eranos/src/lib/linkEmbed.ts

import { MASTODON_SERVERS } from '@/lib/mastodonServers';

/** Extract a YouTube video ID from a URL, or null if not a YouTube link. */
export function extractYouTubeId(url: string): string | null {
  try {
    const u = new URL(url);
    if ((u.hostname === 'www.youtube.com' || u.hostname === 'youtube.com' || u.hostname === 'm.youtube.com') && u.pathname === '/watch') {
      return u.searchParams.get('v');
    }
    if ((u.hostname === 'www.youtube.com' || u.hostname === 'youtube.com') && u.pathname.startsWith('/embed/')) {
      return u.pathname.split('/')[2] || null;
    }
    if ((u.hostname === 'www.youtube.com' || u.hostname === 'youtube.com') && u.pathname.startsWith('/shorts/')) {
      return u.pathname.split('/')[2] || null;
    }
    if (u.hostname === 'youtu.be') {
      return u.pathname.slice(1) || null;
    }
  } catch {
    // not a valid URL
  }
  return null;
}

/** Extract a tweet/post ID from a Twitter or X URL, or null if not a tweet link. */
export function extractTweetId(url: string): string | null {
  try {
    const u = new URL(url);
    const host = u.hostname.replace(/^www\./, '').replace(/^mobile\./, '');
    if (host !== 'twitter.com' && host !== 'x.com') return null;
    const match = u.pathname.match(/^\/[^/]+\/status\/(\d+)/);
    return match ? match[1] : null;
  } catch {
    return null;
  }
}

/** Bluesky post info extracted from a bsky.app URL. */
export interface BlueskyPostInfo {
  /** Handle or DID of the author. */
  author: string;
  /** Record key of the post. */
  rkey: string;
}

/** Extract Bluesky post info from a bsky.app URL, or null if not a Bluesky post link. */
export function extractBlueskyPost(url: string): BlueskyPostInfo | null {
  try {
    const u = new URL(url);
    const host = u.hostname.replace(/^www\./, '');
    if (host !== 'bsky.app') return null;
    // Match /profile/{handle-or-did}/post/{rkey}
    const match = u.pathname.match(/^\/profile\/([^/]+)\/post\/([a-z0-9]+)$/i);
    return match ? { author: match[1], rkey: match[2] } : null;
  } catch {
    return null;
  }
}

/** Extract a Mastodon post URL if the domain is a known Mastodon instance. */
export function extractMastodonPost(url: string): string | null {
  try {
    const u = new URL(url);
    const host = u.hostname.replace(/^www\./, '');
    if (!MASTODON_SERVERS.has(host)) return null;
    // Match /@{user}/{id} or /@{user}@{domain}/{id} (remote posts)
    if (/^\/@[^/]+\/\d+$/.test(u.pathname)) {
      return url;
    }
    return null;
  } catch {
    return null;
  }
}

/** Spotify embed info extracted from an open.spotify.com URL. */
export interface SpotifyEmbedInfo {
  /** Content type: track, album, playlist, episode, show. */
  type: string;
  /** Spotify content ID. */
  id: string;
}

/** Extract Spotify embed info from an open.spotify.com URL. */
export function extractSpotifyEmbed(url: string): SpotifyEmbedInfo | null {
  try {
    const u = new URL(url);
    const host = u.hostname.replace(/^www\./, '');
    if (host !== 'open.spotify.com') return null;
    // Match /{type}/{id} where type is track, album, playlist, episode, or show
    const match = u.pathname.match(/^\/(track|album|playlist|episode|show)\/([a-zA-Z0-9]+)/);
    return match ? { type: match[1], id: match[2] } : null;
  } catch {
    return null;
  }
}

/** Extract a Reddit post URL, or null if not a Reddit post link. */
export function extractRedditPost(url: string): string | null {
  try {
    const u = new URL(url);
    const host = u.hostname.replace(/^www\./, '').replace(/^old\./, '').replace(/^new\./, '');
    if (host !== 'reddit.com') return null;
    // Match /r/{subreddit}/comments/{id}/...
    if (/^\/r\/[^/]+\/comments\/[a-z0-9]+/i.test(u.pathname)) {
      return url;
    }
    return null;
  } catch {
    return null;
  }
}

/** Extract an Internet Archive item identifier from an archive.org URL, or null if not a match. */
export function extractArchiveOrgId(url: string): string | null {
  try {
    const u = new URL(url);
    const host = u.hostname.replace(/^www\./, '');
    if (host !== 'archive.org') return null;
    // Match /details/{identifier} or /embed/{identifier}
    const match = u.pathname.match(/^\/(details|embed)\/([^/?#]+)/);
    return match ? match[2] : null;
  } catch {
    return null;
  }
}

/**
 * Extract a Wikidata entity ID (Q…, P…, or L…) from a Wikidata URL, or null if not one.
 *
 * Supports the "globally unique" concept URI form described at
 * https://www.wikidata.org/wiki/Wikidata:Identifiers:
 *   - https://www.wikidata.org/entity/Q42
 *   - http://www.wikidata.org/entity/Q42
 * …as well as the browser-facing page URL:
 *   - https://www.wikidata.org/wiki/Q42
 */
export function extractWikidataId(url: string): string | null {
  try {
    const u = new URL(url);
    const host = u.hostname.replace(/^www\./, '');
    if (host !== 'wikidata.org') return null;
    // Match /entity/{id} or /wiki/{id} where id is Q|P|L followed by digits.
    const match = u.pathname.match(/^\/(?:entity|wiki)\/([QPL]\d+)$/);
    return match ? match[1] : null;
  } catch {
    return null;
  }
}

/**
 * Extract a Wikipedia article title from a Wikipedia URL, or null if not a Wikipedia article link.
 * Supports en.wikipedia.org/wiki/{title} and other language subdomains.
 */
export function extractWikipediaTitle(url: string): string | null {
  try {
    const u = new URL(url);
    const host = u.hostname.replace(/^www\./, '');
    // Match {lang}.wikipedia.org
    if (!host.endsWith('.wikipedia.org')) return null;
    // Match /wiki/{title} — ignore special pages, talk pages, etc.
    const match = u.pathname.match(/^\/wiki\/([^:][^#]*)$/);
    if (!match) return null;
    return decodeURIComponent(match[1].replace(/_/g, ' '));
  } catch {
    return null;
  }
}

/**
 * Lookup keys that can identify a Magic: The Gathering card printing on Scryfall.
 * Returned by {@link extractGathererCard}.
 */
export type GathererCard =
  | { kind: 'multiverse'; multiverseId: string }
  | { kind: 'set'; set: string; number: string; lang: string };

/**
 * Extract a Magic card reference from a gatherer.wizards.com URL.
 *
 * Supports two URL shapes Wizards has used over the years:
 *
 *   1. Modern pretty URLs:
 *      `https://gatherer.wizards.com/{SET}/{LANG}/{COLLECTOR_NUMBER}/{slug}`
 *      e.g. `/BNG/en-us/156/xenagos-god-of-revels`
 *
 *   2. Legacy ASPX query-string URLs:
 *      `https://gatherer.wizards.com/Pages/Card/Details.aspx?multiverseid=12345`
 *
 * Returns `null` for any Gatherer URL that can't be mapped to a Scryfall
 * lookup (e.g. name-only searches, card list pages, printings pages).
 */
export function extractGathererCard(url: string): GathererCard | null {
  try {
    const u = new URL(url);
    const host = u.hostname.replace(/^www\./, '');
    if (host !== 'gatherer.wizards.com') return null;

    // Legacy query-string form.
    if (/\/Pages\/Card\/[^/]+\.aspx$/i.test(u.pathname)) {
      const multiverseId = u.searchParams.get('multiverseid');
      if (multiverseId && /^\d+$/.test(multiverseId)) {
        return { kind: 'multiverse', multiverseId };
      }
      return null;
    }

    // Modern /{SET}/{LANG}/{NUM}/{slug} form. The set code is 2-6 alphanum
    // chars (e.g. BNG, MH2, 30A), lang is the bcp-47-ish locale tag used in
    // Gatherer URLs (e.g. en-us, ja-jp), and the collector number can
    // contain letters/digits/stars (e.g. 156, 42a, 1★).
    const match = u.pathname.match(/^\/([A-Za-z0-9]{2,6})\/([a-z]{2}(?:-[a-z]{2})?)\/([^/]+)(?:\/[^/]*)?$/);
    if (match) {
      const [, set, lang, number] = match;
      return { kind: 'set', set, number: decodeURIComponent(number), lang };
    }

    return null;
  } catch {
    return null;
  }
}

/** Returns true if the URL should be rendered as a rich embed rather than a plain link. */
export function isEmbeddableUrl(url: string): boolean {
  return !!extractYouTubeId(url) || !!extractTweetId(url) || !!extractBlueskyPost(url)
    || !!extractMastodonPost(url) || !!extractSpotifyEmbed(url) || !!extractRedditPost(url)
    || !!extractArchiveOrgId(url) || !!extractGathererCard(url);
}

/** Get a short label for the embed type. */
export function embedLabel(url: string): string | null {
  if (extractYouTubeId(url)) return 'YouTube';
  if (extractTweetId(url)) return 'Twitter';
  if (extractBlueskyPost(url)) return 'Bluesky';
  if (extractMastodonPost(url)) return 'Mastodon';
  if (extractSpotifyEmbed(url)) return 'Spotify';
  if (extractRedditPost(url)) return 'Reddit';
  if (extractArchiveOrgId(url)) return 'Internet Archive';
  if (extractGathererCard(url)) return 'Gatherer';
  return null;
}