micromelon-website/scripts/fix-all-mdx.mjs

import fs from 'fs';
import path from 'path';
import { fileURLToPath } from 'url';

const __dirname = path.dirname(fileURLToPath(import.meta.url));
const dir = path.join(__dirname, '..', 'content', 'resources');

let fixed = 0;
for (const f of fs.readdirSync(dir).filter(f => f.endsWith('.mdx'))) {
  const filepath = path.join(dir, f);
  const original = fs.readFileSync(filepath, 'utf8');

  // Split frontmatter from content
  const secondDash = original.indexOf('---', 4);
  if (secondDash === -1) continue;
  let frontmatter = original.substring(0, secondDash + 3);
  let content = original.substring(secondDash + 3);

  // === Fix body content ===

  // Remove BACK TO POSTS and OPEN PAGE AS PDF
  content = content.replace(/\[\s*\*?\*?\s*BACK TO POSTS\s*\*?\*?\s*\]\([^)]*\)/gi, '');
  content = content.replace(/\*?\*?\s*BACK TO POSTS\s*\*?\*?/gi, '');
  content = content.replace(/\|?\s*\[?\s*\*?\*?\s*OPEN PAGE AS PDF\s*\*?\*?\s*\]?\s*\([^)]*\)/gi, '');

  // Remove CSS block definitions
  content = content.replace(/#block-[a-zA-Z0-9_-]+ \{[^}]*?\}/g, '');

  // Remove --> artifacts
  content = content.replace(/^-->\s*$/gm, '');

  // Escape ALL < that aren't part of:
  // 1. Markdown images: ![...](...) - these don't have <
  // 2. iframe tags: <iframe ... > or </iframe>
  // 3. Already escaped: \<
  // Strategy: protect safe tags, escape everything else, restore safe tags

  // Temporarily replace safe tags with placeholders
  const safeTags = [];
  content = content.replace(/<(\/?)iframe([^>]*)>/gi, (match) => {
    safeTags.push(match);
    return `__SAFE_TAG_${safeTags.length - 1}__`;
  });
  content = content.replace(/<(\/?)video([^>]*)>/gi, (match) => {
    safeTags.push(match);
    return `__SAFE_TAG_${safeTags.length - 1}__`;
  });
  content = content.replace(/<source([^>]*)>/gi, (match) => {
    safeTags.push(match);
    return `__SAFE_TAG_${safeTags.length - 1}__`;
  });

  // Now escape all remaining < and > that aren't in image markdown
  // Split by lines to handle image markdown
  const lines = content.split('\n');
  for (let i = 0; i < lines.length; i++) {
    const line = lines[i];
    // Skip lines that are image markdown
    if (line.trim().startsWith('![')) continue;
    // Skip lines that only contain safe tag placeholders
    if (line.trim().match(/^__SAFE_TAG_\d+__$/)) continue;

    // Escape < and > that aren't already escaped or part of placeholders
    lines[i] = line.replace(/<(?!_SAFE_TAG)/g, (match, offset) => {
      // Check if already escaped
      if (offset > 0 && line[offset - 1] === '\\') return match;
      return '&lt;';
    });
    lines[i] = lines[i].replace(/>(?!_)/g, (match, offset) => {
      const before = lines[i].substring(0, offset);
      // Don't escape if it's after a safe tag placeholder
      if (before.match(/__SAFE_TAG_\d+$/)) return match;
      // Don't escape if already escaped
      if (offset > 0 && lines[i][offset - 1] === '\\') return match;
      // Don't escape markdown blockquotes at start of line
      if (before.match(/^\s*$/)) return match;
      return '&gt;';
    });
  }
  content = lines.join('\n');

  // Restore safe tags
  for (let i = 0; i < safeTags.length; i++) {
    content = content.replace(`__SAFE_TAG_${i}__`, safeTags[i]);
  }

  // Clean excessive blank lines
  content = content.replace(/\n{3,}/g, '\n\n');

  // === Fix frontmatter excerpt ===
  const excerptMatch = frontmatter.match(/^excerpt: "(.*)"$/m);
  if (excerptMatch) {
    let excerpt = excerptMatch[1];
    excerpt = excerpt.replace(/\*?\*?BACK TO POSTS\*?\*?/gi, '');
    excerpt = excerpt.replace(/!\[[^\]]*\]\([^)]*\)/g, '');
    excerpt = excerpt.replace(/\[([^\]]*)\]\([^)]*\)/g, '$1');
    excerpt = excerpt.replace(/\*\*/g, '');
    excerpt = excerpt.replace(/#block-[^\s]+ \{[^}]*?\}/g, '');
    excerpt = excerpt.replace(/\s*\|\s*/g, ' ');
    excerpt = excerpt.replace(/\s+/g, ' ').trim();
    excerpt = excerpt.substring(0, 250);
    frontmatter = frontmatter.replace(excerptMatch[0], `excerpt: "${excerpt}"`);
  }

  const result = frontmatter + content;
  if (result !== original) {
    fs.writeFileSync(filepath, result);
    fixed++;
  }
}
console.log(`Fixed ${fixed} files`);