import logging
from concurrent.futures import ThreadPoolExecutor
from datetime import datetime, timedelta
from typing import Dict, List, Optional, Any, Tuple
import gradio as gr
import pandas as pd
import requests
from dataclasses import dataclass
from tenacity import retry, stop_after_attempt, wait_fixed, wait_exponential
import plotly.express as px
import plotly.graph_objects as go
import os
import json
import urllib.parse  # IMPORTANT: For joining URLs

# Configuration du logging
logging.basicConfig(
    level=logging.INFO,
    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
    handlers=[
        logging.FileHandler("pesticide_app.log"),
        logging.StreamHandler()
    ]
)
logger = logging.getLogger(__name__)

@dataclass
class SubstanceDetails:
    name: str
    substance_id: int
    status: Optional[str] = None
    approval_date: Optional[str] = None
    expiry_date: Optional[str] = None
    cas_number: Optional[str] = None
    ec_number: Optional[str] = None

class PesticideDataFetcher:
    BASE_URL = "https://api.datalake.sante.service.ec.europa.eu/sante/pesticides"
    HEADERS = {
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36',
        'Accept': 'application/json',
        'Accept-Language': 'fr-FR,fr;q=0.9,en-US;q=0.8,en;q=0.7',  # Keeping this, but it might not be needed for /active_substances
        "Content-Type": "application/json",
        "Cache-Control": "no-cache"
    }
    CACHE_DIR = "cache"
    SUBSTANCE_CACHE_FILE = os.path.join(CACHE_DIR, "substance_cache.json")
    PRODUCT_CACHE_FILE = os.path.join(CACHE_DIR, "product_cache.json")
    MRL_CACHE_FILE = os.path.join(CACHE_DIR, "mrl_cache.json")

    def __init__(self, use_cache: bool = True):
        self.session = requests.Session()
        self.session.headers.update(self.HEADERS)
        self._substance_cache: Dict[int, SubstanceDetails] = {}
        self._product_cache: Dict[int, Dict[str, Any]] = {}
        self._mrl_cache: Dict[int, List[Dict[str, Any]]] = {}
        self.use_cache = use_cache

        if not os.path.exists(self.CACHE_DIR):
            os.makedirs(self.CACHE_DIR)

        if use_cache:
            self._load_caches()

        # Preload substances only if cache is empty
        if not self._substance_cache:
            self.preload_substance_names()


    def _load_caches(self):
        """Loads cached data from files."""
        try:
            if os.path.exists(self.SUBSTANCE_CACHE_FILE):
                with open(self.SUBSTANCE_CACHE_FILE, 'r', encoding='utf-8') as f:
                    substance_data = json.load(f)
                    self._substance_cache = {int(k): SubstanceDetails(**v) for k, v in substance_data.items()}
                logger.info(f"Loaded substance cache: {len(self._substance_cache)} substances")

            if os.path.exists(self.PRODUCT_CACHE_FILE):
                with open(self.PRODUCT_CACHE_FILE, 'r', encoding='utf-8') as f:
                    self._product_cache = {int(k): v for k, v in json.load(f).items()}
                logger.info(f"Loaded product cache: {len(self._product_cache)} products")

            if os.path.exists(self.MRL_CACHE_FILE):
                with open(self.MRL_CACHE_FILE, 'r', encoding='utf-8') as f:
                    self._mrl_cache = {int(k): v for k, v in json.load(f).items()}
                logger.info(f"Loaded MRL cache: {len(self._mrl_cache)} entries")

        except Exception as e:
            logger.error(f"Error loading caches: {e}")
            self._substance_cache = {}
            self._product_cache = {}
            self._mrl_cache = {}

    def _save_caches(self):
        """Saves cached data to files."""
        try:
            with open(self.SUBSTANCE_CACHE_FILE, 'w', encoding='utf-8') as f:
                substance_data = {str(k): v.__dict__ for k, v in self._substance_cache.items()}
                json.dump(substance_data, f, ensure_ascii=False, indent=2)

            with open(self.PRODUCT_CACHE_FILE, 'w', encoding='utf-8') as f:
                json.dump({str(k): v for k, v in self._product_cache.items()}, f, ensure_ascii=False, indent=2)

            with open(self.MRL_CACHE_FILE, 'w', encoding='utf-8') as f:
                json.dump({str(k): v for k, v in self._mrl_cache.items()}, f, ensure_ascii=False, indent=2)

            logger.info("All caches saved successfully")
        except Exception as e:
            logger.error(f"Error saving caches: {e}")

    @retry(stop=stop_after_attempt(3), wait=wait_exponential(multiplier=1, min=2, max=10))
    def fetch_data(self, url: str, params: Optional[Dict[str, Any]] = None) -> Optional[Dict[str, Any]]:
        """Fetches data from the API with retry mechanism and optional parameters."""
        try:
            response = self.session.get(url, params=params, timeout=15)
            response.raise_for_status()
            data = response.json()
            logger.debug(f"API response for {url}: {str(data)[:200]}...")
            return data
        except requests.RequestException as e:
            logger.error(f"Request failed for {url}: {e}")
            return None  # Return None on failure


    def preload_substance_names(self) -> None:
        """Preloads substance names from the API."""
        logger.info("Preloading substances...")
        url = f"{self.BASE_URL}/active_substances"  # Correct endpoint
        params = {"format": "json", "api-version": "v2.0"}  # Correct parameters, NO LANGUAGE
        substances_loaded = 0

        data = self.fetch_data(url, params=params)
        if not data or "value" not in data:
            logger.warning("Initial fetch returned no data or unexpected format. Skipping preloading.")
            return

        while url:
            if "error" in data:
                logger.error(f"Error preloading substances: {data.get('error', 'Unknown error')}")
                break

            for substance in data.get("value", []):
                substance_id = substance.get("substanceId") # Corrected Key
                if substance_id and substance_id not in self._substance_cache:
                    self._substance_cache[substance_id] = SubstanceDetails(
                        name=substance.get("substanceName", "Unknown"), # Corrected Key
                        substance_id=substance_id,
                        status=substance.get("status"),
                        approval_date=substance.get("approvalDate"),
                        expiry_date=substance.get("expiryDate"),
                        cas_number=substance.get("casNumber"),
                        ec_number=substance.get("ecNumber")
                    )
                    substances_loaded += 1
                    # Removed _load_substance_details

            next_link = data.get("nextLink")
            if next_link:
                # Correctly join the BASE_URL with the relative nextLink
                url = urllib.parse.urljoin(self.BASE_URL, next_link)
                params = None  # IMPORTANT: Don't send params again!
                data = self.fetch_data(url)  # Fetch the next page of data.  NO PARAMS.
                if not data:
                    break # Exit loop if data fetch fail
            else:
                url = None
            logger.info(f"Substances loaded so far: {substances_loaded}")


        if self.use_cache:
            self._save_caches()

        logger.info(f"Preloading complete.  Total substances: {len(self._substance_cache)}")


    def get_substance_name(self, substance_id: int) -> str:
        """Retrieves the name of a substance by its ID."""
        # First, check the cache
        if substance_id in self._substance_cache:
            return self._substance_cache[substance_id].name

        # If not in cache, fetch from API and use v2.0
        url = f"{self.BASE_URL}/active_substances/{substance_id}" # Correct endpoint
        params = {"format": "json", "api-version": "v2.0"} # Use v2.0

        data = self.fetch_data(url, params=params)
        if not data:
             return f"Unknown Substance ({substance_id})"

        # Use correct keys and expect a single object, not a list
        substance_name = data.get("substanceName", f"Substance {substance_id}")
        # Update the cache.  Since we request details, get ALL details
        self._substance_cache[substance_id] = SubstanceDetails(
            name=substance_name,
            substance_id=substance_id,
            status=data.get("status"),
            approval_date=data.get("approvalDate"),
            expiry_date=data.get("expiryDate"),
            cas_number=data.get("casNumber"),
            ec_number=data.get("ecNumber")
        )
        if self.use_cache:
                self._save_caches()
        return substance_name


    def get_product_list(self) -> List[Dict[str, Any]]:
        """Retrieves the list of all products."""
        if self._product_cache:  # Check the cache first
            return list(self._product_cache.values())

        logger.info("Retrieving product list...")
        url = f"{self.BASE_URL}/pesticide_residues_products"  # Correct endpoint
        params = {"format": "json", "language": "FR", "api-version": "v2.0"}  # Correct params
        products_loaded = 0
        all_products = []

        while url:
            data = self.fetch_data(url, params=params)
            if not data:
                logger.warning("No data returned for product list.")
                break
            if "error" in data:
                logger.error(f"Error retrieving products: {data.get('error', 'No info')}")
                break

            # Handle both list and dict responses
            if isinstance(data, list):
                products = data
            elif isinstance(data, dict) and "value" in data:
                products = data["value"]
            else:
                logger.error(f"Unexpected API response format: {data}")
                break

            for product in products:
                product_id = product.get("productId")
                if product_id:
                    self._product_cache[product_id] = product
                    products_loaded += 1
                    all_products.append(product)

            next_link = data.get("nextLink") if isinstance(data, dict) else None #Get nextLink
            #Corrected nextlink verification
            if next_link:
                # Correctly join the BASE_URL with the relative nextLink
                url = urllib.parse.urljoin(self.BASE_URL, next_link)
                params = None  # IMPORTANT: Don't send params again!  The nextLink ALREADY contains them.
            else:
                url = None
            logger.info(f"Products retrieved so far: {products_loaded}")

        if self.use_cache:
            self._save_caches()

        logger.info(f"Total products retrieved: {len(self._product_cache)}")
        return all_products



    def get_mrls(self, product_id: int) -> List[Dict[str, Any]]:
        """Retrieves MRLs for a specific product."""
        if product_id in self._mrl_cache:
            return self._mrl_cache[product_id]

        logger.info(f"Retrieving MRLs for product {product_id}...")
        url = f"{self.BASE_URL}/pesticide_residues_products/{product_id}/mrls" #Correct endpoint
        params = {"format": "json", "language": "FR", "api-version": "v2.0"} # Use v2.0 and language

        mrls = []
        while url:
            data = self.fetch_data(url, params=params)
            if not data:
                break
            if "error" in data:
                logger.error(f"Error retrieving MRLs: {data.get('error', 'No info')}")
                break
            mrls.extend(data.get("value", []))
            next_link = data.get("nextLink")
            if next_link:
                url = urllib.parse.urljoin(self.BASE_URL, next_link)
                params = None
            else:
                url = None


        self._mrl_cache[product_id] = mrls
        if self.use_cache:
            self._save_caches()
        logger.info(f"MRLs retrieved for product {product_id}: {len(mrls)}")
        return mrls

    def search_substances(self, query: str) -> List[SubstanceDetails]:
        """Searches for substances by name."""
        query = query.lower()
        return sorted([
            substance for substance in self._substance_cache.values()
            if query in substance.name.lower()
        ], key=lambda x: x.name)

    def get_substance_mrls(self, substance_id: int) -> List[Dict[str, Any]]:
        """Retrieves all products with MRLs for a given substance."""
        logger.info(f"Retrieving MRLs for substance {substance_id}...")
        url = f"{self.BASE_URL}/pesticide_residues_mrls"  # Correct endpoint
        params = {"format": "json", "api-version": "v2.0", "pesticide_residue_id": substance_id}  # Correct params

        all_mrls = []
        while url:
            data = self.fetch_data(url, params=params)
            if not data:
                break
            if "error" in data:
                logger.error(f"Error retrieving MRLs: {data.get('error', 'No info')}")
                break

            all_mrls.extend(data.get("value", []))
            next_link = data.get("nextLink")
            if next_link:
                url = urllib.parse.urljoin(self.BASE_URL, next_link)
                params = None
            else:
                url = None
        logger.info(f"MRLs retrieved for substance {substance_id}: {len(all_mrls)}")
        return all_mrls
    def format_date(self, date_str: Optional[str]) -> str:
        """Formats an ISO date string into a readable format."""
        if not date_str:
            return "N/C"
        try:
            dt = datetime.fromisoformat(date_str.replace('Z', '+00:00'))
            return dt.strftime("%d/%m/%Y")
        except (ValueError, TypeError):
            return date_str

class PesticideApp:
    def __init__(self, use_cache: bool = True):
        logger.info("Initializing application...")
        self.fetcher = PesticideDataFetcher(use_cache=use_cache)

        logger.info("Retrieving product list...")
        products = self.fetcher.get_product_list()
        # Handle empty product list gracefully
        if products:
            self.product_list = {
                p.get('productName', 'Unnamed'): p.get('productId', 0)
                for p in products
            }
        else:
            self.product_list = {'No products found': 0}  # Provide a default
            logger.warning("No products found. Product dropdown will be empty.")

        self.substances = sorted([
            sd.name for sd in self.fetcher._substance_cache.values()
        ])
        logger.info(f"Application initialized with {len(self.product_list)} products and {len(self.substances)} substances")

    def format_date(self, date_str: Optional[str]) -> str:
        """Formats an ISO date string."""
        return self.fetcher.format_date(date_str)

    def get_product_details(self, product_name: str) -> pd.DataFrame:
        """Retrieves details for a product and returns a DataFrame."""
        logger.info(f"Retrieving details for product: {product_name}")
        product_id = self.product_list.get(product_name)
        if not product_id:
            return pd.DataFrame([{"Error": "Product not found"}])

        mrls = self.fetcher.get_mrls(product_id)
        if not mrls:
            return pd.DataFrame([{"Error": "No MRL data found"}])

        data = []
        for mrl in mrls:
            substance_id = mrl.get("pesticideResidueId", 0)
            substance_name = self.fetcher.get_substance_name(substance_id)
            substance = self.fetcher._substance_cache.get(substance_id)

            data.append({
                "Substance": substance_name,
                "Valeur LMR": mrl.get("mrlValue", "N/C"),
                "Unité": mrl.get("mrlUnit", "mg/kg"),
                "Date d'effet": self.format_date(mrl.get("entryIntoForceDate")),
                "Statut": getattr(substance, "status", "N/C"),
                "CAS": getattr(substance, "cas_number", "N/C"),
                "EC": getattr(substance, "ec_number", "N/C"),
                "Date d'approbation": self.format_date(getattr(substance, "approval_date", None)),
                "Date d'expiration": self.format_date(getattr(substance, "expiry_date", None))
            })

        df = pd.DataFrame(data)
        logger.info(f"Details retrieved for {product_name}: {len(df)} entries")
        return df


    def search_substance(self, substance_query: str) -> pd.DataFrame:
        """Searches for substances by name."""
        logger.info(f"Searching for substances: {substance_query}")
        if not substance_query or len(substance_query) < 3:
            return pd.DataFrame([{"Message": "Please enter at least 3 characters for the search"}])

        results = self.fetcher.search_substances(substance_query)
        if not results:
            return pd.DataFrame([{"Message": "No substances found"}])

        data = [{
            "ID": substance.substance_id,
            "Nom": substance.name,
            "Statut": substance.status or "N/C",
            "N° CAS": substance.cas_number or "N/C",
            "N° EC": substance.ec_number or "N/C",
            "Date d'approbation": self.format_date(substance.approval_date),
            "Date d'expiration": self.format_date(substance.expiry_date)
        } for substance in results]

        df = pd.DataFrame(data)
        logger.info(f"Search results for '{substance_query}': {len(df)} substances found")
        return df

    def get_substance_mrls(self, substance_id: int) -> pd.DataFrame:
        """Retrieves all products with MRLs for a given substance and returns a DataFrame."""
        logger.info(f"Retrieving MRLs for substance ID: {substance_id}")
        if not substance_id:
            return pd.DataFrame([{"Error": "Invalid substance ID"}])

        substance = self.fetcher._substance_cache.get(substance_id)
        if not substance:
            return pd.DataFrame([{"Error": "Substance not found"}])

        all_mrls = self.fetcher.get_substance_mrls(substance_id)
        if not all_mrls:
            return pd.DataFrame([{"Message": f"No MRLs found for {substance.name}"}])

        data = []
        for mrl in all_mrls:
            product_id = mrl.get("productId")
            product_name = "Unknown"
            if product_id in self.fetcher._product_cache:
                product_name = self.fetcher._product_cache[product_id].get("productName", "Unknown")

            data.append({
                "Produit": product_name,
                "Valeur LMR": mrl.get("mrlValue", "N/C"),
                "Unité": mrl.get("mrlUnit", "mg/kg"),
                "Date d'effet": self.format_date(mrl.get("entryIntoForceDate")),
                "Notes": mrl.get("footnotes", "")
            })

        df = pd.DataFrame(data)
        logger.info(f"MRLs retrieved for {substance.name}: {len(df)} entries")
        return df

    def create_histogram(self, df: pd.DataFrame) -> go.Figure:
        """Creates a histogram of MRL values."""
        if "Valeur LMR" not in df.columns or df.empty:
            return go.Figure(data=[], layout=go.Layout(title="No data available for histogram"))

        numeric_values = []
        for val in df["Valeur LMR"]:
            try:
                if isinstance(val, str):
                    if val not in ("*", "LOQ", "N/C"):
                        numeric_values.append(float(val.replace(",", ".")))
                else:
                    numeric_values.append(float(val))
            except (ValueError, TypeError):
                continue

        if not numeric_values:
            return go.Figure(data=[], layout=go.Layout(title="No numeric values available for histogram"))

        fig = go.Figure(data=[go.Histogram(x=numeric_values, nbinsx=20)])
        fig.update_layout(
            title="Distribution of MRL Values",
            xaxis_title="MRL Value (mg/kg)",
            yaxis_title="Frequency",
            bargap=0.1
        )
        return fig

    def create_pie_chart(self, df: pd.DataFrame, column: str) -> go.Figure:
        """Creates a pie chart for a given column."""
        if column not in df.columns or df.empty:
            return go.Figure(data=[], layout=go.Layout(title=f"No data available for {column}"))

        value_counts = df[column].value_counts()
        fig = go.Figure(data=[go.Pie(labels=value_counts.index, values=value_counts.values)])
        fig.update_layout(title=f"Distribution by {column}")
        return fig

    def create_ui(self) -> gr.Blocks:
        """Creates the user interface with Gradio."""
        with gr.Blocks(theme=gr.themes.Soft()) as ui:
            gr.HTML("""
                <div style="text-align: center; padding: 20px; background: #006633; color: white;">
                    <h1>EU Pesticide Database</h1>
                    <p>View Maximum Residue Limits (MRLs) of pesticides in food products.</p>
                </div>
            """)

            with gr.Tab("Search by Product"):
                with gr.Row():
                    with gr.Column(scale=3):
                        product = gr.Dropdown(
                            label="Select a food product",
                            choices=sorted(self.product_list.keys()),
                            interactive=True
                        )
                    with gr.Column(scale=1):
                        search_btn = gr.Button("Search", variant="primary")

                with gr.Row():
                    output = gr.Dataframe(
                        headers=["Substance", "Valeur LMR", "Unité", "Date d'effet", "Statut"],
                        row_count=(20, "fixed"),
                        interactive=False
                    )

                with gr.Row():
                    with gr.Column():
                        histogram = gr.Plot(label="MRL Distribution")
                    with gr.Column():
                        status_pie = gr.Plot(label="Substance Status")

                search_btn.click(
                    fn=lambda p: (
                        self.get_product_details(p),
                        self.create_histogram(self.get_product_details(p)),
                        self.create_pie_chart(self.get_product_details(p), "Statut")
                    ) if p else (pd.DataFrame(), go.Figure(), go.Figure()),  # Handle None selection
                    inputs=[product],
                    outputs=[output, histogram, status_pie]
                )

            with gr.Tab("Search by Substance"):
                with gr.Row():
                    with gr.Column(scale=3):
                        substance_query = gr.Textbox(
                            label="Search for a substance (min. 3 characters)",
                            placeholder="Ex: glyphosate, copper, etc.",
                            interactive=True
                        )
                    with gr.Column(scale=1):
                        substance_search_btn = gr.Button("Search", variant="primary")

                substance_results = gr.Dataframe(
                    headers=["ID", "Nom", "Statut", "N° CAS", "N° EC", "Date d'approbation", "Date d'expiration"],
                    row_count=(15, "fixed"),
                    interactive=True
                )

                substance_select = gr.Number(label="Selected Substance ID", interactive=True, visible=False)
                substance_mrls_btn = gr.Button("View MRLs for this substance", visible=True)

                substance_mrls = gr.Dataframe(
                    headers=["Produit", "Valeur LMR", "Unité", "Date d'effet", "Notes"],
                    row_count=(20, "fixed"),
                    interactive=False
                )

                mrl_histogram = gr.Plot(label="MRL Distribution by Product")

                substance_search_btn.click(
                    fn=self.search_substance,
                    inputs=[substance_query],
                    outputs=[substance_results]
                )

                substance_results.select(  # Use .select() for interactive selection
                        fn=lambda evt: evt.data[0] if evt.data else None,
                        inputs=[],
                        outputs=substance_select
                )


                substance_mrls_btn.click(
                    fn=lambda sid: (
                        self.get_substance_mrls(int(sid) if sid is not None else 0),
                        self.create_histogram(self.get_substance_mrls(int(sid) if sid is not None else 0))
                    ),
                    inputs=[substance_select],
                    outputs=[substance_mrls, mrl_histogram]
                )

            with gr.Tab("About"):
                gr.HTML("""
                    <div style="padding: 20px;">
                        <h2>About this Application</h2>
                        <p>This application allows you to view the Maximum Residue Limits (MRLs) of pesticides
                        allowed in food products in the European Union.</p>

                        <h3>Data Sources</h3>
                        <p>Data is sourced from the official European Commission API:</p>
                        <ul>
                            <li>Pesticides API: <a href="https://api.datalake.sante.service.ec.europa.eu/sante/pesticides">
                            https://api.datalake.sante.service.ec.europa.eu/sante/pesticides</a></li>
                        </ul>

                        <h3>How to Use</h3>
                        <ul>
                            <li><strong>Search by Product</strong>: Select a food product to see all substances and their associated MRLs.</li>
                            <li><strong>Search by Substance</strong>: Search for an active substance to see its status and the products in which it is regulated.</li>
                        </ul>

                        <h3>Legend</h3>
                        <ul>
                            <li><strong>MRL</strong>: Maximum Residue Limit, usually expressed in mg/kg</li>
                            <li><strong>N/C</strong>: Not Communicated</li>
                            <li><strong>*</strong>: MRL set at the limit of quantification</li>
                        </ul>
                    </div>
                """)

        return ui

def main():
    # Command-line parameters (you could add them via argparse)
    use_cache = True  # Use cache by default

    try:
        app = PesticideApp(use_cache=use_cache)
        logger.info("Launching user interface...")
        app.create_ui().launch(
            server_name="0.0.0.0",  # Accessible from outside
            server_port=7860,       # Gradio's standard port
            share=False,            # Don't create a public link
            debug=False             # Disable debug mode in production
        )
    except Exception as e:
        logger.critical(f"Critical error launching application: {e}", exc_info=True)

if __name__ == "__main__":
    main()