decode_msi_name.js

var decode_msi_name = (function() {
/*
NOTES:
- most entries start with 0x4840, which seems to be a marker
- corpus char code range 0x3864-0x4840
- range 0x3800-0x4840 exactly large enough to represent 1 or 2 char sequences
*/
  var CMAP = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz._";
  var decode_msi_char = function(x) { return CMAP[x]||"_"; };

  return function decode_msi_name(str) {
    var i = 0, o = [], c = str.charCodeAt(0);
    if(c == 0x4840) { ++i; o.push("*"); } /* technically '_' */
    for(; i < str.length; ++i) {
      c = str.charCodeAt(i);
      if(c < 0x3800 || c > 0x4840) o.push(String.fromCharCode(c)); /* raw */
      else if(c >= 0x4800) o.push(decode_msi_char(c-0x4800)); /* one char */
      else { /* two chars */
        o.push(decode_msi_char((c -= 0x3800) & 0x3f));
        o.push(decode_msi_char((c>>6) & 0x3f));
      }
    }
    return o.join("");
  };
})();