Server : Apache/2 System : Linux server-15-235-50-60 5.15.0-164-generic #174-Ubuntu SMP Fri Nov 14 20:25:16 UTC 2025 x86_64 User : gositeme ( 1004) PHP Version : 8.2.29 Disable Function : exec,system,passthru,shell_exec,proc_close,proc_open,dl,popen,show_source,posix_kill,posix_mkfifo,posix_getpwuid,posix_setpgid,posix_setsid,posix_setuid,posix_setgid,posix_seteuid,posix_setegid,posix_uname Directory :  /home/gositeme/domains/lavocat.quebec/private_html/scripts/

Upload File : current_dir [ Writeable ] document_root [ Writeable ]

const axios = require('axios');
const cheerio = require('cheerio');
const fs = require('fs');
const path = require('path');
const https = require('https');

const BASE_URL = 'https://www.adwavocats.com';
const TEAM_URL = BASE_URL + '/a-propos/';
const OUTPUT_DIR = path.join(__dirname, '../public/images/lawyers');

const expectedLawyers = [
  'justin-wee',
  'alain-arsenault',
  'audrey-labrecque',
  'marie-claude-tremblay',
  'david-chen',
  'sophie-dubois',
  'marc-andre-bouchard',
  'isabella-rodriguez',
  'thomas-leblanc',
  'virginie-dufresne-lemire',
  'jerome-aucoin',
  'antoine-duranleau-hendrickx',
  'ivan-lazarov',
  'yalda-machouf-khadir',
  'olivia-malenfant',
  'imane-melab',
  'justine-monty',
  'mmah-nora-toure'
];

function slugify(str) {
  return str
    .toLowerCase()
    .replace(/é/g, 'e')
    .replace(/è/g, 'e')
    .replace(/ê/g, 'e')
    .replace(/à/g, 'a')
    .replace(/ç/g, 'c')
    .replace(/ô/g, 'o')
    .replace(/î/g, 'i')
    .replace(/ï/g, 'i')
    .replace(/'/g, '')
    .replace(/[^a-z0-9]+/g, '-')
    .replace(/^-+|-+$/g, '');
}

async function downloadImage(url, filename) {
  const filePath = path.join(OUTPUT_DIR, filename);
  const file = fs.createWriteStream(filePath);
  return new Promise((resolve, reject) => {
    https.get(url, (response) => {
      response.pipe(file);
      file.on('finish', () => {
        file.close(resolve);
      });
    }).on('error', (err) => {
      fs.unlink(filePath, () => {});
      reject(err.message);
    });
  });
}

(async () => {
  if (!fs.existsSync(OUTPUT_DIR)) {
    fs.mkdirSync(OUTPUT_DIR, { recursive: true });
  }
  const { data: html } = await axios.get(TEAM_URL);
  const $ = cheerio.load(html);
  const found = {};
  // 1. Try to get images from <img> tags
  $('img').each((i, el) => {
    const alt = $(el).attr('alt') || '';
    let src = $(el).attr('src') || '';
    let srcset = $(el).attr('srcset') || '';
    let dataSrc = $(el).attr('data-src') || '';
    let name = alt.replace(/(Me |Avocat|Avocate|Me\.|Mme\.|M\.|\(.*\))/gi, '').trim();
    const slug = slugify(name);
    let url = '';
    if (srcset && srcset.includes('.webp')) {
      url = srcset.split(',').find(s => s.includes('.webp')).split(' ')[0];
    } else if (src.endsWith('.webp')) {
      url = src;
    } else if (dataSrc.endsWith('.webp')) {
      url = dataSrc;
    } else if (src) {
      url = src;
    }
    if (slug && url && url.match(/wp-content.*\.(webp|jpg|jpeg|png)$/i)) {
      found[slug] = url.startsWith('http') ? url : BASE_URL + url;
    }
  });
  // 2. Fallback: search for image URLs in raw HTML for any missing
  for (const slug of expectedLawyers) {
    if (!found[slug]) {
      // Try to find a matching image URL in the HTML
      const regex = new RegExp(`wp-content[^"']*${slug}[^"']*\\.(webp|jpg|jpeg|png)`, 'i');
      const match = html.match(regex);
      if (match) {
        found[slug] = match[0].startsWith('http') ? match[0] : BASE_URL + '/' + match[0].replace(/^\//, '');
      }
    }
  }
  // 3. Download all found images
  for (const slug of expectedLawyers) {
    if (found[slug]) {
      const ext = path.extname(found[slug].split('?')[0]);
      const filename = slug + ext;
      try {
        await downloadImage(found[slug], filename);
        console.log(`Downloaded: ${filename}`);
      } catch (e) {
        console.error(`Failed to download ${filename}:`, e);
      }
    } else {
      console.log(`No image found for: ${slug}`);
    }
  }
})();