extract-statement.js

import { spawn } from 'child_process';
import { resolve } from 'path';
import { fileURLToPath } from 'url';
import { parseTransactions, extractPhone, extractSender } from './lib/statement.js';

const jarPath = fileURLToPath(new URL('./node_modules/tabula-js/lib/tabula-java.jar', import.meta.url));

export async function runTabula({
  pdfPath = './statement.pdf',
  format = 'TSV',
  pages = 'all',
  guess = false,
} = {}) {
  const resolvedPdf = resolve(pdfPath);
  const args = ['-jar', jarPath, '--format', format, '--pages', pages];
  if (guess) args.push('--guess');
  args.push(resolvedPdf);

  return new Promise((resolvePromise, reject) => {
    const proc = spawn('java', args, { cwd: process.cwd() });
    let stdout = '';
    let stderr = '';
    proc.stdout.setEncoding('utf8');
    proc.stdout.on('data', chunk => (stdout += chunk));
    proc.stderr.setEncoding('utf8');
    proc.stderr.on('data', chunk => (stderr += chunk));
    proc.on('error', reject);
    proc.on('close', (code, signal) => {
      if (code !== 0 || signal) {
        const reason = signal ? `signal ${signal}` : `exit ${code}`;
        reject(new Error(`tabula exited (${reason}): ${stderr.trim()}`));
        return;
      }
      resolvePromise(stdout);
    });
  });
}

async function main() {
  try {
    const pdfPath = process.argv[2] ?? './statement.pdf';
    const tsv = await runTabula({ pdfPath });
    console.log('Raw extracted TSV (tabula output):');
    console.log(tsv.trim() || '(empty)');
    const rows = parseTransactions(tsv);
    const filtered = rows.filter(row => {
      const detail = (row.details ?? '').toLowerCase();
      return /mobi\s*522522/i.test(detail);
    });
    console.log(`Extracted ${rows.length} rows from ${pdfPath}`);
    const enriched = filtered.map(row => ({
      ...row,
      phone: extractPhone(row.details),
      sender: extractSender(row.details),
    }));
    console.log(`Kept ${filtered.length} rows that mention 522522 in their details`);
    console.log(JSON.stringify(enriched, null, 2));
  } catch (error) {
    console.error('Unable to extract rows:', error.message);
    process.exit(1);
  }
}

const entryScript = fileURLToPath(process.argv[1] ? new URL(process.argv[1], 'file://') : '');

if (process.argv[1] && entryScript === fileURLToPath(import.meta.url)) {
  await main();
}