cortex/src/core/import/markdown.ts

import * as fs from 'fs';
import * as path from 'path';
import { addNode } from '../store';
import { NodeKind } from '../../types';

export interface MarkdownImportOptions {
  kind?: NodeKind;
  tags?: string[];
  dryRun?: boolean;
}

export interface MarkdownImportResult {
  imported: number;
  files: string[];
}

export async function importMarkdown(folderPath: string, options: MarkdownImportOptions = {}): Promise<MarkdownImportResult> {
  const absPath = path.resolve(folderPath);

  if (!fs.existsSync(absPath)) {
    throw new Error(`Folder does not exist: ${absPath}`);
  }

  const files = findMarkdownFiles(absPath);
  const defaultKind = options.kind || 'memory';
  const defaultTags = options.tags || [];

  if (options.dryRun) {
    return {
      imported: files.length,
      files: files.map(f => path.relative(absPath, f)),
    };
  }

  let imported = 0;
  const importedFiles: string[] = [];

  for (const file of files) {
    const content = fs.readFileSync(file, 'utf-8');
    const relativePath = path.relative(absPath, file);
    const title = path.basename(file, '.md');

    // Extract title from first H1 if present
    const h1Match = content.match(/^#\s+(.+)$/m);
    const nodeTitle = h1Match ? h1Match[1] : title;

    // Remove the H1 from content if it was used as title
    const nodeContent = h1Match ? content.replace(/^#\s+.+\n*/, '') : content;

    await addNode({
      kind: defaultKind,
      title: nodeTitle,
      content: nodeContent.trim(),
      tags: ['imported', 'markdown', ...defaultTags],
      metadata: {
        importedFrom: 'markdown',
        originalPath: relativePath,
        importedAt: Date.now(),
      },
    });

    imported++;
    importedFiles.push(relativePath);
  }

  return {
    imported,
    files: importedFiles,
  };
}

function findMarkdownFiles(dir: string): string[] {
  const files: string[] = [];

  function walk(currentDir: string) {
    const entries = fs.readdirSync(currentDir, { withFileTypes: true });

    for (const entry of entries) {
      const fullPath = path.join(currentDir, entry.name);

      if (entry.name.startsWith('.')) continue;

      if (entry.isDirectory()) {
        walk(fullPath);
      } else if (entry.isFile() && entry.name.endsWith('.md')) {
        files.push(fullPath);
      }
    }
  }

  walk(dir);
  return files;
}