app.py

import os
import gradio as gr
import tempfile
import time
from groq import Groq
from gtts import gTTS
from deep_translator import GoogleTranslator
from langdetect import detect

# -----------------------------
# STEP 1: API Key setup (Hugging Face Safe)
# -----------------------------
# Make sure to set GROQ_API_KEY in your Space:
# Go to: Settings → Variables and secrets → Repository secrets → Add new secret
# Name: GROQ_API_KEY
# Value: your key (like gsk_xxxxxxx)

api_key = os.getenv("GROQ_API_KEY")
if not api_key:
    raise ValueError(
        "❌ GROQ_API_KEY not found! Please add it in your Hugging Face Space → Settings → Secrets."
    )

client = Groq(api_key=api_key)


# -----------------------------
# STEP 2: Whisper Transcription
# -----------------------------
def transcribe_audio(audio_path):
    """Transcribe audio using Groq Whisper model."""
    try:
        with open(audio_path, "rb") as f:
            transcription = client.audio.transcriptions.create(
                model="whisper-large-v3",
                file=f
            )
        return transcription.text.strip()
    except Exception as e:
        return f"[Transcription error] {e}"


# -----------------------------
# STEP 3: Language Detection
# -----------------------------
def detect_language(text):
    try:
        return detect(text)
    except:
        return "unknown"


# -----------------------------
# STEP 4: Translation
# -----------------------------
def translate_text(text, dest_lang):
    if not text.strip():
        return "[Translation error] Empty input text"
    try:
        translator = GoogleTranslator(source="auto", target=dest_lang)
        translated = translator.translate(text)
        if translated.strip().lower() == text.strip().lower():
            time.sleep(1)
            translated = translator.translate(text)
        return translated
    except Exception as e:
        return f"[Translation error] {e}"


# -----------------------------
# STEP 5: Text-to-Speech
# -----------------------------
def text_to_speech(text, lang):
    try:
        tts = gTTS(text=text, lang=lang)
        temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
        tts.save(temp_file.name)
        return temp_file.name
    except Exception as e:
        return f"[TTS error] {e}"


# -----------------------------
# STEP 6: Full Pipeline
# -----------------------------
def pipeline_translate(audio, target_lang_hint):
    if not audio:
        return "[Error] No audio input", None, None, None

    transcribed = transcribe_audio(audio)
    if transcribed.startswith("[Transcription error]"):
        return transcribed, None, None, None

    detected = detect_language(transcribed)

    # Auto-switch between English and Urdu if auto selected
    if target_lang_hint == "auto":
        target_lang_hint = "ur" if detected == "en" else "en"

    translated = transcribed
    if target_lang_hint != detected:
        translated = translate_text(transcribed, dest_lang=target_lang_hint)

    if translated.startswith("[Translation error]"):
        return transcribed, detected, translated, None

    audio_output = text_to_speech(translated, target_lang_hint)
    if isinstance(audio_output, str) and audio_output.startswith("[TTS error]"):
        return transcribed, detected, translated, None

    return transcribed, detected, translated, audio_output


# -----------------------------
# STEP 7: Gradio Interface
# -----------------------------
with gr.Blocks(theme=gr.themes.Soft()) as demo:
    gr.Markdown("## 🎙 AI Voice Translator")
    gr.Markdown("Speak in any language — it will transcribe, translate, and speak it back!")

    with gr.Row():
        audio_input = gr.Audio(sources=["microphone"], type="filepath", label="🎤 Record Your Speech")
        lang_dropdown = gr.Dropdown(
            choices=["auto", "en", "ur", "fr", "es", "de", "ar"],
            value="auto",
            label="🌐 Target Language (auto detects and switches)"
        )

    with gr.Row():
        transcribed_text = gr.Textbox(label="📝 Transcribed Text")
        detected_lang = gr.Textbox(label="🌍 Detected Language")
        translated_text = gr.Textbox(label="💬 Translated Text")

    audio_output = gr.Audio(label="🔊 Output Speech")

    translate_button = gr.Button("🚀 Translate & Speak")

    translate_button.click(
        pipeline_translate,
        inputs=[audio_input, lang_dropdown],
        outputs=[transcribed_text, detected_lang, translated_text, audio_output]
    )

# -----------------------------
# STEP 8: Launch App
# -----------------------------
if __name__ == "__main__":
    demo.launch()