import { execFile } from 'node:child_process';
import { promisify } from 'node:util';
import { extname } from 'node:path';
import type { ImageRecognizer } from './pipeline.js';

const run = promisify(execFile);

// Optional adapter: requires a separately installed Tesseract executable.
// Passes arguments directly (no shell), enforces a timeout and output cap.
export const tesseractAdapter: ImageRecognizer = async imagePath => {
  if (!['.png', '.jpg', '.jpeg', '.tif', '.tiff'].includes(extname(imagePath).toLowerCase())) {
    throw new Error('Only image files can be sent to the OCR adapter');
  }
  const { stdout } = await run('tesseract', [imagePath, 'stdout', '-l', 'eng'], {
    timeout: 30_000,
    maxBuffer: 2_000_000,
    windowsHide: true,
  });
  return stdout;
};