Create tts_preprocessor.py

tts_preprocessor.py

@@ -0,0 +1,87 @@
+"""
+TTS Text Preprocessing Utilities with Multilingual Support
+"""
+
+import re
+import json
+from typing import Dict, Set, Optional
+from num2words import num2words
+from pathlib import Path
+
+class TTSPreprocessor:
+    """Text preprocessor for TTS providers with multilingual support"""
+    
+    # Preprocessing flags
+    PREPROCESS_NUMBERS = "numbers"
+    PREPROCESS_CURRENCY = "currency"
+    PREPROCESS_TIME = "time"
+    PREPROCESS_DATE = "date"
+    PREPROCESS_CODES = "codes"
+    PREPROCESS_PERCENTAGE = "percentage"
+    
+    def __init__(self, language: str = "tr"):
+        self.language = language
+        self.locale_data = self._load_locale(language)
+        
+    def _load_locale(self, language: str) -> Dict:
+        """Load locale data from JSON file"""
+        locale_path = Path(__file__).parent / "locales" / f"{language}.json"
+        
+        # Fallback to English if locale not found
+        if not locale_path.exists():
+            print(f"⚠️ Locale file not found for {language}, falling back to English")
+            locale_path = Path(__file__).parent / "locales" / "en.json"
+            
+        try:
+            with open(locale_path, 'r', encoding='utf-8') as f:
+                return json.load(f)
+        except Exception as e:
+            print(f"❌ Error loading locale {language}: {e}")
+            # Return minimal default structure
+            return {
+                "language_code": language,
+                "currency": {"symbols": {}, "codes": {}},
+                "months": {},
+                "numbers": {
+                    "decimal_separator": ".",
+                    "thousands_separator": ",",
+                    "decimal_word": "point"
+                },
+                "small_number_threshold": 100
+            }
+    
+    def preprocess(self, text: str, flags: Set[str]) -> str:
+        """Apply preprocessing based on flags"""
+        
+        if self.PREPROCESS_CURRENCY in flags:
+            text = self._process_currency(text)
+            
+        if self.PREPROCESS_TIME in flags:
+            text = self._process_time(text)
+            
+        if self.PREPROCESS_DATE in flags:
+            text = self._process_date(text)
+            
+        if self.PREPROCESS_CODES in flags:
+            text = self._process_codes(text)
+            
+        if self.PREPROCESS_PERCENTAGE in flags:
+            text = self._process_percentage(text)
+            
+        # Numbers should be processed last to avoid conflicts
+        if self.PREPROCESS_NUMBERS in flags:
+            text = self._process_numbers(text)
+            
+        return text
+    
+    def _process_numbers(self, text: str) -> str:
+        """Convert numbers to words based on locale"""
+        decimal_sep = self.locale_data["numbers"]["decimal_separator"]
+        thousands_sep = self.locale_data["numbers"]["thousands_separator"]
+        decimal_word = self.locale_data["numbers"]["decimal_word"]
+        threshold = self.locale_data.get("small_number_threshold", 100)
+        
+        def replace_number(match):
+            num_str = match.group()
+            
+            # Normalize number format