Moving bits to alter characters

[Gaminic]

That's definitely C++ though.

No idea what the values in eTable and dTable mean, though.

[Texan40]

Sorry about that... this is why I never "work" as a programmer because my documentation skills are nil. Here's a repost with comments:

namespace MyCrypt
{
  const char *DEFAULT_KEY = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789!@#$%^&*()_+-=<>?,./\\|[]{}`~";
  const int eTable[] = {  85,170,-17, 68,153, 17,136, 34,-136, 51,-136, -51,-153, -51,-170 }; // split char value into 15 divisions, tables indicate value to offset
  const int dTable[] = { 136, 17,136,153,170,-85,-17,-68,  51,-34,  51,-170, -51,-153,-136 }; // same as above but reverses offset to original
  const float chapters = 17.0; // 15 * 17 = 255

  unsigned char cenc(const unsigned char cVal)
  {
    return (cVal + eTable[((int)ceil(cVal/chapters))-1]); // divide char val by 17, take ceiling, subtract 1 to get index in eTable
  }

  unsigned char cdec(const unsigned char cVal)
  {
    return (cVal + dTable[((int)ceil(cVal/chapters))-1]); // same as eTable, but for decryption
  }

  std::string encrypt(std::string s, std::string k = DEFAULT_KEY)
  {
    std::string sret;
    for(std::string::size_type i=0; i < s.size(); ++i)
    {
      unsigned char ci = s.at(i), ck = k.at(i % (k.size()-1)); // rotate key index for length of string
      sret += (cenc(ci)^ck); // get new value for s[i] from cenc(), XOR with key value from k[i]
    }
    return sret; // return encrypted string
  }

  std::string decrypt(std::string s, std::string k = DEFAULT_KEY)
  {
    std::string sret;
    for(std::string::size_type i=0; i < s.size(); ++i)
    {
      unsigned char ci = s.at(i), ck = k.at(i % (k.size()-1)); // rotate key index for length of string
      sret += cdec(ci^ck); // XOR s[i] with key value from k[i], get original value from cdec();
    }
    return sret; // return decrypted string
  }
};

[TheBeardedQuack]

Although I still don't quite understand the second still (are you changing the char in first half and then position in second half?) this is a lot more help with comments. :)

Also how did you find the dTable numbers to be opposite because they don't appear consistent. (I can tell that this looks like a fair encryption system that also shouldn't give unreliable results such as my current method of tripling the value)

[TheBeardedQuack]

STILL NOT ANSWERED

[Texan40]

The problem with "encryption" is that you want to eventually reverse it. This particular thing is pretty rudimentary and probably not all that hard to "crack." It was just a pet project, though. The key selection is based off a modulo to keep the key rotating even if the string length is longer than the key length. I originally made this to encrypt whole files not strings.

Example encrypt:

For the ascii value of 84 ('T')
84 / 17 = 4.94
ceil(4.94) = 5
5-1 = 4 (This is probably redundant as I could have just taken the floor of the divison?)
eTable[4] = 153
84 + 153 = 237
237 XOR 65 ('A' from DEFAULT_KEY[0]) = 172.
So cenc('T') = '¼' (ascii 172)

Example decrypt:

Input value of 172 ('¼')
172 XOR 65 (once again 'A' from DEFAULT_KEY[0]) = 237
237 / 17 = 13.94
ceil(13.94) = 14
14 - 1 = 13
dTable[13] = -153
237 - 153 = 84
so denc('¼' XOR 'A') = 'T' (ascii 84, our original character!)

It's only a one-off from basic XOR but for my purposes it was "enough" lol