import type { ExtractionResult } from '@rareindex/shared';
import type { FetchOptions } from '../types.js';
import { policyFor } from '../domains.js';

export const DEFAULT_UA = 'RareIndexBot/0.1 (+https://www.rareindex.io/about/data; market-data research; contact: data@rareindex.io)';
/** Used only where connectors/domains.json sets userAgent:"browser" (site serves bots a degraded page; terms allow). */
export const BROWSER_UA = 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.5 Safari/605.1.15';

export interface HttpEngineOptions {
  userAgent?: string;
  defaultTimeoutMs?: number;
  maxRetries?: number;
}

/**
 * Direct HTTP engine for official APIs, feeds and plain public pages. Honest user agent, no
 * evasion, exponential backoff on 429/5xx, and conditional requests when the caller passes
 * If-None-Match / If-Modified-Since headers (budget engine).
 */
export function createHttpEngine(opts: HttpEngineOptions = {}) {
  const ua = opts.userAgent ?? DEFAULT_UA;

  async function fetchUrl(url: string, o: FetchOptions = {}): Promise<ExtractionResult> {
    const started = Date.now();
    const policy = policyFor(url);
    const maxRetries = opts.maxRetries ?? policy.maxRetries;
    const timeoutMs = o.timeoutMs ?? opts.defaultTimeoutMs ?? policy.timeoutMs;
    let attempt = 0;
    let lastErr: string | null = null;
    while (attempt <= maxRetries) {
      attempt++;
      const ctrl = new AbortController();
      const timer = setTimeout(() => ctrl.abort(), timeoutMs);
      try {
        const res = await fetch(url, {
          method: o.method ?? 'GET',
          headers: {
            'user-agent': o.headers?.['user-agent'] ?? (policyFor(url).userAgent === 'browser' ? BROWSER_UA : ua),
            accept: o.responseType === 'text' ? 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8' : 'application/json, text/plain, */*',
            'accept-language': 'en-US,en;q=0.9',
            ...(o.body !== undefined ? { 'content-type': 'application/json' } : {}),
            ...(o.headers ?? {}),
          },
          body: o.body !== undefined ? JSON.stringify(o.body) : undefined,
          signal: ctrl.signal,
          redirect: 'follow',
        });
        clearTimeout(timer);
        const status = res.status;
        if (status === 304) {
          return { success: true, engine: 'api', url, finalUrl: res.url, httpStatus: 304, html: null, markdown: null, json: null, qualityScore: 1, requiresReview: false, error: null, costCredits: 0, durationMs: Date.now() - started, fetchedAt: new Date() };
        }
        if ((status === 429 || status >= 500) && attempt <= maxRetries) {
          const retryAfter = Number(res.headers.get('retry-after'));
          // Never sleep longer than a minute inside a fetch: a long Retry-After (comics.org sends ~30 min) is a
          // hard rate limit → return the 429 and let the scheduler/circuit breaker back off instead.
          if (Number.isFinite(retryAfter) && retryAfter > 60) {
            return { success: false, engine: 'api', url, finalUrl: res.url, httpStatus: status, html: null, markdown: null, json: null, qualityScore: 0, requiresReview: false, error: `HTTP ${status} (retry-after ${retryAfter}s)`, costCredits: 0, durationMs: Date.now() - started, fetchedAt: new Date() };
          }
          const wait = Number.isFinite(retryAfter) && retryAfter > 0 ? retryAfter * 1000 : Math.min(30_000, 1000 * 2 ** attempt);
          lastErr = `HTTP ${status}`;
          await new Promise((r) => setTimeout(r, wait));
          continue;
        }
        const ctype = res.headers.get('content-type') ?? '';
        if (o.responseType === 'binary') {
          const buf = new Uint8Array(await res.arrayBuffer());
          const okB = res.ok && (o.failOnHttpError === false || status < 400);
          return { success: okB, engine: 'api', url, finalUrl: res.url, httpStatus: status, html: null, markdown: null, json: null, buffer: buf, qualityScore: okB ? 1 : 0, requiresReview: false, error: okB ? null : `HTTP ${status}`, costCredits: 0, durationMs: Date.now() - started, fetchedAt: new Date() };
        }
        const isJson = o.responseType !== 'text' && (ctype.includes('json') || o.responseType === 'json');
        let json: unknown = null;
        let text: string | null = null;
        if (isJson) {
          const raw = await res.text();
          try {
            json = raw ? JSON.parse(raw) : null;
          } catch {
            text = raw;
          }
        } else {
          text = await res.text();
        }
        const ok = res.ok && (o.failOnHttpError === false || status < 400);
        return {
          success: ok,
          engine: 'api',
          url,
          finalUrl: res.url,
          httpStatus: status,
          html: text,
          markdown: null,
          json,
          qualityScore: ok ? 1 : 0,
          requiresReview: false,
          error: ok ? null : `HTTP ${status}`,
          costCredits: 0,
          durationMs: Date.now() - started,
          fetchedAt: new Date(),
        };
      } catch (err) {
        clearTimeout(timer);
        lastErr = err instanceof Error ? err.message : String(err);
        if (attempt > maxRetries) break;
        await new Promise((r) => setTimeout(r, Math.min(15_000, 500 * 2 ** attempt)));
      }
    }
    return { success: false, engine: 'api', url, finalUrl: null, httpStatus: null, html: null, markdown: null, json: null, qualityScore: 0, requiresReview: false, error: lastErr ?? 'fetch failed', costCredits: 0, durationMs: Date.now() - started, fetchedAt: new Date() };
  }

  return { fetch: fetchUrl };
}

export type HttpEngine = ReturnType<typeof createHttpEngine>;