]*>/gi, "");
// Leere divs entfernen
html = html.replace(/]*>\s*<\/div>/gi, "");
return html;
};
const processFile = (filePath, outDir) => {
const raw = fs.readFileSync(filePath, "utf-8");
const fmMatch = raw.match(/^---\n([\s\S]*?)\n---\n([\s\S]*)$/);
if (!fmMatch) return null;
const fmBlock = fmMatch[1];
const htmlContent = fmMatch[2].trim();
const titleMatch = fmBlock.match(/title:\s*"(.+?)"/);
const slugMatch = fmBlock.match(/slug:\s*"(.+?)"/);
const excerptMatch = fmBlock.match(/excerpt:\s*"(.*)"/);
const title = titleMatch?.[1] ?? path.basename(filePath, ".html");
const slug = slugMatch?.[1] ?? path.basename(filePath, ".html");
const excerpt = excerptMatch?.[1] ?? "";
if (SKIP_SLUGS.includes(slug)) return null;
// Content extrahieren und bereinigen
const cleanHtml = extractContent(htmlContent);
// HTML -> Markdown
let md = nhm.translate(cleanHtml);
// Post-Cleanup
md = md
// WordPress-Kommentare entfernen
.replace(//g, "")
// Escaped newlines
.replace(/\\n/g, "")
// Einzelne Backslashes auf eigener Zeile
.replace(/^\s*\\\s*$/gm, "")
// Backslash am Zeilenanfang/-ende (Artefakte von
)
.replace(/^\\ /gm, "")
.replace(/ \\$/gm, "")
// Inline-Backslashes vor/nach Text
.replace(/\\\s{2,}/g, " ")
// Doppelte Leerzeilen reduzieren
.replace(/\n{3,}/g, "\n\n")
// WordPress-Nav Reste
.replace(/M0LZI[_]3D[\s\S]*?\[SYS_BOOT[^\]]*\]/g, "")
.replace(/\[WISSEN\].*\[FAQ\]/gs, "")
.replace(/JOIN_NETWORK.*?→/g, "")
.replace(/M0LZI[_]3D © \d{4}/g, "")
.replace(/\[WHATSAPP\].*?\[DATENSCHUTZ\].*?$/gm, "")
.trim();
// Kaputte Tabellen fixen: Zeilen die nur aus | \ \ \ | bestehen → entfernen
md = md.replace(/^\|[\s\\|]+\|$/gm, "");
// Leere Tabellenkoepfe entfernen
md = md.replace(/^\|\s*\|\s*$/gm, "");
// Nochmal doppelte Leerzeilen
md = md.replace(/\n{3,}/g, "\n\n");
const cat = categorize(slug);
const diff = difficulty(slug, cat);
const frontmatter = `---\ntitle: "${title}"\nslug: "${slug}"\ncategory: "${cat}"\ndifficulty: "${diff}"\nexcerpt: "${excerpt}"\n---`;
const output = `${frontmatter}\n\n${md}\n`;
const outPath = path.join(outDir, `${slug}.md`);
fs.writeFileSync(outPath, output, "utf-8");
return slug;
};
// Main
console.log("=== Re-Konvertierung aus Original-HTML ===\n");
console.log("Posts -> Guides:");
const postFiles = fs.readdirSync(BACKUP_POSTS).filter(f => f.endsWith(".html"));
let ok = 0;
for (const file of postFiles) {
const result = processFile(path.join(BACKUP_POSTS, file), OUT_GUIDES);
if (result) { console.log(` OK: ${result}`); ok++; }
}
console.log(`\n${ok} Guides konvertiert.\n`);
console.log("Pages:");
const pageFiles = fs.readdirSync(BACKUP_PAGES).filter(f => f.endsWith(".html"));
let pOk = 0;
for (const file of pageFiles) {
const result = processFile(path.join(BACKUP_PAGES, file), OUT_PAGES);
if (result) { console.log(` OK: ${result}`); pOk++; }
}
console.log(`\n${pOk} Pages konvertiert.`);
// Verifizierung
console.log("\n=== Verifizierung ===");
const guideFiles = fs.readdirSync(OUT_GUIDES).filter(f => f.endsWith(".md"));
let issues = 0;
for (const file of guideFiles) {
const content = fs.readFileSync(path.join(OUT_GUIDES, file), "utf-8");
const backslashLines = (content.match(/^\s*\\\s*$/gm) || []).length;
const brokenTables = (content.match(/\| \\ \\ \\/g) || []).length;
if (backslashLines > 0 || brokenTables > 0) {
console.log(` ISSUE: ${file} (${backslashLines} backslash-lines, ${brokenTables} broken tables)`);
issues++;
}
}
console.log(issues === 0 ? " Alle Guides sauber!" : ` ${issues} Guides mit Problemen`);