetype');
             }}
             """
+        else:
+            # For system fonts like 'Arial' and 'Times New Roman', no need to specify src
+            pass
 
     custom_css += font_face_css
 
     with gr.Blocks(title="TTS Dataset Collection Tool", css=custom_css) as interface:
+        status = gr.Textbox(
+            label="Status",
+            interactive=False,
+            max_lines=3
+        )
+
         gr.Markdown("# TTS Dataset Collection Tool")
 
         with gr.Row():
@@ -571,7 +568,7 @@ def create_interface():
                     elem_classes=["small-input"]
                 )
                 font_select = gr.Dropdown(
-                    choices=list(FONT_STYLES.keys()),
+                    choices=list(FONT_STYLES.keys()) + list(collector.custom_fonts.keys()),
                     value="english_serif",
                     label="Select Font Style",
                     elem_classes=["small-input"]
@@ -584,12 +581,6 @@ def create_interface():
                 )
                 add_font_btn = gr.Button("Add Custom Font")
 
-                status = gr.Textbox(
-                    label="Status",
-                    interactive=False,
-                    max_lines=3
-                )
-
                 # Dataset Info
                 with gr.Accordion("Dataset Statistics", open=False):
                     dataset_info = gr.JSON(
@@ -757,8 +748,23 @@ def create_interface():
             if not success:
                 return {status: f"❌ {msg}"}
             # Update font dropdown
-            font_choices = list(FONT_STYLES.keys())
+            font_choices = list(FONT_STYLES.keys()) + list(collector.custom_fonts.keys())
             font_select.update(choices=font_choices)
+            # Rebuild CSS to include new font
+            font_face_css = ""
+            for font_style, font_info in FONT_STYLES.items():
+                if font_style in collector.custom_fonts:
+                    font_file_name = font_info['family'] + '.ttf'
+                    font_path = collector.fonts_path / font_file_name
+                    if os.path.exists(font_path):
+                        font_face_css += f"""
+                        @font-face {{
+                            font-family: '{font_info["family"]}';
+                            src: url('file/{font_path}') format('truetype');
+                        }}
+                        """
+            # Update the interface CSS
+            interface.set_css(custom_css + font_face_css)
             return {status: f"✅ {msg}"}
 
         # Event handlers