app.py

import gradio as gr
import pandas as pd
import os
import logging
import matplotlib
matplotlib.use('Agg') # Set backend for Matplotlib to avoid GUI conflicts with Gradio
import matplotlib.pyplot as plt
import time # For profiling if needed
from datetime import datetime

# --- Module Imports ---
from gradio_utils import get_url_user_token
# Functions from newly created/refactored modules
from config import (
    LINKEDIN_CLIENT_ID_ENV_VAR, BUBBLE_APP_NAME_ENV_VAR,
    BUBBLE_API_KEY_PRIVATE_ENV_VAR, BUBBLE_API_ENDPOINT_ENV_VAR)
from state_manager import process_and_store_bubble_token
from sync_logic import sync_all_linkedin_data_orchestrator
from ui_generators import (
    display_main_dashboard,
    run_mentions_tab_display,
    run_follower_stats_tab_display,
    build_analytics_tab_plot_area, 
    BOMB_ICON, EXPLORE_ICON, FORMULA_ICON, ACTIVE_ICON 
)
from analytics_data_processing import prepare_filtered_analytics_data # This is key for data structure
from analytics_plot_generator import (
    generate_posts_activity_plot, 
    generate_mentions_activity_plot, generate_mention_sentiment_plot,
    generate_followers_count_over_time_plot,
    generate_followers_growth_rate_plot,
    generate_followers_by_demographics_plot,
    generate_engagement_rate_over_time_plot,
    generate_reach_over_time_plot,
    generate_impressions_over_time_plot,
    create_placeholder_plot,
    generate_likes_over_time_plot,
    generate_clicks_over_time_plot,
    generate_shares_over_time_plot,
    generate_comments_over_time_plot,
    generate_comments_sentiment_breakdown_plot,
    generate_post_frequency_plot,
    generate_content_format_breakdown_plot,
    generate_content_topic_breakdown_plot
)
from formulas import PLOT_FORMULAS 

# --- NEW CHATBOT MODULE IMPORTS ---
from chatbot_prompts import get_initial_insight_and_suggestions
from chatbot_handler import generate_llm_response
# --- END NEW CHATBOT MODULE IMPORTS ---

# Configure logging
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(module)s - %(message)s')

# Mapping from plot_configs IDs to PLOT_FORMULAS keys
PLOT_ID_TO_FORMULA_KEY_MAP = {
    "posts_activity": "posts_activity", 
    "mentions_activity": "mentions_activity", 
    "mention_sentiment": "mention_sentiment", 
    "followers_count": "followers_count_over_time",
    "followers_growth_rate": "followers_growth_rate",
    "followers_by_location": "followers_by_demographics",
    "followers_by_role": "followers_by_demographics",
    "followers_by_industry": "followers_by_demographics",
    "followers_by_seniority": "followers_by_demographics",
    "engagement_rate": "engagement_rate_over_time",
    "reach_over_time": "reach_over_time",
    "impressions_over_time": "impressions_over_time",
    "likes_over_time": "likes_over_time",
    "clicks_over_time": "clicks_over_time",
    "shares_over_time": "shares_over_time",
    "comments_over_time": "comments_over_time",
    "comments_sentiment": "comments_sentiment_breakdown",
    "post_frequency_cs": "post_frequency",
    "content_format_breakdown_cs": "content_format_breakdown",
    "content_topic_breakdown_cs": "content_topic_breakdown",
    "mention_analysis_volume": "mentions_activity", 
    "mention_analysis_sentiment": "mention_sentiment"
}

# --- Helper function to generate textual data summaries for chatbot ---
def generate_chatbot_data_summaries(
    plot_configs_list,
    filtered_merged_posts_df,
    filtered_mentions_df,
    date_filtered_follower_stats_df,
    raw_follower_stats_df,
    token_state_value # To get column names if needed
    ):
    """
    Generates textual summaries for each plot ID to be used by the chatbot.
    """
    data_summaries = {}
    date_col_posts = token_state_value.get("config_date_col_posts", "published_at")
    # Ensure date columns are datetime objects for proper formatting and comparison
    if date_col_posts in filtered_merged_posts_df.columns:
        filtered_merged_posts_df[date_col_posts] = pd.to_datetime(filtered_merged_posts_df[date_col_posts], errors='coerce')
    if 'date' in date_filtered_follower_stats_df.columns:
         date_filtered_follower_stats_df['date'] = pd.to_datetime(date_filtered_follower_stats_df['date'], errors='coerce')
    # Add more date conversions as needed for other dataframes

    for plot_cfg in plot_configs_list:
        plot_id = plot_cfg["id"]
        summary_text = f"No data summary available for {plot_cfg['label']} for the selected period."

        try:
            if plot_id == "followers_count" and not date_filtered_follower_stats_df.empty:
                df_summary = date_filtered_follower_stats_df[['date', 'follower_gains_monthly']].copy()
                df_summary['date'] = df_summary['date'].dt.strftime('%Y-%m-%d')
                summary_text = f"Follower Count (Monthly Gains):\n{df_summary.tail(5).to_string(index=False)}"
            
            elif plot_id == "followers_growth_rate" and not date_filtered_follower_stats_df.empty:
                df_summary = date_filtered_follower_stats_df[['date', 'growth_rate_monthly']].copy() # Assuming 'growth_rate_monthly' column
                df_summary['date'] = df_summary['date'].dt.strftime('%Y-%m-%d')
                summary_text = f"Follower Growth Rate (Monthly):\n{df_summary.tail(5).to_string(index=False)}"

            elif plot_id == "followers_by_location" and not raw_follower_stats_df.empty and 'follower_geo' in raw_follower_stats_df.columns:
                # Assuming 'follower_geo' contains location names and another column like 'count' exists
                # This part needs to align with how generate_followers_by_demographics_plot processes 'follower_geo' data
                # For simplicity, let's assume raw_follower_stats_df has pre-aggregated data for geo
                # This is a placeholder, actual column names for count and name are needed from your data prep
                if 'geo_name' in raw_follower_stats_df.columns and 'geo_count' in raw_follower_stats_df.columns:
                    df_summary = raw_follower_stats_df[['geo_name', 'geo_count']].nlargest(5, 'geo_count')
                    summary_text = f"Followers by Location (Top 5):\n{df_summary.to_string(index=False)}"
                else:
                    summary_text = "Follower location data structure not as expected for summary."


            elif plot_id == "engagement_rate" and not filtered_merged_posts_df.empty:
                df_summary = filtered_merged_posts_df[[date_col_posts, 'engagement_rate']].copy()
                df_summary[date_col_posts] = df_summary[date_col_posts].dt.strftime('%Y-%m-%d')
                summary_text = f"Engagement Rate Over Time:\n{df_summary.tail(5).to_string(index=False)}"
            
            elif plot_id == "reach_over_time" and not filtered_merged_posts_df.empty:
                df_summary = filtered_merged_posts_df[[date_col_posts, 'reach']].copy()
                df_summary[date_col_posts] = df_summary[date_col_posts].dt.strftime('%Y-%m-%d')
                summary_text = f"Reach Over Time:\n{df_summary.tail(5).to_string(index=False)}"

            elif plot_id == "impressions_over_time" and not filtered_merged_posts_df.empty:
                df_summary = filtered_merged_posts_df[[date_col_posts, 'impressions_sum']].copy() # Assuming 'impressions_sum' or similar
                df_summary[date_col_posts] = df_summary[date_col_posts].dt.strftime('%Y-%m-%d')
                summary_text = f"Impressions Over Time:\n{df_summary.tail(5).to_string(index=False)}"

            elif plot_id == "comments_sentiment" and not filtered_merged_posts_df.empty and 'comment_sentiment' in filtered_merged_posts_df.columns:
                sentiment_counts = filtered_merged_posts_df['comment_sentiment'].value_counts().reset_index()
                sentiment_counts.columns = ['Sentiment', 'Count']
                summary_text = f"Comments Sentiment Breakdown:\n{sentiment_counts.to_string(index=False)}"

            # Add more elif blocks for other plot_ids, extracting relevant data:
            # e.g., likes_over_time, clicks_over_time, shares_over_time, comments_over_time
            # post_frequency_cs, content_format_breakdown_cs, content_topic_breakdown_cs
            # mentions_activity, mention_sentiment

            data_summaries[plot_id] = summary_text
        except KeyError as e:
            logging.warning(f"KeyError generating summary for {plot_id}: {e}. Using default summary.")
            data_summaries[plot_id] = f"Data summary generation error for {plot_cfg['label']} (missing column: {e})."
        except Exception as e:
            logging.error(f"Error generating summary for {plot_id}: {e}", exc_info=True)
            data_summaries[plot_id] = f"Error generating data summary for {plot_cfg['label']}."
            
    return data_summaries

# --- Analytics Tab: Plot Figure Generation Function ---
def update_analytics_plots_figures(token_state_value, date_filter_option, custom_start_date, custom_end_date, current_plot_configs):
    logging.info(f"Updating analytics plot figures. Filter: {date_filter_option}, Custom Start: {custom_start_date}, Custom End: {custom_end_date}")
    num_expected_plots = 19 # Ensure this matches the number of plots generated
    
    plot_data_summaries_for_chatbot = {} # Initialize dict for chatbot summaries

    if not token_state_value or not token_state_value.get("token"):
        message = "❌ Accesso negato. Nessun token. Impossibile generare le analisi."
        logging.warning(message)
        placeholder_figs = [create_placeholder_plot(title="Accesso Negato", message="Nessun token.") for _ in range(num_expected_plots)]
        # For each plot_config, add a default "no data" summary
        for p_cfg in current_plot_configs:
            plot_data_summaries_for_chatbot[p_cfg["id"]] = "Accesso negato, nessun dato per il chatbot."
        return [message] + placeholder_figs + [plot_data_summaries_for_chatbot]
    try:
        (filtered_merged_posts_df, 
         filtered_mentions_df, 
         date_filtered_follower_stats_df, # For time-based follower plots
         raw_follower_stats_df,       # For demographic follower plots
         start_dt_for_msg, end_dt_for_msg) = \
            prepare_filtered_analytics_data(
                token_state_value, date_filter_option, custom_start_date, custom_end_date
            )
        
        # Generate data summaries for chatbot AFTER data preparation
        plot_data_summaries_for_chatbot = generate_chatbot_data_summaries(
            current_plot_configs, # Pass the plot_configs list
            filtered_merged_posts_df,
            filtered_mentions_df,
            date_filtered_follower_stats_df,
            raw_follower_stats_df,
            token_state_value
        )

    except Exception as e:
        error_msg = f"❌ Errore durante la preparazione dei dati per le analisi: {e}"
        logging.error(error_msg, exc_info=True)
        placeholder_figs = [create_placeholder_plot(title="Errore Preparazione Dati", message=str(e)) for _ in range(num_expected_plots)]
        for p_cfg in current_plot_configs:
            plot_data_summaries_for_chatbot[p_cfg["id"]] = f"Errore preparazione dati: {e}"
        return [error_msg] + placeholder_figs + [plot_data_summaries_for_chatbot]
    
    date_column_posts = token_state_value.get("config_date_col_posts", "published_at")
    date_column_mentions = token_state_value.get("config_date_col_mentions", "date")
    media_type_col_name = token_state_value.get("config_media_type_col", "media_type")
    eb_labels_col_name = token_state_value.get("config_eb_labels_col", "li_eb_label")
    
    plot_figs = [] # Initialize list to hold plot figures
    
    plot_titles_for_errors = [p_cfg["label"] for p_cfg in current_plot_configs] 

    try:
        # Dinamiche dei Follower (2 plots)
        plot_figs.append(generate_followers_count_over_time_plot(date_filtered_follower_stats_df, type_value='follower_gains_monthly'))
        plot_figs.append(generate_followers_growth_rate_plot(date_filtered_follower_stats_df, type_value='follower_gains_monthly'))
        
        # Demografia Follower (4 plots)
        plot_figs.append(generate_followers_by_demographics_plot(raw_follower_stats_df, type_value='follower_geo', plot_title="Follower per Località"))
        plot_figs.append(generate_followers_by_demographics_plot(raw_follower_stats_df, type_value='follower_function', plot_title="Follower per Ruolo"))
        plot_figs.append(generate_followers_by_demographics_plot(raw_follower_stats_df, type_value='follower_industry', plot_title="Follower per Settore"))
        plot_figs.append(generate_followers_by_demographics_plot(raw_follower_stats_df, type_value='follower_seniority', plot_title="Follower per Anzianità"))

        # Approfondimenti Performance Post (4 plots)
        plot_figs.append(generate_engagement_rate_over_time_plot(filtered_merged_posts_df, date_column=date_column_posts))
        plot_figs.append(generate_reach_over_time_plot(filtered_merged_posts_df, date_column=date_column_posts))
        plot_figs.append(generate_impressions_over_time_plot(filtered_merged_posts_df, date_column=date_column_posts))
        plot_figs.append(generate_likes_over_time_plot(filtered_merged_posts_df, date_column=date_column_posts))
        
        # Engagement Dettagliato Post nel Tempo (4 plots)
        plot_figs.append(generate_clicks_over_time_plot(filtered_merged_posts_df, date_column=date_column_posts))
        plot_figs.append(generate_shares_over_time_plot(filtered_merged_posts_df, date_column=date_column_posts))
        plot_figs.append(generate_comments_over_time_plot(filtered_merged_posts_df, date_column=date_column_posts))
        plot_figs.append(generate_comments_sentiment_breakdown_plot(filtered_merged_posts_df, sentiment_column='comment_sentiment')) # Make sure 'comment_sentiment' exists
        
        # Analisi Strategia Contenuti (3 plots)
        plot_figs.append(generate_post_frequency_plot(filtered_merged_posts_df, date_column=date_column_posts))
        plot_figs.append(generate_content_format_breakdown_plot(filtered_merged_posts_df, format_col=media_type_col_name))
        plot_figs.append(generate_content_topic_breakdown_plot(filtered_merged_posts_df, topics_col=eb_labels_col_name))
        
        # Analisi Menzioni (Dettaglio) (2 plots)
        plot_figs.append(generate_mentions_activity_plot(filtered_mentions_df, date_column=date_column_mentions))
        plot_figs.append(generate_mention_sentiment_plot(filtered_mentions_df)) # Make sure this function handles empty/malformed df

        if len(plot_figs) != num_expected_plots:
            logging.warning(f"Mismatch in generated plots. Expected {num_expected_plots}, got {len(plot_figs)}. This will cause UI update issues.")
            while len(plot_figs) < num_expected_plots:
                plot_figs.append(create_placeholder_plot(title="Grafico Non Generato", message="Logica di generazione incompleta."))
        
        message = f"📊 Analisi aggiornate per il periodo: {date_filter_option}"
        if date_filter_option == "Intervallo Personalizzato": 
            s_display = start_dt_for_msg.strftime('%Y-%m-%d') if start_dt_for_msg else "Qualsiasi"
            e_display = end_dt_for_msg.strftime('%Y-%m-%d') if end_dt_for_msg else "Qualsiasi"
            message += f" (Da: {s_display} A: {e_display})"
        
        final_plot_figs = []
        for i, p_fig_candidate in enumerate(plot_figs):
            if p_fig_candidate is not None and not isinstance(p_fig_candidate, str): # Basic check for a plot object
                final_plot_figs.append(p_fig_candidate)
            else:
                err_title = plot_titles_for_errors[i] if i < len(plot_titles_for_errors) else f"Grafico {i+1}"
                logging.warning(f"Plot {err_title} (index {i}) non è una figura valida: {p_fig_candidate}. Uso placeholder.")
                final_plot_figs.append(create_placeholder_plot(title=f"Errore: {err_title}", message="Impossibile generare figura."))
        
        return [message] + final_plot_figs[:num_expected_plots] + [plot_data_summaries_for_chatbot]

    except (KeyError, ValueError) as e_plot_data:
        logging.error(f"Errore dati durante la generazione di un grafico specifico: {e_plot_data}", exc_info=True)
        error_msg_display = f"Errore dati in un grafico: {str(e_plot_data)[:100]}"
        
        num_already_generated = len(plot_figs)
        for i in range(num_already_generated, num_expected_plots):
            err_title_fill = plot_titles_for_errors[i] if i < len(plot_titles_for_errors) else f"Grafico {i+1}"
            plot_figs.append(create_placeholder_plot(title=f"Errore Dati: {err_title_fill}", message=f"Precedente errore: {str(e_plot_data)[:50]}"))
        
        for p_cfg in current_plot_configs: # Ensure summaries dict is populated on error
            if p_cfg["id"] not in plot_data_summaries_for_chatbot:
                 plot_data_summaries_for_chatbot[p_cfg["id"]] = f"Errore dati grafico: {e_plot_data}"
        return [error_msg_display] + plot_figs[:num_expected_plots] + [plot_data_summaries_for_chatbot]

    except Exception as e_general:
        error_msg = f"❌ Errore generale durante la generazione dei grafici: {e_general}"
        logging.error(error_msg, exc_info=True)
        placeholder_figs_general = [create_placeholder_plot(title=plot_titles_for_errors[i] if i < len(plot_titles_for_errors) else f"Grafico {i+1}", message=str(e_general)) for i in range(num_expected_plots)]
        for p_cfg in current_plot_configs: # Ensure summaries dict is populated on error
             if p_cfg["id"] not in plot_data_summaries_for_chatbot:
                plot_data_summaries_for_chatbot[p_cfg["id"]] = f"Errore generale grafici: {e_general}"
        return [error_msg] + placeholder_figs_general + [plot_data_summaries_for_chatbot]


# --- Gradio UI Blocks ---
with gr.Blocks(theme=gr.themes.Soft(primary_hue="blue", secondary_hue="sky"),
               title="LinkedIn Organization Dashboard") as app:
    token_state = gr.State(value={
        "token": None, "client_id": None, "org_urn": None,
        "bubble_posts_df": pd.DataFrame(), "bubble_post_stats_df": pd.DataFrame(),
        "bubble_mentions_df": pd.DataFrame(), "bubble_follower_stats_df": pd.DataFrame(),
        "fetch_count_for_api": 0, "url_user_token_temp_storage": None,
        "config_date_col_posts": "published_at", "config_date_col_mentions": "date",
        "config_date_col_followers": "date", "config_media_type_col": "media_type",
        "config_eb_labels_col": "li_eb_label"
    })
    
    chat_histories_st = gr.State({}) 
    current_chat_plot_id_st = gr.State(None) 
    plot_data_for_chatbot_st = gr.State({}) # NEW: Store data summaries for chatbot

    gr.Markdown("# 🚀 LinkedIn Organization Dashboard")
    url_user_token_display = gr.Textbox(label="User Token (Nascosto)", interactive=False, visible=False)
    status_box = gr.Textbox(label="Stato Generale Token LinkedIn", interactive=False, value="Inizializzazione...")
    org_urn_display = gr.Textbox(label="URN Organizzazione (Nascosto)", interactive=False, visible=False)
    
    app.load(fn=get_url_user_token, inputs=None, outputs=[url_user_token_display, org_urn_display], api_name="get_url_params", show_progress=False)

    def initial_load_sequence(url_token, org_urn_val, current_state):
        status_msg, new_state, btn_update = process_and_store_bubble_token(url_token, org_urn_val, current_state)
        dashboard_content = display_main_dashboard(new_state)
        return status_msg, new_state, btn_update, dashboard_content

    with gr.Tabs() as tabs:
        with gr.TabItem("1️⃣ Dashboard & Sync", id="tab_dashboard_sync"):
            gr.Markdown("Il sistema controlla i dati esistenti da Bubble. 'Sincronizza' si attiva se sono necessari nuovi dati.")
            sync_data_btn = gr.Button("🔄 Sincronizza Dati LinkedIn", variant="primary", visible=False, interactive=False)
            sync_status_html_output = gr.HTML("<p style='text-align:center;'>Stato sincronizzazione...</p>")
            dashboard_display_html = gr.HTML("<p style='text-align:center;'>Caricamento dashboard...</p>")
            
            org_urn_display.change(
                fn=initial_load_sequence,
                inputs=[url_user_token_display, org_urn_display, token_state],
                outputs=[status_box, token_state, sync_data_btn, dashboard_display_html],
                show_progress="full"
            )

        with gr.TabItem("2️⃣ Analisi", id="tab_analytics"):
            gr.Markdown("## 📈 Analisi Performance LinkedIn")
            gr.Markdown("Seleziona un intervallo di date. Clicca i pulsanti (💣 Insights, ƒ Formula, 🧭 Esplora) su un grafico per azioni.")
            analytics_status_md = gr.Markdown("Stato analisi...")
            
            with gr.Row():
                date_filter_selector = gr.Radio(
                    ["Sempre", "Ultimi 7 Giorni", "Ultimi 30 Giorni", "Intervallo Personalizzato"],
                    label="Seleziona Intervallo Date", value="Sempre", scale=3
                )
                with gr.Column(scale=2): 
                    custom_start_date_picker = gr.DateTime(label="Data Inizio", visible=False, include_time=False, type="datetime") # Use gr.DateTime
                    custom_end_date_picker = gr.DateTime(label="Data Fine", visible=False, include_time=False, type="datetime") # Use gr.DateTime

            apply_filter_btn = gr.Button("🔍 Applica Filtro & Aggiorna Analisi", variant="primary")

            def toggle_custom_date_pickers(selection):
                is_custom = selection == "Intervallo Personalizzato"
                return gr.update(visible=is_custom), gr.update(visible=is_custom)

            date_filter_selector.change(
                fn=toggle_custom_date_pickers,
                inputs=[date_filter_selector],
                outputs=[custom_start_date_picker, custom_end_date_picker]
            )
            
            # Moved plot_configs to be globally accessible within the Blocks scope if needed by update_analytics_plots_figures
            # or pass it explicitly. For now, it's defined here.
            plot_configs = [
                {"label": "Numero di Follower nel Tempo", "id": "followers_count", "section": "Dinamiche dei Follower"},
                {"label": "Tasso di Crescita Follower", "id": "followers_growth_rate", "section": "Dinamiche dei Follower"},
                {"label": "Follower per Località", "id": "followers_by_location", "section": "Demografia Follower"},
                {"label": "Follower per Ruolo (Funzione)", "id": "followers_by_role", "section": "Demografia Follower"},
                {"label": "Follower per Settore", "id": "followers_by_industry", "section": "Demografia Follower"},
                {"label": "Follower per Anzianità", "id": "followers_by_seniority", "section": "Demografia Follower"},
                {"label": "Tasso di Engagement nel Tempo", "id": "engagement_rate", "section": "Approfondimenti Performance Post"},
                {"label": "Copertura nel Tempo", "id": "reach_over_time", "section": "Approfondimenti Performance Post"},
                {"label": "Visualizzazioni nel Tempo", "id": "impressions_over_time", "section": "Approfondimenti Performance Post"},
                {"label": "Reazioni (Like) nel Tempo", "id": "likes_over_time", "section": "Approfondimenti Performance Post"},
                {"label": "Click nel Tempo", "id": "clicks_over_time", "section": "Engagement Dettagliato Post nel Tempo"},
                {"label": "Condivisioni nel Tempo", "id": "shares_over_time", "section": "Engagement Dettagliato Post nel Tempo"},
                {"label": "Commenti nel Tempo", "id": "comments_over_time", "section": "Engagement Dettagliato Post nel Tempo"},
                {"label": "Ripartizione Commenti per Sentiment", "id": "comments_sentiment", "section": "Engagement Dettagliato Post nel Tempo"},
                {"label": "Frequenza Post", "id": "post_frequency_cs", "section": "Analisi Strategia Contenuti"},
                {"label": "Ripartizione Contenuti per Formato", "id": "content_format_breakdown_cs", "section": "Analisi Strategia Contenuti"},
                {"label": "Ripartizione Contenuti per Argomenti", "id": "content_topic_breakdown_cs", "section": "Analisi Strategia Contenuti"},
                {"label": "Volume Menzioni nel Tempo (Dettaglio)", "id": "mention_analysis_volume", "section": "Analisi Menzioni (Dettaglio)"},
                {"label": "Ripartizione Menzioni per Sentiment (Dettaglio)", "id": "mention_analysis_sentiment", "section": "Analisi Menzioni (Dettaglio)"}
            ]
            assert len(plot_configs) == 19, "Mancata corrispondenza in plot_configs e grafici attesi."

            active_panel_action_state = gr.State(None) 
            explored_plot_id_state = gr.State(None) 
            
            plot_ui_objects = {} 

            with gr.Row(equal_height=False):
                with gr.Column(scale=8) as plots_area_col:
                    plot_ui_objects = build_analytics_tab_plot_area(plot_configs)
                
                with gr.Column(scale=4, visible=False) as global_actions_column_ui: 
                    gr.Markdown("### 💡 Azioni Contestuali Grafico") 
                    
                    insights_chatbot_ui = gr.Chatbot(
                        label="Chat Insights", type="messages", height=450, 
                        bubble_full_width=False, visible=False, show_label=False,
                        placeholder="L'analisi AI del grafico apparirà qui. Fai domande di approfondimento!"
                    )
                    insights_chat_input_ui = gr.Textbox(
                        label="La tua domanda:", placeholder="Chiedi all'AI riguardo a questo grafico...", 
                        lines=2, visible=False, show_label=False
                    )
                    with gr.Row(visible=False) as insights_suggestions_row_ui:
                        insights_suggestion_1_btn = gr.Button(value="Suggerimento 1", size="sm", min_width=50)
                        insights_suggestion_2_btn = gr.Button(value="Suggerimento 2", size="sm", min_width=50)
                        insights_suggestion_3_btn = gr.Button(value="Suggerimento 3", size="sm", min_width=50)
                    
                    formula_display_markdown_ui = gr.Markdown(
                        "I dettagli sulla formula/metodologia appariranno qui.", visible=False
                    )

            async def handle_panel_action( 
                plot_id_clicked: str, 
                action_type: str, 
                current_active_action_from_state: dict, 
                current_chat_histories: dict, 
                current_chat_plot_id: str,
                current_plot_data_for_chatbot: dict # NEW: data summaries
            ):
                logging.info(f"Azione '{action_type}' per grafico: {plot_id_clicked}. Attualmente attivo: {current_active_action_from_state}")
                
                clicked_plot_config = next((p for p in plot_configs if p["id"] == plot_id_clicked), None)
                if not clicked_plot_config:
                    logging.error(f"Configurazione non trovata per plot_id {plot_id_clicked}")
                    num_button_updates = 2 * len(plot_configs) # insights, formula buttons
                    error_updates = [gr.update(visible=False)] * 10 # action_col, chatbot, input, suggestions_row, 3x sugg_btn, formula_md
                    error_updates.extend([current_active_action_from_state, current_chat_plot_id, current_chat_histories]) 
                    error_updates.extend([gr.update()] * num_button_updates) 
                    return error_updates

                clicked_plot_label = clicked_plot_config["label"]
                
                hypothetical_new_active_state = {"plot_id": plot_id_clicked, "type": action_type}
                is_toggling_off = current_active_action_from_state == hypothetical_new_active_state
                
                new_active_action_state_to_set = None
                action_col_visible_update = gr.update(visible=True)
                
                insights_chatbot_visible_update = gr.update(visible=False)
                insights_chat_input_visible_update = gr.update(visible=False)
                insights_suggestions_row_visible_update = gr.update(visible=False)
                formula_display_visible_update = gr.update(visible=False)
                
                chatbot_content_update = gr.update()
                suggestion_1_update = gr.update()
                suggestion_2_update = gr.update()
                suggestion_3_update = gr.update()
                new_current_chat_plot_id = current_chat_plot_id 
                updated_chat_histories = current_chat_histories 

                formula_content_update = gr.update()

                if is_toggling_off:
                    new_active_action_state_to_set = None
                    action_col_visible_update = gr.update(visible=False)
                    new_current_chat_plot_id = None 
                    logging.info(f"Chiusura pannello {action_type} per {plot_id_clicked}")
                else:
                    new_active_action_state_to_set = hypothetical_new_active_state
                    if action_type == "insights":
                        insights_chatbot_visible_update = gr.update(visible=True)
                        insights_chat_input_visible_update = gr.update(visible=True)
                        insights_suggestions_row_visible_update = gr.update(visible=True)
                        
                        new_current_chat_plot_id = plot_id_clicked
                        chat_history_for_this_plot = current_chat_histories.get(plot_id_clicked, [])
                        
                        # Get data summary for this plot
                        plot_specific_data_summary = current_plot_data_for_chatbot.get(plot_id_clicked, f"Nessun sommario dati specifico disponibile per '{clicked_plot_label}'.")

                        if not chat_history_for_this_plot: 
                            initial_insight_msg, suggestions = get_initial_insight_and_suggestions(
                                plot_id_clicked, 
                                clicked_plot_label,
                                plot_specific_data_summary # Pass the summary
                            )
                            chat_history_for_this_plot = [initial_insight_msg] # Gradio expects list of dicts
                            updated_chat_histories = current_chat_histories.copy()
                            updated_chat_histories[plot_id_clicked] = chat_history_for_this_plot
                        else: 
                            # If history exists, still get fresh suggestions, but don't overwrite history's first message
                            _, suggestions = get_initial_insight_and_suggestions(
                                plot_id_clicked, 
                                clicked_plot_label,
                                plot_specific_data_summary # Pass summary for context if needed by suggestions
                            )


                        chatbot_content_update = gr.update(value=chat_history_for_this_plot)
                        suggestion_1_update = gr.update(value=suggestions[0])
                        suggestion_2_update = gr.update(value=suggestions[1])
                        suggestion_3_update = gr.update(value=suggestions[2])
                        logging.info(f"Apertura pannello CHAT per {plot_id_clicked} ('{clicked_plot_label}')")

                    elif action_type == "formula":
                        formula_display_visible_update = gr.update(visible=True)
                        formula_key = PLOT_ID_TO_FORMULA_KEY_MAP.get(plot_id_clicked)
                        formula_text = f"**Formula/Metodologia per: {clicked_plot_label}**\n\nID Grafico: `{plot_id_clicked}`.\n\n"
                        if formula_key and formula_key in PLOT_FORMULAS:
                            formula_data = PLOT_FORMULAS[formula_key]
                            formula_text += f"### {formula_data['title']}\n\n"
                            formula_text += f"**Descrizione:**\n{formula_data['description']}\n\n"
                            formula_text += "**Come viene calcolato:**\n"
                            for step in formula_data['calculation_steps']:
                                formula_text += f"- {step}\n"
                        else:
                            formula_text += "(Nessuna informazione dettagliata sulla formula trovata per questo ID grafico in `formulas.py`)"
                        formula_content_update = gr.update(value=formula_text)
                        new_current_chat_plot_id = None 
                        logging.info(f"Apertura pannello FORMULA per {plot_id_clicked} (mappato a {formula_key})")
                
                all_button_icon_updates = []
                for cfg_item in plot_configs:
                    p_id_iter = cfg_item["id"]
                    # Update insights button icon
                    if new_active_action_state_to_set == {"plot_id": p_id_iter, "type": "insights"}:
                        all_button_icon_updates.append(gr.update(value=ACTIVE_ICON))
                    else:
                        all_button_icon_updates.append(gr.update(value=BOMB_ICON))
                    # Update formula button icon
                    if new_active_action_state_to_set == {"plot_id": p_id_iter, "type": "formula"}:
                        all_button_icon_updates.append(gr.update(value=ACTIVE_ICON))
                    else:
                        all_button_icon_updates.append(gr.update(value=FORMULA_ICON))
                
                final_updates = [
                    action_col_visible_update,
                    insights_chatbot_visible_update, chatbot_content_update,
                    insights_chat_input_visible_update,
                    insights_suggestions_row_visible_update, suggestion_1_update, suggestion_2_update, suggestion_3_update,
                    formula_display_visible_update, formula_content_update,
                    new_active_action_state_to_set, 
                    new_current_chat_plot_id,    
                    updated_chat_histories     
                ] + all_button_icon_updates
                
                return final_updates

            async def handle_chat_message_submission(
                user_message: str, 
                current_plot_id: str, 
                chat_histories: dict, 
                current_plot_data_for_chatbot: dict # NEW: data summaries
            ):
                if not current_plot_id or not user_message.strip():
                    history_for_plot = chat_histories.get(current_plot_id, [])
                    # Yield current state if no action needed
                    yield history_for_plot, gr.update(value=""), chat_histories # Clear input, return current history
                    return

                plot_config = next((p for p in plot_configs if p["id"] == current_plot_id), None)
                plot_label = plot_config["label"] if plot_config else "Grafico Selezionato"
                
                # Retrieve the specific data summary for the current plot
                plot_specific_data_summary = current_plot_data_for_chatbot.get(current_plot_id, f"Nessun sommario dati specifico disponibile per '{plot_label}'.")


                history_for_plot = chat_histories.get(current_plot_id, []).copy()
                history_for_plot.append({"role": "user", "content": user_message})
                
                # Update UI immediately with user message
                yield history_for_plot, gr.update(value=""), chat_histories # Clear input

                # Pass the data summary to the LLM along with the history
                bot_response_text = await generate_llm_response(
                    user_message, 
                    current_plot_id, 
                    plot_label, 
                    history_for_plot, # This history now includes the initial insight with summary + user message
                    plot_specific_data_summary # Explicitly pass for this turn if needed by LLM handler logic
                )
                
                history_for_plot.append({"role": "assistant", "content": bot_response_text})
                
                updated_chat_histories = chat_histories.copy()
                updated_chat_histories[current_plot_id] = history_for_plot
                
                yield history_for_plot, "", updated_chat_histories


            async def handle_suggested_question_click(
                suggestion_text: str,
                current_plot_id: str, 
                chat_histories: dict, 
                current_plot_data_for_chatbot: dict # NEW: data summaries
            ):
                if not current_plot_id or not suggestion_text.strip():
                    history_for_plot = chat_histories.get(current_plot_id, [])
                    yield history_for_plot, gr.update(value=""), chat_histories 
                    return

                # This is essentially the same as submitting a message, so reuse logic
                # The suggestion_text becomes the user_message
                async for update in handle_chat_message_submission(
                    suggestion_text, 
                    current_plot_id, 
                    chat_histories, 
                    current_plot_data_for_chatbot
                ):
                    yield update


            def handle_explore_click(plot_id_clicked, current_explored_plot_id_from_state):
                logging.info(f"Click su Esplora per: {plot_id_clicked}. Attualmente esplorato da stato: {current_explored_plot_id_from_state}")
                if not plot_ui_objects:
                    logging.error("plot_ui_objects non popolato durante handle_explore_click.")
                    updates_for_missing_ui = [current_explored_plot_id_from_state] 
                    for _ in plot_configs: # panel_component, explore_button
                        updates_for_missing_ui.extend([gr.update(), gr.update()]) 
                    return updates_for_missing_ui
                
                new_explored_id_to_set = None
                is_toggling_off = (plot_id_clicked == current_explored_plot_id_from_state)
                
                if is_toggling_off:
                    new_explored_id_to_set = None
                    logging.info(f"Interruzione esplorazione grafico: {plot_id_clicked}")
                else:
                    new_explored_id_to_set = plot_id_clicked
                    logging.info(f"Esplorazione grafico: {plot_id_clicked}")
                                        
                panel_and_button_updates = []
                for cfg in plot_configs:
                    p_id = cfg["id"]
                    if p_id in plot_ui_objects:
                        panel_visible = not new_explored_id_to_set or (p_id == new_explored_id_to_set)
                        panel_and_button_updates.append(gr.update(visible=panel_visible)) 
                                                
                        if p_id == new_explored_id_to_set: 
                            panel_and_button_updates.append(gr.update(value=ACTIVE_ICON))
                        else:
                            panel_and_button_updates.append(gr.update(value=EXPLORE_ICON))
                    else: 
                        panel_and_button_updates.extend([gr.update(), gr.update()])
                                        
                final_updates = [new_explored_id_to_set] + panel_and_button_updates 
                return final_updates
            
            # Outputs for panel actions
            action_panel_outputs_list = [
                global_actions_column_ui, 
                insights_chatbot_ui, insights_chatbot_ui, # Target chatbot UI for visibility and value
                insights_chat_input_ui, 
                insights_suggestions_row_ui, insights_suggestion_1_btn, insights_suggestion_2_btn, insights_suggestion_3_btn, 
                formula_display_markdown_ui, formula_display_markdown_ui, # Target markdown for visibility and value
                active_panel_action_state, 
                current_chat_plot_id_st,
                chat_histories_st
            ]
            for cfg_item_action in plot_configs: 
                pid_action = cfg_item_action["id"]
                if pid_action in plot_ui_objects:
                    action_panel_outputs_list.append(plot_ui_objects[pid_action]["bomb_button"])
                    action_panel_outputs_list.append(plot_ui_objects[pid_action]["formula_button"])
                else: 
                    action_panel_outputs_list.extend([None, None]) 

            # Outputs for explore actions
            explore_buttons_outputs_list = [explored_plot_id_state]
            for cfg_item_explore in plot_configs:
                pid_explore = cfg_item_explore["id"]
                if pid_explore in plot_ui_objects:
                    explore_buttons_outputs_list.append(plot_ui_objects[pid_explore]["panel_component"])
                    explore_buttons_outputs_list.append(plot_ui_objects[pid_explore]["explore_button"])
                else: 
                    explore_buttons_outputs_list.extend([None, None])

            # Inputs for panel actions
            action_click_inputs = [
                active_panel_action_state, 
                chat_histories_st,
                current_chat_plot_id_st,
                plot_data_for_chatbot_st # NEW: pass data summaries state
            ]
            # Inputs for explore actions
            explore_click_inputs = [explored_plot_id_state]

            def create_panel_action_handler(p_id, action_type_str):
                async def _handler(current_active_val, current_chats_val, current_chat_pid, current_plot_data_summaries): # Add summaries
                    logging.debug(f"Entering _handler for plot_id: {p_id}, action: {action_type_str}")
                    result = await handle_panel_action(p_id, action_type_str, current_active_val, current_chats_val, current_chat_pid, current_plot_data_summaries) # Pass summaries
                    logging.debug(f"_handler for plot_id: {p_id}, action: {action_type_str} completed.")
                    return result
                return _handler

            for config_item in plot_configs:
                plot_id = config_item["id"]
                if plot_id in plot_ui_objects:
                    ui_obj = plot_ui_objects[plot_id]
                    
                    ui_obj["bomb_button"].click(
                        fn=create_panel_action_handler(plot_id, "insights"), 
                        inputs=action_click_inputs, 
                        outputs=action_panel_outputs_list,
                        api_name=f"action_insights_{plot_id}"
                    )
                    ui_obj["formula_button"].click(
                        fn=create_panel_action_handler(plot_id, "formula"), 
                        inputs=action_click_inputs,
                        outputs=action_panel_outputs_list,
                        api_name=f"action_formula_{plot_id}"
                    )
                    ui_obj["explore_button"].click(
                        fn=lambda current_explored_val, p_id=plot_id: handle_explore_click(p_id, current_explored_val),
                        inputs=explore_click_inputs,
                        outputs=explore_buttons_outputs_list,
                        api_name=f"action_explore_{plot_id}"
                    )
                else:
                    logging.warning(f"Oggetto UI per plot_id '{plot_id}' non trovato durante il tentativo di associare i gestori di click.")
            
            chat_submission_outputs = [insights_chatbot_ui, insights_chat_input_ui, chat_histories_st]
            chat_submission_inputs = [insights_chat_input_ui, current_chat_plot_id_st, chat_histories_st, plot_data_for_chatbot_st] # Add data summaries state

            insights_chat_input_ui.submit(
                fn=handle_chat_message_submission,
                inputs=chat_submission_inputs, 
                outputs=chat_submission_outputs,
                api_name="submit_chat_message"
            )

            suggestion_click_inputs = [current_chat_plot_id_st, chat_histories_st, plot_data_for_chatbot_st] # Add data summaries state
            insights_suggestion_1_btn.click(
                fn=handle_suggested_question_click,
                inputs=[insights_suggestion_1_btn] + suggestion_click_inputs, # Pass button value as first arg
                outputs=chat_submission_outputs, 
                api_name="click_suggestion_1"
            )
            insights_suggestion_2_btn.click(
                fn=handle_suggested_question_click,
                inputs=[insights_suggestion_2_btn] + suggestion_click_inputs, 
                outputs=chat_submission_outputs,
                api_name="click_suggestion_2"
            )
            insights_suggestion_3_btn.click(
                fn=handle_suggested_question_click,
                inputs=[insights_suggestion_3_btn] + suggestion_click_inputs, 
                outputs=chat_submission_outputs,
                api_name="click_suggestion_3"
            )

            def refresh_all_analytics_ui_elements(current_token_state, date_filter_val, custom_start_val, custom_end_val, current_chat_histories):
                logging.info("Aggiornamento di tutti gli elementi UI delle analisi e reset delle azioni/chat.")
                
                # Pass plot_configs to the update function so it can be used by generate_chatbot_data_summaries
                plot_generation_results = update_analytics_plots_figures(
                    current_token_state, date_filter_val, custom_start_val, custom_end_val, plot_configs
                )
                status_message_update = plot_generation_results[0]
                generated_plot_figures = plot_generation_results[1:-1] # All items except first (status) and last (summaries)
                new_plot_data_summaries = plot_generation_results[-1] # Last item is the summaries dict
                
                all_updates = [status_message_update] 
                
                for i in range(len(plot_configs)): 
                    if i < len(generated_plot_figures):
                        all_updates.append(generated_plot_figures[i]) 
                    else: 
                        all_updates.append(create_placeholder_plot("Errore Figura", f"Figura mancante per grafico {plot_configs[i]['id']}"))
                
                all_updates.extend([
                    gr.update(visible=False), # global_actions_column_ui
                    gr.update(value=[], visible=False), # insights_chatbot_ui (value)
                    gr.update(value="", visible=False), # insights_chat_input_ui (value)
                    gr.update(visible=False), # insights_suggestions_row_ui
                    gr.update(value="Suggerimento 1", visible=True), # insights_suggestion_1_btn
                    gr.update(value="Suggerimento 2", visible=True), # insights_suggestion_2_btn
                    gr.update(value="Suggerimento 3", visible=True), # insights_suggestion_3_btn
                    gr.update(value="I dettagli sulla formula/metodologia appariranno qui.", visible=False), # formula_display_markdown_ui
                    None, # active_panel_action_state
                    None, # current_chat_plot_id_st
                    current_chat_histories, # chat_histories_st (preserve or reset as needed, here preserving)
                    new_plot_data_summaries # NEW: plot_data_for_chatbot_st
                ])
                
                for cfg in plot_configs: 
                    pid = cfg["id"]
                    if pid in plot_ui_objects:
                        all_updates.append(gr.update(value=BOMB_ICON))   
                        all_updates.append(gr.update(value=FORMULA_ICON)) 
                        all_updates.append(gr.update(value=EXPLORE_ICON)) 
                        all_updates.append(gr.update(visible=True))      # panel_component visibility
                    else: 
                        all_updates.extend([None, None, None, None])
                
                all_updates.append(None) # explored_plot_id_state
                
                logging.info(f"Preparati {len(all_updates)} aggiornamenti per il refresh delle analisi.")
                return all_updates

            apply_filter_and_sync_outputs_list = [analytics_status_md] 
            for config_item_filter_sync in plot_configs:
                pid_filter_sync = config_item_filter_sync["id"]
                if pid_filter_sync in plot_ui_objects and "plot_component" in plot_ui_objects[pid_filter_sync]:
                    apply_filter_and_sync_outputs_list.append(plot_ui_objects[pid_filter_sync]["plot_component"])
                else:
                    apply_filter_and_sync_outputs_list.append(None) 
            
            apply_filter_and_sync_outputs_list.extend([
                global_actions_column_ui,         # Reset visibility
                insights_chatbot_ui,              # Reset content & visibility
                insights_chat_input_ui,           # Reset content & visibility
                insights_suggestions_row_ui,      # Reset visibility
                insights_suggestion_1_btn,        # Reset text & visibility
                insights_suggestion_2_btn,
                insights_suggestion_3_btn,
                formula_display_markdown_ui,      # Reset content & visibility
                active_panel_action_state,        # Reset state
                current_chat_plot_id_st,          # Reset state
                chat_histories_st,                # Preserve or reset state
                plot_data_for_chatbot_st          # NEW: Update this state
            ])

            for cfg_filter_sync_btns in plot_configs:
                pid_filter_sync_btns = cfg_filter_sync_btns["id"]
                if pid_filter_sync_btns in plot_ui_objects:
                    apply_filter_and_sync_outputs_list.append(plot_ui_objects[pid_filter_sync_btns]["bomb_button"])
                    apply_filter_and_sync_outputs_list.append(plot_ui_objects[pid_filter_sync_btns]["formula_button"])
                    apply_filter_and_sync_outputs_list.append(plot_ui_objects[pid_filter_sync_btns]["explore_button"])
                    apply_filter_and_sync_outputs_list.append(plot_ui_objects[pid_filter_sync_btns]["panel_component"]) 
                else:
                    apply_filter_and_sync_outputs_list.extend([None, None, None, None]) 
            
            apply_filter_and_sync_outputs_list.append(explored_plot_id_state) # Reset state
            
            logging.info(f"Output totali definiti per apply_filter/sync: {len(apply_filter_and_sync_outputs_list)}")
            
            apply_filter_btn.click(
                fn=refresh_all_analytics_ui_elements,
                inputs=[token_state, date_filter_selector, custom_start_date_picker, custom_end_date_picker, chat_histories_st],
                outputs=apply_filter_and_sync_outputs_list,
                show_progress="full"
            )

        with gr.TabItem("3️⃣ Menzioni", id="tab_mentions"):
            refresh_mentions_display_btn = gr.Button("🔄 Aggiorna Visualizzazione Menzioni", variant="secondary")
            mentions_html = gr.HTML("Dati menzioni...")
            mentions_sentiment_dist_plot = gr.Plot(label="Distribuzione Sentiment Menzioni")
            refresh_mentions_display_btn.click(
                fn=run_mentions_tab_display, inputs=[token_state],
                outputs=[mentions_html, mentions_sentiment_dist_plot],
                show_progress="full"
            )

        with gr.TabItem("4️⃣ Statistiche Follower", id="tab_follower_stats"):
            refresh_follower_stats_btn = gr.Button("🔄 Aggiorna Visualizzazione Statistiche Follower", variant="secondary")
            follower_stats_html = gr.HTML("Statistiche follower...")
            with gr.Row(): 
                fs_plot_monthly_gains = gr.Plot(label="Guadagni Mensili Follower")
            with gr.Row():
                fs_plot_seniority = gr.Plot(label="Follower per Anzianità (Top 10 Organici)")
                fs_plot_industry = gr.Plot(label="Follower per Settore (Top 10 Organici)")
            
            refresh_follower_stats_btn.click(
                fn=run_follower_stats_tab_display, inputs=[token_state],
                outputs=[follower_stats_html, fs_plot_monthly_gains, fs_plot_seniority, fs_plot_industry],
                show_progress="full"
            )
            
    sync_event_part1 = sync_data_btn.click(
        fn=sync_all_linkedin_data_orchestrator,
        inputs=[token_state], outputs=[sync_status_html_output, token_state], show_progress="full"
    )
    sync_event_part2 = sync_event_part1.then( 
        fn=process_and_store_bubble_token, 
        inputs=[url_user_token_display, org_urn_display, token_state],
        outputs=[status_box, token_state, sync_data_btn], show_progress=False 
    )
    sync_event_part3 = sync_event_part2.then(
        fn=display_main_dashboard, 
        inputs=[token_state], outputs=[dashboard_display_html], show_progress=False
    )
    sync_event_final = sync_event_part3.then(
        fn=refresh_all_analytics_ui_elements, # This will now also update chatbot data summaries
        inputs=[token_state, date_filter_selector, custom_start_date_picker, custom_end_date_picker, chat_histories_st],
        outputs=apply_filter_and_sync_outputs_list, 
        show_progress="full"
    )

if __name__ == "__main__":
    if not os.environ.get(LINKEDIN_CLIENT_ID_ENV_VAR):
        logging.warning(f"ATTENZIONE: Variabile d'ambiente '{LINKEDIN_CLIENT_ID_ENV_VAR}' non impostata.")
    if not os.environ.get(BUBBLE_APP_NAME_ENV_VAR) or \
       not os.environ.get(BUBBLE_API_KEY_PRIVATE_ENV_VAR) or \
       not os.environ.get(BUBBLE_API_ENDPOINT_ENV_VAR):
        logging.warning("ATTENZIONE: Variabili d'ambiente Bubble non completamente impostate.")
    
    try:
        logging.info(f"Versione Matplotlib: {matplotlib.__version__}, Backend: {matplotlib.get_backend()}")
    except ImportError: 
        logging.warning("Matplotlib non trovato direttamente, ma potrebbe essere usato dai generatori di grafici.")
        
    app.launch(server_name="0.0.0.0", server_port=7860, debug=True)