import type { Parser, ExtractedRecord, RawDocument } from "@dci/connectors";
import { resolvePlace, regionRecord, type Place } from "./shared.js";

/**
 * Microsoft Azure — https://datacenters.microsoft.com/globe/data/geo/regions.json
 * Public JSON behind the "Azure global infrastructure" globe. One object per region:
 *   id, displayName, location (city or state), geographyId, latitude/longitude (region marker, NOT a
 *   facility), typeId (regionaz | aregionaz | region | regionaaz), isOpen, yearOpen, monthOpen,
 *   availabilityZonesId (available | soon | nearest | nearestSoon), availableTo, announcementLink.
 */
interface AzureRegion {
  id: string;
  displayName: string;
  location?: string;
  geographyId?: string;
  continent?: string;
  latitude?: number;
  longitude?: number;
  typeId?: string;
  isOpen?: boolean;
  yearOpen?: number | null;
  monthOpen?: string | null;
  availabilityZonesId?: string;
  availabilityZonesNearestRegionIds?: string[];
  availableTo?: string;
  announcementLink?: string;
  dataResidency?: string;
  hasGroundStation?: boolean;
}

/** Azure geographyId → ISO-2 (fallback when `location` is a state/area that the city table cannot resolve). */
const GEOGRAPHY_ISO: Record<string, string> = {
  africa: "ZA", australia: "AU", austria: "AT", belgium: "BE", brazil: "BR", canada: "CA", chile: "CL", china: "CN", denmark: "DK", finland: "FI", france: "FR", germany: "DE", greece: "GR", india: "IN", indonesia: "ID", israel: "IL", italy: "IT", japan: "JP", korea: "KR", malaysia: "MY", mexico: "MX", newzealand: "NZ", norway: "NO", poland: "PL", qatar: "QA", saudiarabia: "SA", spain: "ES", sweden: "SE", switzerland: "CH", taiwan: "TW", unitedarabemirates: "AE", unitedkingdom: "GB", unitedstates: "US",
};
/** Regions whose `location` is ambiguous without a country (e.g. "Hong Kong" under geography asiapacific). */
const LOCATION_ISO_OVERRIDES: Record<string, string> = { eastasia: "HK", southeastasia: "SG", northeurope: "IE", westeurope: "NL", chinanorth3: "CN" };

/** Sub-national areas Microsoft names that are neither a city nor a US state. */
const AREA_LOCATIONS: Record<string, { regionName: string; iso: string }> = { Hebei: { regionName: "Hebei", iso: "CN" } };

const MONTHS: Record<string, string> = { january: "01", february: "02", march: "03", april: "04", may: "05", june: "06", july: "07", august: "08", september: "09", october: "10", november: "11", december: "12" };

export const azureRegionsParser: Parser = {
  name: "cloud_azure_regions_v1",
  version: "1.0.0",
  pageTypes: ["cloud_region"],
  parse(doc: RawDocument, ctx): ExtractedRecord[] {
    let list: AzureRegion[];
    try { const parsed = JSON.parse(doc.text) as unknown; list = Array.isArray(parsed) ? (parsed as AzureRegion[]) : ((parsed as { regions?: AzureRegion[] }).regions ?? []); } catch (e) { ctx.log("warn", `azure regions.json not JSON: ${(e as Error).message}`); return []; }
    const out: ExtractedRecord[] = [];
    for (const r of list) {
      if (!r?.id || !r.displayName) continue;
      const isoHint = LOCATION_ISO_OVERRIDES[r.id] ?? (r.geographyId ? GEOGRAPHY_ISO[r.geographyId] : undefined) ?? null;
      const area = r.location ? AREA_LOCATIONS[r.location] : undefined;
      let place: Place = area ? { city: null, regionName: area.regionName, countryIso2: area.iso, lat: null, lng: null, geoPrecision: null } : resolvePlace(r.location ?? null, isoHint);
      if (!place.countryIso2 && isoHint) place.countryIso2 = isoHint;
      // Curated city coordinates first. When Microsoft only names a sub-national area (a US state, Hebei, …) fall
      // back to their own region marker flagged "approximate". When they only name a country ("Netherlands",
      // "Ireland", "Israel") no marker is emitted: their country markers can sit across a border or offshore.
      let geoSource: string | undefined;
      if (place.lat == null && place.regionName && !place.city && typeof r.latitude === "number" && typeof r.longitude === "number") {
        place = { ...place, lat: r.latitude, lng: r.longitude, geoPrecision: "approximate" };
        geoSource = "provider-marker:azure";
      }
      const open = r.isOpen === true;
      const restricted = /business presence in China|government|by invitation/i.test(r.availableTo ?? "");
      const status = !open ? "announced" : restricted ? "restricted" : "operational";
      const launchedOn = open && r.yearOpen ? (r.monthOpen && MONTHS[r.monthOpen.toLowerCase()] ? `${r.yearOpen}-${MONTHS[r.monthOpen.toLowerCase()]}` : String(r.yearOpen)) : null;
      const azState = r.availabilityZonesId ?? "";
      const externalIds: Record<string, string | number> = { azure_region_id: r.id, azure_type: r.typeId ?? "", azure_geography: r.geographyId ?? "", availability_zones_state: azState };
      if (r.availableTo) externalIds.available_to = r.availableTo;
      if (r.announcementLink) externalIds.announcement_url = r.announcementLink;
      if (r.dataResidency) externalIds.data_residency = r.dataResidency;
      if (r.location) externalIds.published_location = r.location;
      out.push(
        regionRecord({
          connectorId: ctx.connectorId,
          keyPrefix: "azure:region",
          providerName: "Microsoft Azure",
          code: r.id,
          name: r.displayName,
          place,
          // Microsoft publishes whether AZs are available, not how many; leave the count null when unknown.
          availabilityZones: null,
          launchedOn,
          status,
          isSovereign: false,
          externalIds,
          geoSource,
          url: doc.finalUrl,
          methods: { code: "json:id", name: "json:displayName", city: "json:location+curated-city-table", status: "json:isOpen", launchedOn: "json:yearOpen", ...(geoSource ? { lat: "json:latitude(region marker)" } : {}) },
          certainty: 0.98,
        }),
      );
    }
    ctx.log("info", `azure: ${out.length} regions (${out.filter((r) => r.data.status === "announced").length} announced)`);
    return out;
  },
};