Update app.py

app.py

@@ -8,10 +8,7 @@ import tarfile
 from pathlib import Path
 import soundfile as sf
 import sherpa_onnx
-from deep_translator import GoogleTranslator
 import numpy as np
-from iso639 import Lang
-import pycountry
 
 
 models = [
@@ -21,9 +18,9 @@ models = [
     ['vits-piper-fa-gyro-medium','https://github.com/k2-fsa/sherpa-onnx/releases/download/tts-models/vits-piper-fa_IR-gyro-medium.tar.bz2'],
     ['piper-fa-amir-medium','https://github.com/k2-fsa/sherpa-onnx/releases/download/tts-models/vits-piper-fa_IR-amir-medium.tar.bz2'],
     ['vits-mimic3-fa-haaniye_low','https://github.com/k2-fsa/sherpa-onnx/releases/download/tts-models/vits-mimic3-fa-haaniye_low.tar.bz2'],
-    ['',''],
+    # ['',''],
 ]
-model_info = models[model_id]
+dropdown_choices = list([i[0] for i in models])
 def download_and_extract_model(url, destination):
     """Download and extract the model files."""
     print(f"Downloading from URL: {url}")
@@ -176,6 +173,9 @@ def dl_espeak_data():
             print("  Subdirectories:", dirs)
         if files:
             print("  Files:", files)
+
+dl_espeak_data()
+
 def find_model_files(model_dir):
     """Find model files in the given directory and its subdirectories."""
     model_files = {}
@@ -212,15 +212,18 @@ def find_model_files(model_dir):
 def generate_audio(text, model_info):
     """Generate audio from text using the specified model."""
     try:
-        model_dir = os.path.join("./models", model_info['id'])
+        model_dir = os.path.join("./models", model_info)
         
         print(f"\nLooking for model in: {model_dir}")
         
         # Download model if it doesn't exist
         if not os.path.exists(model_dir):
-            print(f"Model directory doesn't exist, downloading {model_info['id']}...")
+            print(f"Model directory doesn't exist, downloading {model_info}...")
             os.makedirs(model_dir, exist_ok=True)
-            download_and_extract_model(model_info['url'], model_dir)
+            for i in models:
+                if model_info == i[0]:
+                    model_url=i[1]
+            download_and_extract_model(model_url, model_dir)
         
         print(f"Contents of {model_dir}:")
         for item in os.listdir(model_dir):
@@ -267,7 +270,7 @@ def generate_audio(text, model_info):
                 
             # Set data dir if it exists
             espeak_data = os.path.join(os.path.dirname(model_files['model']), 'espeak-ng-data')
-            data_dir = espeak_data if os.path.exists(espeak_data) else ''
+            data_dir = espeak_data if os.path.exists(espeak_data) else 'espeak-ng-data'
             
             # Get lexicon path if it exists
             lexicon = model_files.get('lexicon', '') if os.path.exists(model_files.get('lexicon', '')) else ''
@@ -328,28 +331,22 @@ def tts_interface(selected_model, text, status_output):
         if not text.strip():
             return None, "Please enter some text"
             
-        # Get model ID from the display name mapping
-        model_id = models_by_display.get(selected_model)
-        if not model_id or model_id not in models:
-            return None, "Please select a model"
-            
-        
+
+        model_id = selected_model
         # Store original text for status message
         original_text = text
         
 
         try:
             # Update status with language info
-            lang_info = model_info.get('language', [{}])[0]
-            lang_name = lang_info.get('language_name', 'Unknown')
-            voice_name = model_info.get('name', model_id)
-            status = f"Generating speech using {voice_name} ({lang_name})..."
+            voice_name = model_id
+            status = f"Generating speech using {voice_name} ..."
             
             # Generate audio
-            audio_data, sample_rate = generate_audio(text, model_info)
+            audio_data, sample_rate = generate_audio(text, model_id)
             
             # Include translation info in final status if text was actually translated
-            final_status = f"Generated speech using {voice_name} ({lang_name})"
+            final_status = f"Generated speech using {voice_name}"
             final_status += f"\nText: '{text}'"
             
             return (sample_rate, audio_data), final_status