Utility to extract PDF info

pdfinfo.js

fs = require('fs');

function fromPdfString(s) {
  var t = '';
  for (var i = 0; i < s.length; i++) {
    if (s[i] != '\\') {
      t += s[i];
      continue;
    }
    if (s[i + 1] < "A") {
      if (s[i + 1] === '0' || s[i + 1] === '1' ||
          s[i + 1] === '2' || s[i + 1] === '3') {
        t += String.fromCharCode(parseInt(s.substring(i + 1, i + 4), 8));
        i += 3;
      } else {
        t += s[++i];
      }
      continue;
    }
    var j = s[i + 1] === 'x' ? i + 4 : i + 2;
    t += JSON.parse('"' + s.substring(i, j) + '"');
    i = j - 1;
  }
  if (t.charCodeAt(0) === 0xFE && t.charCodeAt(1) === 0xFF) {
    var u = '';
    for (var i = 2; i < t.length; i += 2)
      u += String.fromCharCode((t.charCodeAt(i) << 8) |
                               t.charCodeAt(i + 1));
    t = u;
  }
  return t;
}

process.argv.forEach(function (val, index, array) {
  if (index < 2) return;
  try {
    var data = fs.readFileSync(val, {encoding: 'binary'})
    var m  = /^%PDF-(\d\.\d)/.exec(data);
    var version = m ? m[1] : '';
    m = /\/Producer\s*\((.*?)\)\s*[\r\n\/>]/.exec(data);
    var producer = m ? fromPdfString(m[1]) : '';
    m = /\/Creator\s*\((.*?)\)\s*[\r\n\/>]/.exec(data);
    var creator = m ? fromPdfString(m[1]) : '';
    console.log(val + '\t' + version + '\t' + producer + '\t' + creator);
  } catch (ex) {}
});