#!/usr/bin/env node

const fs = require('fs');
const path = require('path');
const { glob } = require('glob');

/**
 * Clean MDX files by removing unnecessary HTML wrappers and data attributes
 */
async function cleanMDXFiles() {
  const mdxFiles = await glob('data/blog/**/*.mdx');
  
  console.log(`Found ${mdxFiles.length} MDX files to clean...`);
  
  for (const filePath of mdxFiles) {
    try {
      let content = fs.readFileSync(filePath, 'utf-8');
      const originalContent = content;
      
      // Remove all data-start and data-end attributes
      content = content.replace(/\s*data-start="[^"]*"/g, '');
      content = content.replace(/\s*data-end="[^"]*"/g, '');
      
      // Remove all class="" attributes
      content = content.replace(/\s*class="[^"]*"/g, '');
      
      // Remove all dir="auto" attributes
      content = content.replace(/\s*dir="auto"/g, '');
      
      // Remove all data-message-* attributes
      content = content.replace(/\s*data-message-[^=]*="[^"]*"/g, '');
      
      // Remove unnecessary wrapper divs (gizmo, flex containers, etc.)
      content = content.replace(/<div class="flex-shrink-0[^>]*>[\s\S]*?<div class="markdown prose[^>]*>/g, '');
      content = content.replace(/<div class="flex max-w-full[^>]*>[\s\S]*?<div class="markdown prose[^>]*>/g, '');
      content = content.replace(/<\/div>\s*<\/div>\s*<\/div>\s*<\/div>\s*<\/div>\s*<\/div>\s*<\/div>\s*<\/div>\s*<\/div>/g, '');
      content = content.replace(/<\/div>\s*<\/div>\s*<\/div>\s*<\/div>/g, '');
      
      // Clean up h2 tags - remove attributes but keep content
      content = content.replace(/<h2[^>]*>(.*?)<\/h2>/g, '## $1');
      
      // Clean up p tags - remove attributes but keep content
      content = content.replace(/<p[^>]*>(.*?)<\/p>/g, (match, p1) => {
        // If it's just whitespace, skip it
        if (!p1.trim()) return '';
        return p1.trim() + '\n';
      });
      
      // Clean up li tags
      content = content.replace(/<li[^>]*>\s*<p[^>]*>(.*?)<\/p>\s*<\/li>/g, '- $1');
      content = content.replace(/<li[^>]*>(.*?)<\/li>/g, '- $1');
      
      // Clean up ul tags
      content = content.replace(/<ul[^>]*>/g, '');
      content = content.replace(/<\/ul>/g, '');
      
      // Clean up strong tags - remove data attributes
      content = content.replace(/<strong[^>]*>(.*?)<\/strong>/g, '**$1**');
      
      // Clean up anchor tags - remove data attributes and nofollow
      content = content.replace(/<a href="([^"]*)"[^>]*rel="noopener noreferrer nofollow"[^>]*>(.*?)<\/a>/g, '[$2]($1)');
      content = content.replace(/<a href="([^"]*)"[^>]*>(.*?)<\/a>/g, '[$2]($1)');
      
      // Convert Visual Link Preview HTML to component
      content = content.replace(
        /<div class="vlp-link-container[^>]*>[\s\S]*?<a href="([^"]*)"[^>]*title="([^"]*)"[^>]*>[\s\S]*?<img[^>]*src="([^"]*)"[^>]*>[\s\S]*?<div class="vlp-block-0[^>]*>([^<]*)<\/div>[\s\S]*?<div class="vlp-block-1[^>]*>([^<]*)<\/div>[\s\S]*?<\/div>/g,
        (match, url, title, image, blockTitle, summary) => {
          return `\n<VisualLinkPreview \n  url="${url}"\n  title="${blockTitle || title}"\n  summary="${summary}"\n  image="${image}"\n/>\n`;
        }
      );
      
      // Remove empty lines (more than 2 consecutive)
      content = content.replace(/\n{3,}/g, '\n\n');
      
      // Ensure frontmatter has excerpt and category if missing
      const frontmatterMatch = content.match(/^---\n([\s\S]*?)\n---/);
      if (frontmatterMatch) {
        let frontmatter = frontmatterMatch[1];
        
        // Add excerpt if missing
        if (!frontmatter.includes('excerpt:')) {
          // Try to extract first paragraph as excerpt
          const firstParagraph = content.match(/---\n\n(.*?)\n\n/s);
          if (firstParagraph) {
            const excerpt = firstParagraph[1]
              .replace(/[#*\[\]]/g, '')
              .replace(/\n/g, ' ')
              .trim()
              .substring(0, 200);
            frontmatter += `\nexcerpt: ${excerpt}`;
          }
        }
        
        // Add category if missing (try to infer from content or use default)
        if (!frontmatter.includes('category:')) {
          frontmatter += `\ncategory: Kabel Technologie`;
        }
        
        content = content.replace(/^---\n[\s\S]*?\n---/, `---\n${frontmatter}\n---`);
      }
      
      // Only write if content changed
      if (content !== originalContent) {
        fs.writeFileSync(filePath, content, 'utf-8');
        console.log(`✓ Cleaned: ${filePath}`);
      } else {
        console.log(`- Skipped (no changes): ${filePath}`);
      }
      
    } catch (error) {
      console.error(`✗ Error processing ${filePath}:`, error.message);
    }
  }
  
  console.log('\nDone!');
}

cleanMDXFiles().catch(console.error);