export interface ArchiveResolution {
  htmlOrText: string;
  archiveUrl: string;
  provider: "wikipedia-rest" | "wayback-cdx-raw" | "wayback-avail" | "jina-reader";
  isPlainText?: boolean;
}

const UNARCHIVABLE_PATHS =
  /\b(wp-content\/uploads\/.*\/\/|\/search\?|\/login|\/signin|merriam-webster\.com\/dictionary)\b/i;

/**
 * Unwraps URLs that are already Wayback / Archive links so we never double-wrap them.
 */
export function unwrapArchiveUrl(rawUrl: string): string {
  const match = rawUrl.match(
    /^https?:\/\/web\.archive\.org\/web\/[0-9a-z_*]+\/(https?:\/\/.+)$/i
  );
  if (match && match[1]) {
    return match[1];
  }
  return rawUrl;
}

export async function fetchFromArchiveWaterfall(
  inputUrl: string,
  log?: (msg: string) => void
): Promise<ArchiveResolution | null> {
  const targetUrl = unwrapArchiveUrl(inputUrl.trim());

  if (UNARCHIVABLE_PATHS.test(targetUrl)) {
    return null;
  }

  // 1. Fast-Path Wikipedia REST API
  const wikiMatch = targetUrl.match(
    /^https?:\/\/([a-z]{2,3})\.wikipedia\.org\/wiki\/([^#?]+)/i
  );
  if (wikiMatch) {
    const lang = wikiMatch[1].toLowerCase();
    const rawTitle = decodeURIComponent(wikiMatch[2]);
    const wikiContent = await fetchWikipediaRestArticle(lang, rawTitle);
    if (wikiContent) {
      log?.(`[ARCHIVE] Resolved Wikipedia article via REST API: ${rawTitle}`);
      return {
        htmlOrText: wikiContent,
        archiveUrl: `https://${lang}.wikipedia.org/wiki/${encodeURIComponent(rawTitle)}`,
        provider: "wikipedia-rest",
        isPlainText: true,
      };
    }
  }

  // 2. Wayback CDX + raw id_
  const cdxSnap = await fetchWaybackCdxRaw(targetUrl);
  if (cdxSnap) {
    log?.(`[ARCHIVE] Recovered via Wayback CDX raw snapshot: ${cdxSnap.archiveUrl}`);
    return cdxSnap;
  }

  // 3. Wayback Availability API
  const availSnap = await fetchWaybackAvailability(targetUrl);
  if (availSnap) {
    log?.(`[ARCHIVE] Recovered via Wayback Availability API: ${availSnap.archiveUrl}`);
    return availSnap;
  }

  // 4. Clean Reader Mirror
  const readerSnap = await fetchCleanReaderMirror(targetUrl);
  if (readerSnap) {
    log?.(`[ARCHIVE] Recovered via Clean Reader Mirror for: ${targetUrl}`);
    return readerSnap;
  }

  return null;
}

async function fetchWikipediaRestArticle(
  lang: string,
  slug: string
): Promise<string | null> {
  const candidates = [
    slug,
    slug
      .split("_")
      .map((w) => (w.length > 0 ? w[0].toUpperCase() + w.slice(1) : w))
      .join("_"),
  ];

  for (const title of new Set(candidates)) {
    try {
      const apiUrl = `https://${lang}.wikipedia.org/w/api.php?action=query&prop=extracts&explaintext=1&titles=${encodeURIComponent(
        title
      )}&redirects=1&format=json`;
      const res = await fetch(apiUrl, {
        headers: {
          "User-Agent": "DeepSwarmResearch/5.9.2 (https://saverapps.uk; local@saverapps.uk)",
          Accept: "application/json",
        },
        signal: AbortSignal.timeout(4000),
      });
      if (!res.ok) continue;
      const data: any = await res.json();
      const pages = data?.query?.pages;
      if (!pages) continue;
      const pageObj: any = Object.values(pages)[0];
      if (
        pageObj &&
        !("missing" in pageObj) &&
        typeof pageObj.extract === "string" &&
        pageObj.extract.length > 150
      ) {
        return `# ${pageObj.title}\n\n${pageObj.extract}`;
      }
    } catch {
      // Try next candidate
    }
  }
  return null;
}

async function fetchWaybackCdxRaw(targetUrl: string): Promise<ArchiveResolution | null> {
  try {
    const cdxUrl = `https://web.archive.org/cdx/search/cdx?url=${encodeURIComponent(
      targetUrl
    )}&output=json&fl=timestamp,original&filter=statuscode:200&collapse=digest&limit=-1`;

    const cdxRes = await fetch(cdxUrl, {
      headers: { "User-Agent": "DeepSwarmResearch/5.9.2" },
      signal: AbortSignal.timeout(4500),
    });
    if (!cdxRes.ok) return null;

    const rows: any = await cdxRes.json();
    if (!Array.isArray(rows) || rows.length < 2 || !Array.isArray(rows[1])) {
      return null;
    }

    const [timestamp, original] = rows[1];
    const rawSnapshotUrl = `https://web.archive.org/web/${timestamp}id_/${original}`;

    const snapRes = await fetch(rawSnapshotUrl, {
      headers: { "User-Agent": "Mozilla/5.0 (compatible; DeepSwarmResearch/5.9.2)" },
      signal: AbortSignal.timeout(6500),
    });
    if (!snapRes.ok) return null;

    const html = await snapRes.text();
    if (html.length < 250) return null;

    return {
      htmlOrText: html,
      archiveUrl: `https://web.archive.org/web/${timestamp}/${original}`,
      provider: "wayback-cdx-raw",
    };
  } catch {
    return null;
  }
}

async function fetchWaybackAvailability(targetUrl: string): Promise<ArchiveResolution | null> {
  try {
    const availUrl = `https://archive.org/wayback/available?url=${encodeURIComponent(targetUrl)}`;
    const res = await fetch(availUrl, { signal: AbortSignal.timeout(4000) });
    if (!res.ok) return null;

    const data: any = await res.json();
    const closest = data?.archived_snapshots?.closest;
    if (!closest?.available || !closest?.url) return null;

    const rawUrl = String(closest.url)
      .replace(/^http:\/\//i, "https://")
      .replace(/\/web\/(\d+)\//, "/web/$1id_/");

    const pageRes = await fetch(rawUrl, { signal: AbortSignal.timeout(6500) });
    if (!pageRes.ok) return null;

    const html = await pageRes.text();
    if (html.length < 250) return null;

    return {
      htmlOrText: html,
      archiveUrl: String(closest.url).replace(/^http:\/\//i, "https://"),
      provider: "wayback-avail",
    };
  } catch {
    return null;
  }
}

async function fetchCleanReaderMirror(targetUrl: string): Promise<ArchiveResolution | null> {
  if (/\b(academia\.edu|researchgate\.net|sciencedirect\.com)\b/i.test(targetUrl)) {
    return null;
  }
  try {
    const mirrorUrl = `https://r.jina.ai/${targetUrl}`;
    const res = await fetch(mirrorUrl, {
      headers: {
        Accept: "text/plain",
        "X-Return-Format": "markdown",
      },
      signal: AbortSignal.timeout(6000),
    });
    if (!res.ok) return null;

    const text = (await res.text()).trim();
    if (
      text.length < 300 ||
      /Access Denied|Just a moment\.\.\.|Enable JavaScript and cookies/i.test(text.slice(0, 400))
    ) {
      return null;
    }

    return {
      htmlOrText: text,
      archiveUrl: targetUrl,
      provider: "jina-reader",
      isPlainText: true,
    };
  } catch {
    return null;
  }
}

export interface ArchiveResolution {
  htmlOrText: string;
  archiveUrl: string;
  provider: "wikipedia-rest" | "wayback-cdx-raw" | "wayback-avail" | "jina-reader";
  isPlainText?: boolean;
}

const UNARCHIVABLE_PATHS =
  /\b(wp-content\/uploads\/.*\/\/|\/search\?|\/login|\/signin|merriam-webster\.com\/dictionary)\b/i;

/**
 * Unwraps URLs that are already Wayback / Archive links so we never double-wrap them.
 */
export function unwrapArchiveUrl(rawUrl: string): string {
  const match = rawUrl.match(
    /^https?:\/\/web\.archive\.org\/web\/[0-9a-z_*]+\/(https?:\/\/.+)$/i
  );
  if (match && match[1]) {
    return match[1];
  }
  return rawUrl;
}

export async function fetchFromArchiveWaterfall(
  inputUrl: string,
  log?: (msg: string) => void
): Promise<ArchiveResolution | null> {
  const targetUrl = unwrapArchiveUrl(inputUrl.trim());

  if (UNARCHIVABLE_PATHS.test(targetUrl)) {
    return null;
  }

  // 1. Fast-Path Wikipedia REST API
  const wikiMatch = targetUrl.match(
    /^https?:\/\/([a-z]{2,3})\.wikipedia\.org\/wiki\/([^#?]+)/i
  );
  if (wikiMatch) {
    const lang = wikiMatch[1].toLowerCase();
    const rawTitle = decodeURIComponent(wikiMatch[2]);
    const wikiContent = await fetchWikipediaRestArticle(lang, rawTitle);
    if (wikiContent) {
      log?.(`[ARCHIVE] Resolved Wikipedia article via REST API: ${rawTitle}`);
      return {
        htmlOrText: wikiContent,
        archiveUrl: `https://${lang}.wikipedia.org/wiki/${encodeURIComponent(rawTitle)}`,
        provider: "wikipedia-rest",
        isPlainText: true,
      };
    }
  }

  // 2. Wayback CDX + raw id_
  const cdxSnap = await fetchWaybackCdxRaw(targetUrl);
  if (cdxSnap) {
    log?.(`[ARCHIVE] Recovered via Wayback CDX raw snapshot: ${cdxSnap.archiveUrl}`);
    return cdxSnap;
  }

  // 3. Wayback Availability API
  const availSnap = await fetchWaybackAvailability(targetUrl);
  if (availSnap) {
    log?.(`[ARCHIVE] Recovered via Wayback Availability API: ${availSnap.archiveUrl}`);
    return availSnap;
  }

  // 4. Clean Reader Mirror
  const readerSnap = await fetchCleanReaderMirror(targetUrl);
  if (readerSnap) {
    log?.(`[ARCHIVE] Recovered via Clean Reader Mirror for: ${targetUrl}`);
    return readerSnap;
  }

  return null;
}

async function fetchWikipediaRestArticle(
  lang: string,
  slug: string
): Promise<string | null> {
  const candidates = [
    slug,
    slug
      .split("_")
      .map((w) => (w.length > 0 ? w[0].toUpperCase() + w.slice(1) : w))
      .join("_"),
  ];

  for (const title of new Set(candidates)) {
    try {
      const apiUrl = `https://${lang}.wikipedia.org/w/api.php?action=query&prop=extracts&explaintext=1&titles=${encodeURIComponent(
        title
      )}&redirects=1&format=json`;
      const res = await fetch(apiUrl, {
        headers: {
          "User-Agent": "DeepSwarmResearch/5.9.2 (https://saverapps.uk; local@saverapps.uk)",
          Accept: "application/json",
        },
        signal: AbortSignal.timeout(4000),
      });
      if (!res.ok) continue;
      const data: any = await res.json();
      const pages = data?.query?.pages;
      if (!pages) continue;
      const pageObj: any = Object.values(pages)[0];
      if (
        pageObj &&
        !("missing" in pageObj) &&
        typeof pageObj.extract === "string" &&
        pageObj.extract.length > 150
      ) {
        return `# ${pageObj.title}\n\n${pageObj.extract}`;
      }
    } catch {
      // Try next candidate
    }
  }
  return null;
}

async function fetchWaybackCdxRaw(targetUrl: string): Promise<ArchiveResolution | null> {
  try {
    const cdxUrl = `https://web.archive.org/cdx/search/cdx?url=${encodeURIComponent(
      targetUrl
    )}&output=json&fl=timestamp,original&filter=statuscode:200&collapse=digest&limit=-1`;

    const cdxRes = await fetch(cdxUrl, {
      headers: { "User-Agent": "DeepSwarmResearch/5.9.2" },
      signal: AbortSignal.timeout(4500),
    });
    if (!cdxRes.ok) return null;

    const rows: any = await cdxRes.json();
    if (!Array.isArray(rows) || rows.length < 2 || !Array.isArray(rows[1])) {
      return null;
    }

    const [timestamp, original] = rows[1];
    const rawSnapshotUrl = `https://web.archive.org/web/${timestamp}id_/${original}`;

    const snapRes = await fetch(rawSnapshotUrl, {
      headers: { "User-Agent": "Mozilla/5.0 (compatible; DeepSwarmResearch/5.9.2)" },
      signal: AbortSignal.timeout(6500),
    });
    if (!snapRes.ok) return null;

    const html = await snapRes.text();
    if (html.length < 250) return null;

    return {
      htmlOrText: html,
      archiveUrl: `https://web.archive.org/web/${timestamp}/${original}`,
      provider: "wayback-cdx-raw",
    };
  } catch {
    return null;
  }
}

async function fetchWaybackAvailability(targetUrl: string): Promise<ArchiveResolution | null> {
  try {
    const availUrl = `https://archive.org/wayback/available?url=${encodeURIComponent(targetUrl)}`;
    const res = await fetch(availUrl, { signal: AbortSignal.timeout(4000) });
    if (!res.ok) return null;

    const data: any = await res.json();
    const closest = data?.archived_snapshots?.closest;
    if (!closest?.available || !closest?.url) return null;

    const rawUrl = String(closest.url)
      .replace(/^http:\/\//i, "https://")
      .replace(/\/web\/(\d+)\//, "/web/$1id_/");

    const pageRes = await fetch(rawUrl, { signal: AbortSignal.timeout(6500) });
    if (!pageRes.ok) return null;

    const html = await pageRes.text();
    if (html.length < 250) return null;

    return {
      htmlOrText: html,
      archiveUrl: String(closest.url).replace(/^http:\/\//i, "https://"),
      provider: "wayback-avail",
    };
  } catch {
    return null;
  }
}

async function fetchCleanReaderMirror(targetUrl: string): Promise<ArchiveResolution | null> {
  if (/\b(academia\.edu|researchgate\.net|sciencedirect\.com)\b/i.test(targetUrl)) {
    return null;
  }
  try {
    const mirrorUrl = `https://r.jina.ai/${targetUrl}`;
    const res = await fetch(mirrorUrl, {
      headers: {
        Accept: "text/plain",
        "X-Return-Format": "markdown",
      },
      signal: AbortSignal.timeout(6000),
    });
    if (!res.ok) return null;

    const text = (await res.text()).trim();
    if (
      text.length < 300 ||
      /Access Denied|Just a moment\.\.\.|Enable JavaScript and cookies/i.test(text.slice(0, 400))
    ) {
      return null;
    }

    return {
      htmlOrText: text,
      archiveUrl: targetUrl,
      provider: "jina-reader",
      isPlainText: true,
    };
  } catch {
    return null;
  }
}