import streamlit as st
from googletrans import Translator  # Free Google Translate API
from transformers import pipeline
import requests

# Initialize the Google Translator
translator = Translator()

# Streamlit UI setup
st.set_page_config(page_title="AI-Powered Language Learning Assistant", page_icon="🧠", layout="wide")

# Header and introduction
st.title("🧠 AI-Powered Language Learning Assistant")
st.markdown("""
Welcome to your AI-powered language assistant! Here you can:
- Translate words or sentences to different languages
- Learn and practice new vocabulary
- Get grammar feedback.
""")

# Input field for text
text_input = st.text_input("Enter the text you want to translate or practice", "")

# Select target language for translation
language = st.selectbox("Select the language to translate to", ["es", "fr", "de", "it", "pt", "ru"])

if text_input:
    # Translate text using googletrans
    st.subheader(f"Original Text: {text_input}")
    translated_text = translator.translate(text_input, dest=language).text

    # Display translation
    st.markdown(f"### Translated Text to {language.upper()}:")
    st.write(translated_text)

    # Show pronunciation tip
    st.subheader("Pronunciation Tip:")
    st.write("Use Google Translate or Forvo to practice pronunciation.")

    # Grammar Check (using LanguageTool)
    st.subheader("Grammar Feedback:")
    grammar_check_url = "https://api.languagetool.org/v2/check"
    params = {
        "text": text_input,
        "language": "en-US"
    }
    response = requests.post(grammar_check_url, data=params)
    if response.status_code == 200:
        result = response.json()
        if result['matches']:
            st.write("### Grammar Issues Found:")
            for match in result['matches']:
                st.write(f"- **{match['message']}** at position {match['offset']}-{match['offset']+match['length']}")
        else:
            st.write("No grammar issues found!")
    else:
        st.write("Grammar check failed. Try again later.")

# Vocabulary practice section using Hugging Face's BERT
st.markdown("---")
st.header("Vocabulary Practice")
word_input = st.text_input("Enter a word to get its definition and synonyms", "")
if word_input:
    # Using Hugging Face's BERT model for related words (synonyms)
    try:
        word_model = pipeline("fill-mask", model="bert-base-uncased")  # Using BERT to predict related words
        result = word_model(f"The synonym of {word_input} is [MASK].")
        st.write(f"Synonyms or related words for **{word_input}**: {result}")
    except Exception as e:
        st.error("Error fetching vocabulary practice data.")

# Footer for engagement
st.markdown("""
---
**Need more practice?** Visit [Google Translate](https://translate.google.com) for real-time translations and pronunciation!
""")