src/lib/scraper/apify.ts

/**
 * Apify Client Wrapper
 * Simplified interface for running Apify actors and handling results
 */

import { ApifyClient } from 'apify-client';

const apifyClient = new ApifyClient({
  token: process.env.APIFY_API_KEY,
});

export interface ScrapingResult {
  success: boolean;
  content?: string;
  title?: string;
  description?: string;
  images?: string[];
  metadata?: Record<string, unknown>;
  error?: string;
}

/**
 * Run Apify actor and get results
 */
export async function runApifyActor(
  actorId: string,
  input: Record<string, unknown>
): Promise<ScrapingResult> {
  try {
    if (!process.env.APIFY_API_KEY) {
      return {
        success: false,
        error: 'Apify API key not configured',
      };
    }

    // Start actor run
    const run = await apifyClient.actor(actorId).call(input);

    // Get dataset results
    const dataset = await apifyClient.dataset(run.defaultDatasetId).listItems();

    if (!dataset.items || dataset.items.length === 0) {
      return {
        success: false,
        error: 'No data returned from actor',
      };
    }

    // Parse results based on actor type
    const item = dataset.items[0] as Record<string, unknown>;

    return {
      success: true,
      content: extractText(item),
      title: (item.title as string) || (item.name as string) || undefined,
      description: (item.description as string) || undefined,
      images: extractImages(item),
      metadata: item,
    };
  } catch (error) {
    console.error('[apify] actor failed:', error);
    return {
      success: false,
      error: error instanceof Error ? error.message : 'Actor execution failed',
    };
  }
}

/**
 * Extract text content from actor result
 */
function extractText(item: Record<string, unknown>): string {
  // Try common text fields
  if (typeof item.text === 'string') return item.text;
  if (typeof item.content === 'string') return item.content;
  if (typeof item.description === 'string') return item.description;
  if (typeof item.caption === 'string') return item.caption;

  // Try combining multiple fields
  const parts: string[] = [];
  if (typeof item.title === 'string') parts.push(item.title);
  if (typeof item.subtitle === 'string') parts.push(item.subtitle);
  if (typeof item.body === 'string') parts.push(item.body);

  return parts.join('\n\n');
}

/**
 * Extract image URLs from actor result
 */
function extractImages(item: Record<string, unknown>): string[] {
  const images: string[] = [];

  // Check common image fields
  const imageFields = ['image', 'images', 'src', 'imageUrl', 'photo', 'photos'];
  for (const field of imageFields) {
    const value = item[field];
    if (typeof value === 'string') {
      images.push(value);
    } else if (Array.isArray(value)) {
      for (const img of value) {
        if (typeof img === 'string') {
          images.push(img);
        } else if (typeof img === 'object' && img !== null) {
          const url = (img as Record<string, unknown>).url;
          if (typeof url === 'string') {
            images.push(url);
          }
        }
      }
    }
  }

  return Array.from(new Set(images)); // Remove duplicates
}

/**
 * Popular actor IDs (ready-to-use)
 */
export const ACTOR_IDS = {
  instagram: 'apify/instagram-scraper',
  twitter: 'apidojo/tweet-scraper',
  tiktok: 'clockworks/tiktok-scraper',
  threads: 'apify/instagram-scraper', // Threads can be scraped via Instagram's API
  web: 'apify/website-content-crawler',
};