recommender/recommender.py

import ast
import re
import pandas as pd
import tempfile
import os
from intent_classification.fd_classification import find_intent
from intent_classification.retrieval_classification import handle_query_classification,find_matching_card,generate_card_response_with_context
from recommender.retrieval_ranking import generate_multi_queries,convert_to_direct_query_gradio,retrieve_and_rank_cards,generate_credit_card_recommendation_gemini,cross_encoder
from data import eligibility_lookup,card_features_lookup,all_card_names,features
from recommender.graph_retrieval_vectordb import generate_cypher,run_cypher_query,Neo4jConnectionError

#function to pass the retrieved cards and generated response to the UI
def recommend_cards_gradio(user_query, preferences, income, cibil, age,
                           min_joining_fee, max_joining_fee,
                           min_annual_fee, max_annual_fee,
                           use_eligibility=True,):
    try:
        # print(user_query)
        if(user_query):
            result = handle_query_classification(user_query)

            if result["intent"] == "no_retrieval":
                return (
                    f"<div style='background-color:#e3f2fd;padding:20px;border-radius:10px;'>"
                    f"<pre style='white-space:pre-wrap;font-size:13px;color:#212121;'>{result['response']}</pre></div>",
                    [["No retrieval required", "Answered using LLM"]],
                    None,
                    [],
                    {},
                    "Answered without retrieval"
                )
            elif result["intent"] == "specific":
                matched_card = find_matching_card(user_query)
                if matched_card:
                    gemini_answer = generate_card_response_with_context(user_query, matched_card)
                    card_name = matched_card["name"]
                    card_desc = matched_card["description"]
                    card_lookup = {card_name: card_desc}
            
                    # Constructing eligibility info if available
                    eligibility_info = eligibility_lookup.get(card_name, "No eligibility or fee information available.")
                    chat_history_entry = f"{card_name}:\n{card_desc}\n\nEligibility & Fees:\n{eligibility_info}"
            
                    return (
                        f"<div style='background-color:#fffde7;padding:20px;border-radius:10px;'>"
                        f"<pre style='white-space:pre-wrap;font-size:13px;color:#212121;'>{gemini_answer}</pre></div>",
                        [["Specific card detected", card_name]],
                        None,
                        [],
                        card_lookup,
                        user_query 
                    )
                else:
                    return (
                        "<b style='color:red;'>Card mentioned not found in database.</b>",
                        [["Card not found", "Try another card name."]],
                        None,
                        [],
                        {},
                        "Card not found"
                    )
        
        direct_query,excluded_cards = convert_to_direct_query_gradio(user_query, preferences, all_card_names=all_card_names,feature_list=features)
        queries = generate_multi_queries(direct_query)

        if cibil < 700 and use_eligibility:
            query_intent = True
        else:
            query_intent = find_intent(user_query)
            print(query_intent)
        cypher_query = generate_cypher(direct_query, query_intent)
        print("Generated Cypher:\n", cypher_query)

        try:
            faiss_index, filtered_mapping = run_cypher_query(
                user_query, cypher_query, use_eligibility,
                income, cibil, age,
                min_joining_fee, max_joining_fee,
                min_annual_fee, max_annual_fee,excluded_cards
            )
        except Neo4jConnectionError as graph_err:
            return (
                "<b style='color:red;'>Graph database connection failed. Please try again later.</b>",
                [["Graph database error", str(graph_err)]],
                None,
                [],
                {},
                "Graph DB connection error"
            )

        cards = retrieve_and_rank_cards(faiss_index, filtered_mapping, direct_query, queries, top_k=10)
        gemini_summary = generate_credit_card_recommendation_gemini(user_query, direct_query, cards)

        if not cards:
            return (
                "<b style='color:red;'>No eligible cards found.</b>",
                [["No eligible cards found", "Please try a different query or check your input values."]],
                None,
                [],
                {},
                "No eligible card found"
            )

        match = re.search(r"f\.name IN (\[.*?\])", cypher_query)
        query_features = set(ast.literal_eval(match.group(1))) if match else set()

        card_rows = []
        for score, card in sorted(
            zip(cross_encoder.predict([[direct_query, card["description"]] for card in cards]), cards),
            reverse=True,
            key=lambda x: x[0]
        ):
            card_name = card["name"]
            card_desc = card["description"]
            matched_features = query_features.intersection(card_features_lookup.get(card_name, set()))
            feature_str = ", ".join(matched_features) if matched_features else "None"
            card_rows.append([card_name, feature_str, card_desc])

        
        card_names = [row[0] for row in card_rows]
        card_lookup = {row[0]: row[2] for row in card_rows}
        top_card_html = f"""
            <div style="
                background-color: #fff3e0;
                color: #212121;
                border-radius: 16px;
                padding: 20px;
                border: 2px solid #ffa726;
                box-shadow: 2px 2px 8px rgba(0,0,0,0.1);
                margin-bottom: 16px;
                font-family: sans-serif;
                font-size: 8px;
            ">
                <pre style="white-space: pre-wrap; font-size: 13px; color: #212121;">{gemini_summary}</pre>
            </div>
        """

        df_cards = pd.DataFrame(card_rows, columns=["Card Name", "Matched Features", "Description"])
        filename = "recommended_cards.csv"
        temp_dir = tempfile.gettempdir()
        file_path = os.path.join(temp_dir, filename)
        df_cards.to_csv(file_path, index=False)
        
        return top_card_html, card_rows, file_path, card_names, card_lookup, direct_query

    except Exception as e:
        print("Error:", e)
        return (
            "An unexpected error occurred. Please try again in a few minutes.",
            [["Something went wrong", "Please try again."]],
            None,
            [],
            {},
            "Unexpected error occurred, please try again in a while"
        )