Duplicate the last char in Plain text

[mailhog]

Duplicate the last char in Plain text

[scollon42]

Seems that

function decodeQuotedPrintableHelper

return NaN for the last prefix = after Merci.
and when the result go to

function convertBytesToUnicodeCodePoints

it will duplicate the previous bytes while where there is a NaN
for exemple

var ret = decodeQuotedPrintableHelper("BLACKLISTED WAKE UP SMS SENT ([phone]) : Votre code de v=C3=A9rifica=tion Koolicar : 140035=", "=")

ret /*=> [66, 76, 65, 67, 75, 76, 73, 83, 84, 69, 68, 32, 87, 65, 75, 69, 32, 85, 80, 32, 83, 77, 83, 32, 83, 69, 78, 84, 32, 40, 43, 51, 51, 54, 49, 50, 51, 52, 53, 54, 55, 56, 41, 32, 58, 32, 86, 111, 116, 114, 101, 32, 99, 111, 100, 101, 32, 100, 101, 32, 118, 195, 169, 114, 105, 102, 105, 99, 97, NaN, 111, 110, 32, 75, 111, 111, 108, 105, 99, 97, 114, 32, 58, 32, 49, 52, 48, 48, 51, 53, NaN] */

convertBytesToUnicodeCodePoints(ret, "UTF8") /*=> [66, 76, 65, 67, 75, 76, 73, 83, 84, 69, 68, 32, 87, 65, 75, 69, 32, 85, 80, 32, 83, 77, 83, 32, 83, 69, 78, 84, 32, 40, 43, 51, 51, 54, 49, 50, 51, 52, 53, 54, 55, 56, 41, 32, 58, 32, 86, 111, 116, 114, 101, 32, 99, 111, 100, 101, 32, 100, 101, 32, 118, 233, 114, 105, 102, 105, 99, 97, 97, 111, 110, 32, 75, 111, 111, 108, 105, 99, 97, 114, 32, 58, 32, 49, 52, 48, 48, 51, 53, 53] */

I think for this last char in decodeQuotedPrintableHelper function there is an edge case.
here prefix for quoted printable is = the function will go through this condition :

if (str.charAt(i) == prefix ) {
  ret= decoded_bytes.push(parseInt(str.substr(i + 1, 2), 16));
  i += 3;
}

but here :

parseInt(str.substr(i + 1, 2), 16)
/* i + 1 == str.length  so str.substr will return an empty string and parseInt will return NaN */

This can fix the NaN case :

function decodeQuotedPrintableHelper(str, prefix) {
  var decoded_bytes = [];
  for (var i = 0; i < str.length;) {
    if (str.charAt(i) == prefix) {

     // here we check if it's not the end of the str
      if (i + 1 < str.length) {
      	decoded_bytes.push(parseInt(str.substr(i + 1, 2), 16));
      }

      i += 3;
    } else {
      decoded_bytes.push(str.charCodeAt(i));
	  ++i;
    }
  }
  return decoded_bytes;
}

hope this can help 😃

[scollon42]

it should be fixed by this Pull Request: #33
I did a quick fix so don't hesitate to challenge it and review the code 😄

[TheMatrix97]

I have the same issue, using encoding cp1252.
plain text: "enlla seggent"
Source code: "l'enlla=E7 seg=FCent"
@scollon42 How could I include this modification to a running docker container?