Unified language codes --> #9

tts_wrapper/engines/elevenlabs/client.py

@@ -4,7 +4,9 @@
 from ...tts import FileFormat
 from ...exceptions import UnsupportedFileFormat
 import requests
-
+from ...engines.utils import (
+    getISOLangCode,
+)
 audio_format = ("pcm_22050",)
 
 
@@ -199,11 +201,11 @@ def get_voices(self):
                 )  # Default to 'en-US'
                 if voice["high_quality_base_model_ids"] == "eleven_multilingual_v1":
                     voice["language_codes"] = [
-                        language_code for language_code in supported_languages_v1.keys()
+                        getISOLangCode(language_code) for language_code in supported_languages_v1.keys()
                     ]
                 else:
                     voice["language_codes"] = [
-                        language_code for language_code in supported_languages_v2.keys()
+                        getISOLangCode(language_code) for language_code in supported_languages_v2.keys()
                     ]
                 voice["name"] = voice["name"]
                 voice["gender"] = "Unknown"

tts_wrapper/engines/google/client.py

@@ -3,7 +3,9 @@
 import struct
 from google.cloud import texttospeech_v1beta1 as texttospeech
 from google.oauth2 import service_account
-
+from ...engines.utils import (
+    getISOLangCode,
+)
 
 class GoogleClient:
     def __init__(self, credentials: Union[str, Dict]) -> None:
@@ -137,7 +139,7 @@ def get_voices(self) -> List[Dict[str, Any]]:
             voice_data = {
                 "id": voice.name,
                 "name": voice.name,
-                "language_codes": voice.language_codes,
+                "language_codes": [getISOLangCode(item) for item in voice.language_codes],
                 "gender": voice.ssml_gender.name,  # 'MALE', 'FEMALE', 'NEUTRAL'
             }
             standardized_voices.append(voice_data)

tts_wrapper/engines/googletrans/client.py

@@ -4,6 +4,9 @@
 from io import BytesIO
 import logging
 import mp3
+from ...engines.utils import (
+    getISOLangCode,
+)
 
 try:
     from gtts import gTTS
@@ -133,7 +136,7 @@ def get_voices(self) -> List[Dict[str, Any]]:
                     standardized_voices.append(
                         {
                             "id": f"{lang_code}-{accent}",
-                            "language_codes": [lang_code],
+                            "language_codes": [getISOLangCode(lang_code)],
                             "name": f"{lang_name} ({accent})",
                             "gender": "Unknown",
                         }
@@ -142,7 +145,7 @@ def get_voices(self) -> List[Dict[str, Any]]:
                 standardized_voices.append(
                     {
                         "id": lang_code,
-                        "language_codes": [lang_code],
+                        "language_codes": [getISOLangCode(lang_code)],
                         "name": lang_name,
                         "gender": "Unknown",
                     }

tts_wrapper/engines/microsoft/client.py

@@ -1,9 +1,12 @@
+import logging
 from typing import Tuple, List, Dict, Any, Optional
 from tts_wrapper.tts import FileFormat
 import io
 import wave
 from ...exceptions import ModuleNotInstalled
-
+from ...engines.utils import (
+    getISOLangCode,
+)
 try:
     import requests
 except ImportError:
@@ -77,7 +80,7 @@ def get_available_voices(self) -> List[Dict[str, Any]]:
         for voice in voices:
             voice_dict = {
                 "id": voice["ShortName"],
-                "language_codes": [voice["Locale"]],
+                "language_codes": [getISOLangCode(voice["Locale"])],
                 "name": voice["LocalName"],
                 "gender": voice["Gender"],  # 'Gender' is already a string
             }

tts_wrapper/engines/polly/client.py

@@ -4,7 +4,9 @@
 from ...exceptions import ModuleNotInstalled
 import json
 import io
-
+from ...engines.utils import (
+    getISOLangCode,
+)
 Credentials = Tuple[str, str, str]
 
 FORMATS = {
@@ -98,7 +100,7 @@ def get_voices(self) -> List[Dict[str, Any]]:
         standardized_voices = []
         for voice in voices:
             voice["id"] = voice["Id"]
-            voice["language_codes"] = [voice["LanguageCode"]]
+            voice["language_codes"] = [getISOLangCode(voice["LanguageCode"])]
             voice["name"] = voice["Name"]
             voice["gender"] = voice["Gender"]
             standardized_voices.append(voice)

tts_wrapper/engines/sherpaonnx/client.py

@@ -4,6 +4,9 @@
 import json
 import logging
 import threading
+from ...engines.utils import (
+    getISOLangCode,
+)
 
 try:
     import numpy as np
@@ -268,7 +271,7 @@ def get_voices(self) -> List[Dict[str, str]]:
                 "id": voice["Iso Code"],
                 "name": voice["Language Name"],
                 "gender": "N",
-                "language_codes": [voice["Iso Code"]],
+                "language_codes": [getISOLangCode(voice["Iso Code"])],
             }
             for voice in self.voices_cache
         ]

tts_wrapper/engines/utils.py

@@ -6,6 +6,7 @@
 from io import BytesIO
 import re
 from typing import List, Dict, Tuple
+from langcodes import *
 
 
 def process_wav(raw: bytes) -> bytes:
@@ -22,19 +23,20 @@ def create_temp_filename(suffix="") -> str:
         tempfile.gettempdir(), f"{tempfile.gettempprefix()}_{random_seq}{suffix}"
     )
 
+
 def estimate_word_timings(text: str, wpm: int = 150) -> List[Tuple[float, float, str]]:
     # Remove SSML tags
     text = re.sub('<[^<]+?>', '', text)
-    
+
     # Split text into words, keeping punctuation
     words = re.findall(r'\b[\w\']+\b|[.,!?;]', text)
-    
+
     words_per_second = wpm / 60
     base_seconds_per_word = 1 / words_per_second
-    
+
     timings = []
     current_time = 0.0
-    
+
     for i, word in enumerate(words):
         # Adjust timing based on word length and type
         if len(word) <= 3:
@@ -43,7 +45,7 @@ def estimate_word_timings(text: str, wpm: int = 150) -> List[Tuple[float, float,
             duration = base_seconds_per_word * 1.2
         else:
             duration = base_seconds_per_word
-        
+
         # Adjust for punctuation
         if word in '.,!?;':
             duration = base_seconds_per_word * 0.5
@@ -52,14 +54,21 @@ def estimate_word_timings(text: str, wpm: int = 150) -> List[Tuple[float, float,
                 prev_start, prev_end, prev_word = timings[-1]
                 timings[-1] = (prev_start, prev_end + 0.2, prev_word)
                 current_time += 0.2
-        
+
         # Add natural variations
         variation = (hash(word) % 20 - 10) / 100  # -10% to +10% variation
         duration *= (1 + variation)
-        
+
         end_time = current_time + duration
         timings.append((current_time, end_time, word))
         current_time = end_time
-    
+
     return timings
 
+
+def getISOLangCode(dialect: str):
+    try:
+        return str(Language.get(dialect))
+    except Exception as dialectError:
+        dialect = dialect.split('-')[0]
+        return str(Language.get(dialect))

tts_wrapper/engines/uwp/client.py

@@ -1,7 +1,9 @@
 import asyncio
 from typing import List, Dict, Any, Optional
 from ...exceptions import ModuleNotInstalled
-
+from ...engines.utils import (
+    getISOLangCode,
+)
 try:
     from winrt.windows.media.speechsynthesis import SpeechSynthesizer
     from winrt.windows.storage.streams import DataReader
@@ -42,7 +44,7 @@ def get_voices(self) -> List[Dict[str, Any]]:
         for voice in voices:
             standardized_voice = {
                 "id": voice.id,
-                "language_codes": [voice.language],
+                "language_codes": [getISOLangCode(voice.language)],
                 "name": voice.display_name,
                 "gender": voice.gender.value,
             }

tts_wrapper/engines/watson/client.py

@@ -5,7 +5,9 @@
 import threading
 import json
 import logging
-
+from ...engines.utils import (
+    getISOLangCode,
+)
 from ...exceptions import ModuleNotInstalled
 
 Credentials = Tuple[str, str, str]  # api_key, region, instance_id
@@ -161,7 +163,7 @@ def get_voices(self) -> List[Dict[str, Any]]:
         for voice in voices:
             standardized_voice = {
                 "id": voice["name"],
-                "language_codes": [voice["language"]],
+                "language_codes": [getISOLangCode(voice["language"])],
                 "name": voice["name"].split("_")[1].replace("V3Voice", ""),
                 "gender": voice["gender"],
             }

tts_wrapper/engines/witai/client.py

@@ -3,7 +3,9 @@
 from ...exceptions import UnsupportedFileFormat
 import logging
 import requests
-
+from ...engines.utils import (
+    getISOLangCode,
+)
 FORMATS = {"mp3": "mp3", "pcm": "raw", "wav": "wav"}
 
 
@@ -48,7 +50,7 @@ def get_voices(self) -> List[Dict[str, Any]]:
                     standardized_voices.append(
                         {
                             "id": voice["name"],
-                            "language_codes": [locale],
+                            "language_codes": [getISOLangCode(locale)],
                             "name": voice["name"].split("$")[1],
                             "gender": voice["gender"],
                             "styles": voice.get("styles", []),