Remove API dependency for lang detection using langdetect lib

backend/audio_to_tgt.py

@@ -2,6 +2,7 @@ import whisper
 import numpy as np
 from pydub import AudioSegment
 from deep_translator import GoogleTranslator, detection
+from langdetect import detect
 import os
 
 available_languages = GoogleTranslator().get_supported_languages(as_dict=True)
@@ -24,7 +25,8 @@ def src_audio_to_eng_translator(audio_file_input, model_size = "turbo", target_l
     result = model.transcribe(audio_data)
     input_text = result["text"]
 
-    src_lang_code = detection.single_detection(input_text, api_key = lang_detect_key)
+    # src_lang_code = detection.single_detection(input_text, api_key = lang_detect_key)
+    src_lang_code = detect(input_text)
     src_lang = formatted_codes.get(src_lang_code, 'Source language not detected')
     target_lang_code = formatted_languages.get(target_lang, 'en')
     translated_text = GoogleTranslator(source='auto', target=target_lang_code).translate(input_text)

backend/text_to_tgt.py

@@ -1,4 +1,5 @@
 from deep_translator import GoogleTranslator, detection
+from langdetect import detect
 import os
 
 available_languages = GoogleTranslator().get_supported_languages(as_dict=True)
@@ -9,7 +10,8 @@ lang_detect_key = os.getenv("detect_language_api_key")
 
 def src_txt_to_eng_translator(input_text, target_lang = 'English'):
     target_lang_code = formatted_languages.get(target_lang, 'en')
-    src_lang_code = detection.single_detection(input_text, api_key = lang_detect_key)
+    # src_lang_code = detection.single_detection(input_text, api_key = lang_detect_key)
+    src_lang_code = detect(input_text)
     src_lang = formatted_codes.get(src_lang_code, 'Source language not detected')
     translated = GoogleTranslator(source='auto', target=target_lang_code).translate(input_text)
     return translated, src_lang

requirements.txt

@@ -4,4 +4,5 @@ pillow==11.0.0
 pytesseract==0.3.13
 openai-whisper==20240930
 pydub==0.25.1
-langcodes==3.5.0
+langcodes==3.5.0
+langdetect==1.0.9