app.py

import os
import firebase_admin
from firebase_admin import credentials, firestore
from dotenv import load_dotenv, find_dotenv
from datetime import date
import logging
import json
import streamlit as st

# Set up logging configuration.
logging.basicConfig(
    format="%(asctime)s [%(levelname)s] %(message)s",
    level=logging.INFO
)

# --------------- Firebase Initialization Functions ---------------

def initialize_environment():
    """
    Loads environment variables from the .env file.
    """
    logging.info("Loading ")
    dotenv_path = find_dotenv()
    load_dotenv(dotenv_path)
    logging.info(f"Environment variables loaded from {dotenv_path}")

def initialize_firebase():
    """
    Initializes Firebase using credentials.
    Returns a Firestore client.
    """
    
    if not firebase_admin._apps:
        data = json.loads(os.getenv("firebase_detailed_cred"))
        cred = credentials.Certificate(data)
        firebase_admin.initialize_app(cred)
        logging.info("Firebase initialized using provided credentials.")
    else:
        logging.info("Firebase already initialized.")
    return firestore.client()

# # --------------- Firestore Query Function ---------------
# def fetch_news_firestore(start_date: str, end_date: str, categories, keyword, db_client):
#     """
#     Queries Firestore for news articles in the "ProcessedNews" collection that
#     match the provided date range and categories.
#     Performs client-side filtering for keywords.
#     Limits the results to 100 articles.
#     """
#     logging.info(f"Querying Firestore for articles from {start_date} to {end_date}.")
#     query = db_client.collection("ProcessedNews") \
#                      .where("Date", ">=", start_date) \
#                      .where("Date", "<=", end_date) \
#                      .limit(100)
    
#     if categories:
#         logging.info(f"Filtering by categories: {categories}")
#         query = query.where("category", "in", categories)
    
#     docs = query.stream()
#     results = []
#     for doc in docs:
#         data = doc.to_dict()
#         results.append(data)
#     logging.info(f"Retrieved {len(results)} articles from Firestore before keyword filtering.")
    
#     if keyword:
#         kw = keyword.lower()
#         results = [
#             article for article in results
#             if kw in article.get("title", "").lower() or kw in article.get("summary", "").lower()
#         ]
#         logging.info(f"{len(results)} articles remain after keyword filtering with keyword '{keyword}'.")
#     return results


from firebase_admin.firestore import FieldFilter
from datetime import datetime

def fetch_news_firestore(start_date: str, end_date: str, categories, keyword, db_client):
    """
    Queries Firestore for news articles using correct filter syntax and date handling,
    and removes duplicate articles based on the "title" field.
    """
    logging.info(f"Querying Firestore for articles from {start_date} to {end_date}.")

    # Convert input dates to datetime objects
    try:
        start_date_dt = datetime.strptime(start_date, '%Y-%m-%d')
        end_date_dt = datetime.strptime(end_date, '%Y-%m-%d')
    except ValueError as e:
        logging.error(f"Invalid date format: {e}. Use YYYY-MM-DD.")
        return []

    # Assuming 'Date' is stored as a string in Firestore. Adjust if using Timestamp.
    start_date_str = start_date_dt.strftime('%Y-%m-%d')
    end_date_str = end_date_dt.strftime('%Y-%m-%d')

    # Build query with FieldFilter
    query = db_client.collection("ProcessedNews")
    query = query.where(filter=FieldFilter("Date", ">=", start_date_str))
    query = query.where(filter=FieldFilter("Date", "<=", end_date_str))
    query = query.limit(100)

    if categories:
        logging.info(f"Filtering by categories: {categories}")
        query = query.where(filter=FieldFilter("category", "in", categories))

    docs = query.stream()
    results = [doc.to_dict() for doc in docs]
    logging.info(f"Retrieved {len(results)} articles from Firestore before keyword filtering.")

    # Keyword filtering remains the same
    if keyword:
        kw = keyword.lower()
        results = [
            article for article in results
            if kw in article.get("title", "").lower() or kw in article.get("summary", "").lower()
        ]
        logging.info(f"{len(results)} articles remain after keyword filtering.")

    # Remove duplicate articles based on title:
    unique_results = []
    seen_titles = set()
    for article in results:
        title = article.get("title", "").strip()
        # Only add article if title is non-empty and hasn't been seen yet
        if title and title not in seen_titles:
            seen_titles.add(title)
            unique_results.append(article)
    logging.info(f"{len(unique_results)} unique articles remain after duplicate removal.")
    
    return unique_results





# --------------- Streamlit UI & Styles ---------------
def local_css(css_text):
    st.markdown(f'<style>{css_text}</style>', unsafe_allow_html=True)

light_mode_styles = """
/* Global Styles for Light Mode */
body, .stApp {
    background-color: #f8f9fa !important;
    color: #212529 !important;
    font-family: 'Helvetica Neue', Helvetica, Arial, sans-serif;
}

/* Overall text decoration */
h1, h2, h3, h4, h5, h6 {
    text-shadow: 1px 1px 2px rgba(0,0,0,0.1);
    font-weight: 700;
}
.container h2 {
    text-align: center;
    text-transform: uppercase;
    color: #FF7E5F;
    border-bottom: 2px solid #FF7E5F;
    padding-bottom: 0.5rem;
    margin-bottom: 1.5rem;
}
p {
    font-size: 1rem;
    line-height: 1.6;
    margin-bottom: 1rem;
}
a {
    text-decoration: underline;
    transition: color 0.2s ease;
}
a:hover {
    color: #FF7E5F;
}

/* Main Container */
.container {
    max-width: 800px;
    margin: 2rem auto;
    padding: 0 1rem;
}

/* Hero Section with Brighter Orange */
.hero {
    background: linear-gradient(135deg, rgba(255,140,0,0.9), rgba(255,69,0,0.9)),
                url('https://images.unsplash.com/photo-1529257414775-5d6c2509fbe6?fit=crop&w=1600&q=80');
    background-size: cover;
    background-position: center;
    border-radius: 12px;
    padding: 6rem 2rem;
    text-align: center;
    color: #fff;
    box-shadow: 0 4px 20px rgba(0,0,0,0.2);
    animation: fadeIn 1s ease-in-out;
}
@keyframes fadeIn {
    from { opacity: 0; transform: translateY(20px); }
    to { opacity: 1; transform: translateY(0); }
}
.hero h1 {
    font-size: 3rem;
    margin-bottom: 1.5rem;
    text-transform: uppercase;
    letter-spacing: 2px;
}
.hero p {
    font-size: 1.3rem;
    max-width: 800px;
    margin: 0 auto;
    line-height: 1.6;
}

/* News Cards with Enhanced Shadow & Animation */
.news-container {
    display: flex;
    flex-direction: column;
    gap: 2.5rem;
    margin-bottom: 2rem;
}
.news-card {
    background-color: #fff;
    color: #212529;
    border-radius: 8px;
    border-left: 4px solid #FF7E5F;
    box-shadow: 0 4px 10px rgba(0,0,0,0.1);
    padding: 1.5rem;
    transition: transform 0.3s ease, box-shadow 0.3s ease;
    margin-bottom: 1rem;
}
.news-card:hover {
    transform: translateY(-5px) scale(1.03);
    box-shadow: 0 8px 16px rgba(255,126,95,0.3);
}
.news-card h2 {
    font-size: 1.2rem;
    margin-bottom: 0.5rem;
}
.news-card .meta {
    font-size: 0.9rem;
    color: #6c757d;
    margin-bottom: 0.8rem;
}
.news-card p {
    font-size: 0.95rem;
    line-height: 1.4;
}
.news-card a {
    color: #007bff;
    text-decoration: none;
    font-weight: 500;
}
"""

# --------------- Main App ---------------
def main():
    st.set_page_config(page_title="Modern News Feed", layout="centered")
    st.sidebar.title("Refine Your Feed")
    local_css(light_mode_styles)

    # Initialize Firebase and environment variables once.
    if "db_client" not in st.session_state:
        initialize_environment()
        st.session_state.db_client = initialize_firebase()

    db_client = st.session_state.db_client

    # Sidebar filters.
    today_date = date.today()
    default_start = today_date.replace(day=1)
    default_end = today_date

    date_range = st.sidebar.date_input("Select Date Range", [default_start, default_end])
    if len(date_range) != 2:
        st.sidebar.error("Please select both a start and an end date.")
        return

    start_date_obj, end_date_obj = date_range
    start_date_str = start_date_obj.strftime("%Y-%m-%d")
    end_date_str = end_date_obj.strftime("%Y-%m-%d")

    categories_list = ["World", "Sports", "Business", "Sci/Tech", "Politics", "Entertainment","Others"]
    selected_categories = st.sidebar.multiselect("Select Categories", options=categories_list)
    keyword = st.sidebar.text_input("Search Keywords")

    # Two sidebar buttons.
    col1, col2 = st.sidebar.columns(2)
    with col1:
        get_news = st.button("Get News")
    with col2:
        clear_news = st.button("Clear News")

    # When "Get News" is pressed, query Firestore.
    if get_news:
        logging.info("Get News button pressed.")
        st.session_state.news_pressed = True
        with st.spinner("Fetching articles..."):
            articles = fetch_news_firestore(start_date_str, end_date_str, selected_categories, keyword, db_client)
        st.session_state.articles = articles
        logging.info(f"{len(articles)} articles stored in session state after filtering.")

    # When "Clear News" is pressed, clear the session.
    if clear_news:
        logging.info("Clear News button pressed. Clearing articles and flags.")
        st.session_state.pop("articles", None)
        st.session_state.pop("news_pressed", None)

    # --------------- MAIN CONTENT ---------------
    st.markdown("<div class='container'>", unsafe_allow_html=True)

    # If Get News has been pressed, show results or an info message.
    if st.session_state.get("news_pressed", False):
        articles_to_show = st.session_state.get("articles", [])
        if articles_to_show:
            st.markdown("## Explore the Stories That Matter")
            st.success(f"Found {len(articles_to_show)} articles")
            st.markdown("<div class='news-container'>", unsafe_allow_html=True)
            for article in articles_to_show:
                st.markdown(f"""
                    <div class="news-card">
                        <h2>{article.get("title", "No Title")}</h2>
                        <div class="meta">
                            <strong>Date:</strong> {article.get("Date", "Unknown")} &nbsp;&nbsp; 
                            <strong>Category:</strong> {article.get("category", "Uncategorized")}
                        </div>
                        <p>{article.get("summary", "No Summary Available")}</p>
                        {'<a href="' + article.get("url", "#") + '" target="_blank">Read More</a>' if article.get("url") else ""}
                    </div>
                """, unsafe_allow_html=True)
            st.markdown("</div>", unsafe_allow_html=True)
        else:
            st.info("No news meet the requirement. Please adjust your filters and try again.")
            logging.info("No articles found matching the filters.")
    else:
        # If Get News hasn't been pressed, show the hero section.
        st.markdown("""
            <div class="hero">
                <h1>News Spotlight</h1>
                <p>Discover fresh stories and insights that ignite your curiosity. Your journey into groundbreaking news starts here.</p>
            </div>
        """, unsafe_allow_html=True)

    st.markdown("</div>", unsafe_allow_html=True)

if __name__ == "__main__":
    main()