app.py

from fastapi import FastAPI, HTTPException
from fastapi import UploadFile, File
from gtts import gTTS
from io import BytesIO
from fastapi.responses import StreamingResponse
from pydantic import BaseModel
from transformers import MarianMTModel, MarianTokenizer
import speech_recognition as sr
# import pyttsx3
import sounddevice as sd
from scipy.io.wavfile import write
import tempfile
import os
# import uvicorn

# Initialize FastAPI app
app = FastAPI()

# CORS configuration
# app.add_middleware(
#     CORSMiddleware,
#     allow_origins=["*"],  # Allow all origins for development; adjust for production
#     allow_credentials=True,
#     allow_methods=["*"],  # Allow all HTTP methods
#     allow_headers=["*"],  # Allow all headers
# )

# Initialize TTS engine for speaking the translated text
# engine = pyttsx3.init()

# Supported languages dictionary
supported_languages = {
    "en": "English", "fr": "French", "es": "Spanish", "de": "German",
    "it": "Italian", "ru": "Russian", "zh": "Chinese", "ar": "Arabic",
    "hi": "Hindi", "ja": "Japanese", "ko": "Korean", "pt": "Portuguese",
    "nl": "Dutch", "sv": "Swedish", "pl": "Polish", "tr": "Turkish",
    "vi": "Vietnamese", "th": "Thai", "he": "Hebrew", "id": "Indonesian"
}

# Model for input data validation
class TranslationRequest(BaseModel):
    src_lang: str
    tgt_lang: str
    text: str

class SpeakRequest(BaseModel):
    text: str

def load_model(src_lang, tgt_lang):
    """Load the appropriate translation model."""
    model_name = f'Helsinki-NLP/opus-mt-{src_lang}-{tgt_lang}'
    try:
        tokenizer = MarianTokenizer.from_pretrained(model_name)
        model = MarianMTModel.from_pretrained(model_name)
        return model, tokenizer
    except Exception as e:
        print(f"Model loading error: {e}")  # Log error
        raise

def translate_text(text, model, tokenizer):
    """Translate input text."""
    inputs = tokenizer(text, return_tensors="pt", padding=True)
    translated = model.generate(**inputs)
    translated_text = tokenizer.decode(translated[0], skip_special_tokens=True)
    return translated_text

@app.post("/translate")
async def translate(request: TranslationRequest):
    """Translate text from source to target language."""
    if request.src_lang not in supported_languages or request.tgt_lang not in supported_languages:
        raise HTTPException(status_code=400, detail="Unsupported language.")
    
    try:
        print(f"Translating text: {request.text} from {request.src_lang} to {request.tgt_lang}")  # Debug info
        model, tokenizer = load_model(request.src_lang, request.tgt_lang)
        translated_text = translate_text(request.text, model, tokenizer)
        print(f"Translated text: {translated_text}")  # Debug info
        return {"translated_text": translated_text}
    except Exception as e:
        print(f"Error: {e}")  # Log error
        raise HTTPException(status_code=500, detail=str(e))

@app.post("/recognize")
async def recognize(language: str = 'en', file: UploadFile = File(...)):
    """Recognize speech from uploaded WAV file"""
    try:
        recognizer = sr.Recognizer()
        with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio:
            temp_audio.write(await file.read())
            temp_audio_path = temp_audio.name

        with sr.AudioFile(temp_audio_path) as source:
            audio = recognizer.record(source)

        recognized_text = recognizer.recognize_google(audio, language=language)

        os.remove(temp_audio_path)

        return {"recognized_text": recognized_text}
    except sr.UnknownValueError:
        raise HTTPException(status_code=400, detail="Could not understand the audio")
    except sr.RequestError as e:
        raise HTTPException(status_code=500, detail=f"Recognition error: {e}")
    except Exception as e:
        print(f"Error: {e}")
        raise HTTPException(status_code=500, detail=str(e))
# @app.post("/recognize")
# async def recognize(language: str = 'en'):
#     """Capture speech and recognize it using sounddevice."""
#     duration = 10  # Duration of recording in seconds
#     samplerate = 16000  # Sample rate for recording

#     try:
#         print("Recording audio...")  # Debug info
#         # Record audio using sounddevice
#         recording = sd.rec(int(duration * samplerate), samplerate=samplerate, channels=1, dtype='int16')
#         sd.wait()  # Wait until recording is finished

#         # Save audio to a temporary file
#         with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio_file:
#             write(temp_audio_file.name, samplerate, recording)
#             temp_audio_path = temp_audio_file.name

#         # Recognize speech using speech_recognition
#         recognizer = sr.Recognizer()
#         with sr.AudioFile(temp_audio_path) as source:
#             audio = recognizer.record(source)

#         recognized_text = recognizer.recognize_google(audio, language=language)

#         # Clean up temporary audio file
#         os.remove(temp_audio_path)

#         return {"recognized_text": recognized_text}
#     except sr.UnknownValueError:
#         raise HTTPException(status_code=400, detail="Could not understand the audio")
#     except sr.RequestError as e:
#         raise HTTPException(status_code=500, detail=f"Recognition error: {e}")
#     except Exception as e:
#         print(f"Error: {e}")  # Log error
#         raise HTTPException(status_code=500, detail=str(e))


@app.get("/speak")
def speak(text: str):
    tts = gTTS(text)
    mp3_fp = BytesIO()
    tts.write_to_fp(mp3_fp)
    mp3_fp.seek(0)
    return StreamingResponse(mp3_fp, media_type="audio/mpeg")
    
# @app.post("/speak")
# async def speak(request: SpeakRequest):
#     """Speak out a given text."""
#     if not request.text:
#         raise HTTPException(status_code=400, detail="No text provided.")
#     try:
#         engine.say(request.text)
#         engine.runAndWait()
#         return {"message": "Text spoken successfully"}
#     except Exception as e:
#         raise HTTPException(status_code=500, detail=str(e))

# if __name__ == "__main__":
#     uvicorn.run("app:app", host="0.0.0.0", port=8000, reload=True)