I have some encoded strings which are probably encoded in a base64-variant - except that there aren't only 64 chars, but all the (printable) ASCII chars between 33 and 126 (0x21 to 0x7E).

I've searched for other base64-variants but only found variants like base85 which excludes some characters that may cause problems.

But in my case, the string also contains those characters like ` (backtick), {, }, [, and all the others. Every printable ASCII char.

Here's an example of one of the encoded strings:


Any way to decode such a string without knowing what encoding was used? Does anybody know about an encoding like base64 which uses every printable ASCII char?

  • $\begingroup$ Such encodings can be produced by any large arithmetics library that supports conversion of octet strings to integers and of integers to arbitrary base representation. It is however not on topic here. $\endgroup$ – Henrick Hellström Oct 5 '15 at 22:17
  • $\begingroup$ can you determine the entire alphabet space from a large number of strings? $\endgroup$ – Richie Frame Oct 5 '15 at 23:20
  • $\begingroup$ Those strings can contain any printable ASCII char - so, from 0x21 ("!") to 0x7E ("~"). $\endgroup$ – Florian Bach Oct 6 '15 at 6:59
  • $\begingroup$ There are several less used and/or non-standard baseN encodings like base91 and base94. $\endgroup$ – otus Oct 6 '15 at 7:03
  • $\begingroup$ I have tested both base91 and base94, no success. I begin to doubt if this even is some baseN encoding ... $\endgroup$ – Florian Bach Oct 6 '15 at 8:15

Browse other questions tagged or ask your own question.