app.py

from tensorflow.keras.models import load_model
from flask import Flask, render_template, request, session, redirect, url_for
import pandas as pd

import pandas as pd
import numpy as np

from helpers.helper import get_img, get_processed, get_reviews, download_stopwords
from flask_session import Session


app = Flask(__name__)
app.secret_key = "123"
app.config["SESSION_TYPE"] = "filesystem"  # Use filesystem for session storage
app.config["SESSION_PERMANENT"] = False  # Sessions are not permanent by default
app.config["SESSION_FILE_DIR"] = "./flask_sessions"

Session(app)

df = None
lstm_model = None
suggestions = pd.DataFrame()


def preload():
    global df
    global lstm_model
    global suggestions
    df = pd.read_csv(r"./static/db/movie.csv")
    lstm_model = load_model(r"./static/db/lstm123.keras")
    download_stopwords()


# Home Page - User Input
@app.route("/", methods=["GET", "POST"])
def index():
    if request.method == "GET":
        query = request.args.get("query", "")
        if query:
            session["search_text"] = query
            return redirect(url_for("search"))
    return render_template(
        "index.html",
    )


# Suggestion Page
@app.route("/search", methods=["GET", "POST"])
def search():
    global suggestions
    if request.method == "POST":
        selected_index = request.form.get("selected_index")
        session["selected_index"] = selected_index
        if selected_index:
            return redirect(url_for("reviews"))
    query = session.get("search_text", [])
    if query:
        suggestions = df[df["primaryTitle"].str.contains(query, case=False, na=False)]
        # imgdata = get_img(suggestions)
        imgdata = [None] * len(suggestions)
        suggestions["imgurl"] = imgdata
        session["suggestions"] = suggestions
        dft_list = suggestions.to_dict(orient="records")
    return render_template(
        "suggestions.html",
        dft_list=dft_list,
        suggestions=suggestions,
        svg_file="images/no.svg",
    )


# Reviews Page
@app.route("/reviews", methods=["GET", "POST"])
def reviews():
    qs = session.get("search_text", None)
    selected_index = session.get("selected_index")
    print("SEL", selected_index)
    if qs:
        suggestions = session.get("suggestions")
        if suggestions is not None and not suggestions.empty:
            sugg = suggestions[
                suggestions["primaryTitle"].str.contains(qs, case=False, na=False)
            ]
            if selected_index is not None:
                selected_movie = sugg.iloc[int(selected_index)]
                tconst = selected_movie["tconst"]
                print("TConst", tconst)
                if tconst:
                    reviews = get_reviews(tconst)
                    if reviews is not None and not reviews.empty:
                        unseen_padded = get_processed(reviews["user_review"])
                        print(unseen_padded)

                        unseen_sentiments = lstm_model.predict(unseen_padded)
                        reviews["predicted_score"] = np.round(unseen_sentiments * 10, 1)

                        reviews_list = reviews.to_dict(orient="records")
                        return render_template("review.html", reviews=reviews_list)
                    else:
                        return render_template("noreview.html")
    return "Failed"


if __name__ == "__main__":
    preload()
    app.run(debug=True)