import os
from sqlalchemy import create_engine, text
import logging
from dotenv import load_dotenv
from sqlalchemy.engine import URL
from typing import List, Dict, Any
import pandas as pd
import json
import logging

class DBManager:
    def __init__(self):
        load_dotenv()
        USERNAME = os.getenv("DB_USERNAME") 
        PASSWORD = os.getenv("DB_PASSWORD") 
        HOST = os.getenv("DB_HOST") 
        NAME = os.getenv("DB_NAME") 
        PORT = os.getenv("DB_PORT") 

        if not all([USERNAME, PASSWORD, HOST, NAME]):
            raise ValueError("One or more database environment variables are not set.")

        connection_url = URL.create(
            drivername="postgresql+psycopg2",
            username=USERNAME,
            password=PASSWORD,
            host=HOST,
            port=PORT,
            database=NAME,
        )

        self.engine = create_engine(connection_url, pool_pre_ping=True)

        try:
            safe_url = connection_url.render_as_string(hide_password=True)
        except Exception:
            safe_url = str(connection_url).replace(PASSWORD, "***")

        logging.info(f"Database engine created with URL: {safe_url}")

    def _execute_query(self, query: str, params: dict = None) -> pd.DataFrame:
        """Executes a SQL query and returns the result as a pandas DataFrame."""
        try:
            with self.engine.connect() as connection:
                df = pd.read_sql_query(sql=text(query), con=connection, params=params)
            return df
        except Exception as e:
            logging.error(f"Failed to execute query: {e}", exc_info=True)
            return pd.DataFrame()
        
    def get_profile_name_by_app_id(self, application_id: int) -> str | None:
        """
        Gets the applicant's full name from their profile.
        This is the most reliable source for the name.
        """
        query = """
        SELECT 
            p.first_name || ' ' || p.last_name as name
        FROM 
            application a 
        JOIN profile p ON p.id = a.profile_id
        WHERE a.id = :application_id
        LIMIT 1
        """
        params = {"application_id": application_id}
        df = self._execute_query(query, params)
        if not df.empty:
            return df.iloc[0]['name']
        return None
    
    # --- NEW METHOD ADDED ---
    def get_visa_photo_metadata_per_service(self) -> Dict[int, Dict[str, Any]]:
        """
        Retrieves visa photo specifications for each service.
        Returns a dictionary mapping service_id to its photo specifications.
        """
        query = """
            select s."name", metadata from service_document sd
            join service s on s.id = sd.service_id
            join document d on d.id = sd.document_id 
            where d.metadata  is not null and s.is_active = true and s.has_limited_discoverability = False;
        """
        df = self._execute_query(query)
        result = {}
        for _, row in df.iterrows():
            service_name = row['name']
            metadata = row['metadata']
            try:
                metadata_dict = json.loads(metadata)
                result[service_name] = metadata_dict
            except json.JSONDecodeError:
                logging.warning(f"Invalid JSON metadata for service {service_name}")
        return result
        
    def get_destination_country(self, application_id: int) -> str | None:
        """
        Gets the official destination country name for the application.
        """
        query = """
        SELECT DISTINCT
            c."name" 
        FROM country c 
        JOIN visa_service_availability vsa ON vsa.destination_country_id = c.id 
        JOIN service s ON vsa.service_id  = s.id
        JOIN service_unit su ON s.id = su.service_id 
        JOIN application a ON a.service_unit_id  = su.id 
        WHERE a.id = :application_id
        LIMIT 1;
        """
        params = {"application_id": application_id}
        df = self._execute_query(query, params)
        if not df.empty:
            return df.iloc[0]['name']
        return None
    # --- END OF NEW METHOD ---
    
    def get_form_data(self, application_id: int) -> pd.DataFrame:
        """
        HANYA mengambil data dari form standar (non-SMART_UPLOAD).
        """
        sql_query = """
        SELECT
            d.name AS name,
            ad.value AS value
        FROM
            application_document AS ad
        JOIN
            document AS d ON ad.document_id = d.id
        LEFT JOIN 
            document_tag AS dt ON ad.document_id = dt.document_id
        WHERE
            ad.application_id = :application_id
            AND (dt.use_case IS NULL OR dt.use_case != 'SMART_UPLOAD')
            AND ad.value IS NOT NULL
            AND ad.value <> ''
        """
        params = {"application_id": application_id}
        df = self._execute_query(sql_query, params)
        
        if not df.empty:
            # Use document name as the key
            df = df.drop_duplicates(subset=['name'], keep='first')
        return df

    def get_smart_upload_results(self, application_id: int) -> pd.DataFrame:
        """
        Mengambil SEMUA hasil ekstraksi JSON yang unik dari smart upload
        sebagai JSON utuh, BUKAN key-value yang dibongkar.
        Menambahkan 'created_at' untuk sorting data terbaru.
        """
        sql_query = """
        SELECT DISTINCT
            suj.extraction_result,
            COALESCE(suj.extraction_result ->> 'doc_type', 'unknown') AS doc_type,
            suj.created_at
        FROM
            application_document AS ad
        JOIN
            smart_upload_job AS suj
            ON ad.document_id = suj.document_id
            AND ad.application_id = ANY(suj.application_ids)
        WHERE
            ad.application_id = :application_id
            AND suj.status IN ('COMPLETED', 'PROCESSED')
            AND suj.extraction_result IS NOT NULL
            AND suj.extraction_result::text <> '{}'
        ORDER BY suj.created_at DESC;
        """
        params = {"application_id": application_id}
        df = self._execute_query(sql_query, params)
        return df