from __future__ import annotations
import re
from datetime import datetime
from typing import Any
import requests
from bs4 import BeautifulSoup
from . import db
from .models import ObraSocialCatalog, ObraSocialPageSnapshot
from .utils import hash_payload, safe_json_dumps

GEOREF_BASE = 'https://apis.datos.gob.ar/georef/api'
OBRAS_SOCIALES_URL = 'https://www.sssalud.gob.ar/?page=listRnosc&tipo={tipo}'


class IntegrationError(Exception):
    pass


def normalize_space(text: str) -> str:
    return re.sub(r'\s+', ' ', (text or '')).strip()


def get_requests_session():
    session = requests.Session()
    session.headers.update({
        'User-Agent': 'BookAppointmentsPro/3.0 (+Flask admin integration)'
    })
    return session


def parse_categoria_y_origen(html: str):
    soup = BeautifulSoup(html, 'html.parser')
    categoria = ''
    origen = ''

    for div in soup.select('div.col-md-8.col-md-offset-2.text-center, div.text-center'):
        text = normalize_space(div.get_text(' ', strip=True))
        if 'Orígen de datos:' in text or 'Origen de datos:' in text:
            b = div.find('b')
            if b:
                categoria = normalize_space(b.get_text(' ', strip=True))
            m = re.search(r'Or[ií]gen de datos:\s*(.+)', text, re.I)
            if m:
                origen = normalize_space(m.group(1))
            break

    if not categoria:
        header_candidates = soup.find_all(['h1', 'h2', 'h3', 'b'])
        for node in header_candidates:
            txt = normalize_space(node.get_text(' ', strip=True))
            if txt and 'RNAS' not in txt and len(txt) > 8:
                categoria = txt
                break
    return {
        'categoria_oficial': categoria,
        'origen_datos': origen,
    }


def parse_obras_sociales_rows(html: str) -> dict[str, Any]:
    soup = BeautifulSoup(html, 'html.parser')
    meta = parse_categoria_y_origen(html)
    body_text = normalize_space(soup.get_text(' ', strip=True))
    status = 'ok'
    if 'No se reportan datos' in body_text:
        status = 'no_data'
    if 'error de conexión' in body_text.lower() or 'error de base' in body_text.lower() or 'error de conexion' in body_text.lower():
        status = 'warning'

    rows = []
    table = soup.find('table')
    if table:
        trs = table.find_all('tr')
    else:
        trs = soup.find_all('tr')

    for tr in trs:
        cells = [normalize_space(td.get_text(' ', strip=True)) for td in tr.find_all(['td', 'th'])]
        if len(cells) < 5:
            continue
        header_join = ' '.join(cells[:5]).lower()
        if 'rnas' in header_join and 'denominación' in header_join:
            continue
        row = {
            'rnas': cells[0],
            'denominacion': cells[1],
            'domicilio': cells[2],
            'localidad': cells[3],
            'telefono': cells[4] if len(cells) > 4 else '',
            'linea_gratuita': cells[5] if len(cells) > 5 else '',
            'habilitada_opciones': cells[6] if len(cells) > 6 else '',
        }
        if row['rnas'] and row['denominacion']:
            rows.append(row)

    return {
        'categoria_oficial': meta['categoria_oficial'],
        'origen_datos': meta['origen_datos'],
        'status': status,
        'rows': rows,
    }


def fetch_obras_sociales_tipo(tipo: int) -> dict[str, Any]:
    session = get_requests_session()
    resp = session.get(OBRAS_SOCIALES_URL.format(tipo=tipo), timeout=30)
    resp.raise_for_status()
    parsed = parse_obras_sociales_rows(resp.text)
    parsed['tipo'] = tipo
    return parsed


def sync_obras_sociales(tipo_desde: int = 1, tipo_hasta: int = 20) -> dict[str, Any]:
    summary = {
        'pages_processed': 0,
        'pages_changed': 0,
        'rows_new': 0,
        'rows_updated': 0,
        'pages_with_error': 0,
        'messages': [],
    }
    now = datetime.utcnow()
    seen_ids = set()

    for tipo in range(tipo_desde, tipo_hasta + 1):
        page = ObraSocialPageSnapshot.query.filter_by(tipo=tipo).first()
        if not page:
            page = ObraSocialPageSnapshot(tipo=tipo)
            db.session.add(page)
            db.session.flush()
        try:
            snapshot = fetch_obras_sociales_tipo(tipo)
            payload_for_hash = {
                'tipo': tipo,
                'categoria_oficial': snapshot['categoria_oficial'],
                'origen_datos': snapshot['origen_datos'],
                'status': snapshot['status'],
                'rows': snapshot['rows'],
            }
            current_hash = hash_payload(payload_for_hash)
            changed = page.last_hash != current_hash
            if changed:
                summary['pages_changed'] += 1
            page.categoria_oficial = snapshot['categoria_oficial']
            page.origen_datos = snapshot['origen_datos']
            page.status = snapshot['status']
            page.row_count = len(snapshot['rows'])
            page.last_hash = current_hash
            page.last_error = None
            page.last_synced_at = now

            for row in snapshot['rows']:
                row_hash = hash_payload({
                    'tipo': tipo,
                    'categoria_oficial': snapshot['categoria_oficial'],
                    **row,
                })
                item = ObraSocialCatalog.query.filter_by(rnas=row['rnas']).first()
                if not item:
                    item = ObraSocialCatalog(rnas=row['rnas'])
                    db.session.add(item)
                    summary['rows_new'] += 1
                else:
                    if item.row_hash != row_hash:
                        summary['rows_updated'] += 1
                item.tipo = tipo
                item.categoria_oficial = snapshot['categoria_oficial']
                item.denominacion = row['denominacion']
                item.domicilio = row['domicilio']
                item.localidad = row['localidad']
                item.telefono = row['telefono']
                item.linea_gratuita = row['linea_gratuita']
                item.habilitada_opciones = row['habilitada_opciones']
                item.vigente = True
                item.row_hash = row_hash
                item.last_seen_at = now
                item.page_snapshot_id = page.id
                db.session.flush()
                seen_ids.add(item.id)

            summary['pages_processed'] += 1
        except Exception as exc:
            page.status = 'error'
            page.last_error = str(exc)
            page.last_synced_at = now
            summary['pages_with_error'] += 1
            summary['messages'].append(f'tipo={tipo}: {exc}')

    # marcar como no vigentes las no vistas en esta corrida
    if seen_ids:
        ObraSocialCatalog.query.filter(~ObraSocialCatalog.id.in_(seen_ids)).update(
            {'vigente': False}, synchronize_session=False
        )
    db.session.commit()
    return summary


def georef_get(path: str, params: dict[str, Any] | None = None):
    session = get_requests_session()
    resp = session.get(f'{GEOREF_BASE}/{path}', params=params or {}, timeout=30)
    resp.raise_for_status()
    return resp.json()


def get_provinces():
    data = georef_get('provincias', {'campos': 'id,nombre', 'max': 100})
    return data.get('provincias', [])


def get_municipios(provincia_id: str):
    data = georef_get('municipios', {'provincia': provincia_id, 'campos': 'id,nombre', 'max': 500})
    return data.get('municipios', [])


def get_localidades(provincia_id: str = '', municipio_id: str = ''):
    params = {'campos': 'id,nombre', 'max': 500}
    if provincia_id:
        params['provincia'] = provincia_id
    if municipio_id:
        params['municipio'] = municipio_id
    data = georef_get('localidades', params)
    return data.get('localidades', [])


def normalize_sisa_item(item: Any) -> dict[str, Any]:
    if isinstance(item, dict):
        getter = item.get
    else:
        getter = lambda k, default=None: getattr(item, k, default)

    name_parts = [getter('apellido'), getter('nombre')]
    visible = normalize_space(' '.join([p for p in name_parts if p]))
    if not visible:
        visible = getter('nombreCompleto') or getter('displayName') or getter('nombre_apellido') or ''
    specialty = getter('especialidad') or getter('specialty') or getter('especialidadNombre') or ''
    jurisdiction = getter('jurisdiccion') or getter('jurisdiccionNombre') or getter('provincia') or ''
    return {
        'documento': getter('dni') or getter('documento') or '',
        'display_name': normalize_space(visible),
        'matricula': getter('matricula') or getter('matriculaNacional') or getter('numeroMatricula') or '',
        'profession_name': getter('profesion') or getter('profesionNombre') or getter('titulo') or '',
        'specialty': specialty,
        'jurisdiction_name': jurisdiction,
        'state_name': getter('estado') or getter('estadoRegistro') or getter('situacion') or '',
        'raw': safe_json_dumps(item),
    }


def _zeep_client(wsdl: str):
    from zeep import Client
    from zeep.transports import Transport
    session = get_requests_session()
    transport = Transport(session=session, timeout=30)
    return Client(wsdl=wsdl, transport=transport)


def sisa_test_connection(wsdl: str, user: str, password: str, operation: str) -> dict[str, Any]:
    if not wsdl:
        raise IntegrationError('Falta configurar la URL WSDL.')
    if not user or not password:
        raise IntegrationError('Faltan las credenciales SISA.')
    client = _zeep_client(wsdl)
    service = client.service
    if not hasattr(service, operation):
        ops = ', '.join(sorted(client.wsdl.bindings[next(iter(client.wsdl.bindings))].all())[:10]) if client.wsdl.bindings else ''
        raise IntegrationError(f'La operación {operation} no existe en el WSDL. Operaciones detectadas: {ops}')
    return {'ok': True, 'message': f'WSDL cargado correctamente. Operación disponible: {operation}'}


def sisa_search_professionals(wsdl: str, user: str, password: str, operation: str, dni: str = '', query: str = '', matricula: str = '') -> list[dict[str, Any]]:
    if not wsdl:
        raise IntegrationError('Falta configurar la URL WSDL.')
    if not user or not password:
        raise IntegrationError('Faltan las credenciales SISA.')
    client = _zeep_client(wsdl)
    service = client.service
    if not hasattr(service, operation):
        raise IntegrationError(f'La operación {operation} no existe en el WSDL configurado.')

    method = getattr(service, operation)
    candidates = [
        {'usuario': user, 'clave': password, 'dni': dni, 'apellidoNombre': query, 'matricula': matricula},
        {'usuario': user, 'password': password, 'dni': dni, 'apellidoNombre': query, 'matricula': matricula},
        {'user': user, 'password': password, 'dni': dni, 'apellidoNombre': query, 'matricula': matricula},
        {'usuario': user, 'clave': password, 'documento': dni, 'nombreApellido': query, 'matricula': matricula},
        {'usuario': user, 'clave': password, 'matricula': matricula, 'query': query, 'dni': dni},
        {'dni': dni, 'apellidoNombre': query, 'matricula': matricula, 'usuario': user, 'clave': password},
    ]
    last_error = None
    response = None
    for kwargs in candidates:
        clean_kwargs = {k: v for k, v in kwargs.items() if v not in (None, '')}
        try:
            response = method(**clean_kwargs)
            break
        except Exception as exc:
            last_error = exc
            continue
    if response is None:
        raise IntegrationError(f'No se pudo ejecutar la operación SISA: {last_error}')

    if response is None:
        return []
    if isinstance(response, list):
        items = response
    elif isinstance(response, dict):
        for key in ('profesionales', 'items', 'return', 'resultado', 'results'):
            if key in response and response[key]:
                items = response[key]
                break
        else:
            items = [response]
    else:
        possible = None
        for key in ('profesionales', 'items', 'return', 'resultado', 'results'):
            possible = getattr(response, key, None)
            if possible:
                break
        items = possible if possible is not None else [response]

    normalized = [normalize_sisa_item(item) for item in items if item]
    normalized = [item for item in normalized if item['display_name'] or item['documento'] or item['matricula']]
    return normalized