#importing all the libraries needed for basic classification and presentation of cover page
import streamlit as st
from transformers import pipeline
#adding the css presentation from the styles file in utils
from utils.styles import load_css

#configure the page to display the intended title
st.set_page_config(
    page_title="Daily Mirror News Analyzer",
    layout="wide"
)

#calling the css changes
load_css()

#writing the title
st.markdown('<div class="main_title">News App — Created by Group 09</div>',
    unsafe_allow_html=True)

#writing the description 
st.write('''<div class="feature_text">Welcome! In this app users can freely 
carry out classification of news content, ask our model questions and play 
our mini-game to see your classification skills </div> <br>''',unsafe_allow_html=True)
st.write('<div class="feature_text">The features of our webapp are shown below: </div>', unsafe_allow_html=True)

#dividing the page
st.divider()

#creating the features 
col1, col2 = st.columns(2)

with col1:
    #writing the topic fo the first feature of the app
    st.markdown('<div class="topic_header">📂 CSV Content Features</div>', unsafe_allow_html=True)
    #writing the content inside the first feature 
    st.markdown("""
        <div class="feature_box">
            <div class="feature_text">
                • <b>Classification:</b> Upload a CSV and label the data<br>
                • <b>Analysis:</b> See explanatory data analysis here <br>
                • <b>Q & A:</b> Ask the model questions from your data
            </div>
        </div>
    """, unsafe_allow_html=True)

with col2:
    #writing the topic fo the second feature of the app
    st.markdown('<div class="topic_header">🎮 Interactive Mini-Game</div>', unsafe_allow_html=True)
    #writing the content inside the second feature 
    st.markdown("""
        <div class="feature_box">
            <div class="feature_text">
                • <b>Classification Test:</b> Play a game where you classify news snippets <br>
            </div>
        </div>
    """, unsafe_allow_html=True)

#dividing the page
st.divider()

#creating the model used for quick classification
st.markdown('<div class="topic_header">🔍 Quick Model Test</div>', unsafe_allow_html=True)
st.write('<div class="feature_text">Type a sentence below to see how our fine-tuned model classifies news:</div>',unsafe_allow_html=True)

#calling and loading the fine tuned model which is uploaded to the hugging face 
@st.cache_resource
def load_classifier():
    return pipeline("text-classification", model="SuchithTissera/text_analytics_model", truncation=True)

#calling the model which is now loaded into a simple variable 
classifier = load_classifier()

#getting the user text
user_text = st.text_area(" ", placeholder="Enter news content here...", height=100)

#carrying out the classification
if st.button("Classify Text"):
    if not user_text.strip():
        st.warning("Please enter some text.")
    else:
        with st.spinner("Analyzing..."):
            result = classifier(user_text)
        
        st.success("Analysis Complete")
        res_col1, res_col2 = st.columns(2)
        with res_col1:
            st.metric(label="Predicted Category", value=result[0]["label"])
        with res_col2:
            confidence = round(result[0]["score"] * 100, 2)
            st.metric(label="Confidence Score", value=f"{confidence}%")