f1-circuits/ergast_service.py

import re
import logging
import urllib.request
import json
from functools import lru_cache
from typing import Optional, TYPE_CHECKING

if TYPE_CHECKING:
    from circuits_service import CircuitService


def slugify(text: str) -> str:
    text = text.lower().strip()
    replacements = {
        'ü': 'u', 'ö': 'o', 'ä': 'a', 'é': 'e', 'è': 'e', 'ê': 'e', 'ë': 'e',
        'à': 'a', 'â': 'a', 'î': 'i', 'ï': 'i', 'ô': 'o', 'û': 'u', 'ñ': 'n',
        'ã': 'a', 'á': 'a', 'í': 'i', 'ó': 'o', 'ú': 'u',
    }
    for k, v in replacements.items():
        text = text.replace(k, v)
    text = re.sub(r'[^a-z0-9]+', '-', text)
    return text.strip('-')


class ErgastService:
    JOLPICA_BASE = 'https://api.jolpi.ca/ergast/f1'

    # Country name normalization: Jolpica country → our country slug
    country_map = {
        'UK': 'uk',
        'UAE': 'united-arab-emirates',
        'USA': 'united-states',
        'Korea': 'korea',
    }

    def __init__(self, circuits_service: 'CircuitService' = None):
        self.circuits_service = circuits_service
        self.logger = logging.getLogger(__name__)
        self._circuit_name_index = self._build_circuit_name_index()

    def _build_circuit_name_index(self) -> dict[str, tuple[str, str, str]]:
        """Build a lookup from slugified circuit name → (country_slug, locality_slug, circuit_slug)."""
        index = {}
        if not self.circuits_service:
            return index
        for country_slug, country in self.circuits_service.circuit_data.items():
            for locality_slug, locality in country.localities.items():
                for circuit_slug, circuit in locality.circuits.items():
                    # Index by our circuit slug (which is the slugified circuit name)
                    index[circuit_slug] = (country_slug, locality_slug, circuit_slug)
                    # Also index by ASCII-slugified version for unicode mismatches
                    ascii_slug = slugify(circuit.name)
                    if ascii_slug not in index:
                        index[ascii_slug] = (country_slug, locality_slug, circuit_slug)
        return index

    @lru_cache(maxsize=128)
    def _fetch_season_schedule(self, season: int) -> list[dict]:
        """Fetch and cache the race schedule for a season from Jolpica."""
        url = f'{self.JOLPICA_BASE}/{season}.json?limit=100'
        try:
            resp = urllib.request.urlopen(url, timeout=10)
            data = json.loads(resp.read())
            return data['MRData']['RaceTable']['Races']
        except Exception as e:
            self.logger.error(f"Failed to fetch Jolpica schedule for {season}: {e}")
            return []

    def find_slugs_for_grand_prix(self, grand_prix_name: str, season: int) -> Optional[tuple[str, str, str]]:
        """Resolve a grand prix name + season to (country_slug, locality_slug, circuit_slug)."""
        races = self._fetch_season_schedule(season)

        # Find the matching race
        race = next((r for r in races if r['raceName'] == grand_prix_name), None)
        if not race:
            self.logger.error(f"Race '{grand_prix_name}' not found in {season} schedule")
            return None, None, None

        circuit = race['Circuit']
        circuit_name_slug = slugify(circuit['circuitName'])

        # Try matching by slugified circuit name
        if circuit_name_slug in self._circuit_name_index:
            return self._circuit_name_index[circuit_name_slug]

        self.logger.warning(f"Circuit '{circuit['circuitName']}' (slug: {circuit_name_slug}) not found in index")
        return None, None, None