Encoding, Compression

Question

I wanted to know whether

• A file encryption algorithm can be designed in such a way that it can perform the activities of a file compression as well (any live example?).
• Also can I integrate it to a Mobile SMS Service, I mean for text messages?

Also I wanted to know about binary files...if a plain text file is encoded in binary does its size reduces? Also is it better to encode a plain text file to a binary one rather than encoding it to any other format (in case if anyone wants it to encode for any purpose)

Answer 1

In fact all decent encryption programs (take PGP for example) compress data before encryption. They use something mainstream like ZIP for compression. The reason is once data is encrypted it looks like random noise and becomes incompressible, so it can only be compressed before encryption. You likely can't do that for SMS - you have to obey the specifications for the SMS so you'd better check those specifications.

Answer 2

First question: No reason why it shouldn't be possible. Since you can still encrypt compressed data and vise versa, you can test the benefits by trying. I don't know if it can be used for text messaging, but you'd have to explain what you're trying to do if anyone wants to give a reasonable and fitting answer.

Second question: plain text is still binary, it's just readable and encoded in a certain character set. Main difference is that plain text usually conforms to a certain encoding, say UTF-8 or ISO-8859-1.

You can still compress plain text, which makes it "binary" in the traditional sense that you only see weird bytes. :-)

Answer 3

Compression removes redundant information. Redundant information makes breaking an encryption easier. So yes, encryption and compression are compatible. I don't know if there is an algorithm designed to do both things though.

Yes, a binary file will usually be smaller than a plain text file. For instance, the number 34 written out in text takes 2 bytes (at least) whereas in those same 2 bytes you could write a number as large as 65000.

What makes an encoding "better" than another is the purpose to which it is put. If you are optimizing for size, binary is probably better. If you are optimizing for readability or graceful failure, text might be better.