I have a method that turns a byte array into an integer
public int Encode(string input)
{
var bytes = Encoding.Unicode.GetBytes(input.ToLowerInvariant());
return BitConverter.ToInt64(bytes,0);
}
Why is this integer not different for any input string?
For example
input = "http://www.google.com => 31525695615402088
and
input = "http://www.microsoft.com => 31525695615402088
Because 64 bits is 8 bytes, and so ToInt64 consumes only the first 8 bytes of the input array. What are the first eight bytes of the strings you've used?
And, as alexm notes, Encoding.Unicode specifies UTF-16, in which each character is actually two bytes (usually), so only the first 4 characters count.
'h' == 0x68
't' == 0x74
'p' == 0x70
Little endian, two-byte characters, so "http" gives you an array that starts with :
{ 0x68, 0x00, 0x74, 0x00, 0x74, 0x00, 0x70, 0x00 ...
Interpret this as a little-endian 32-bit integer, and you get:
0x0070007400740068
Which, of course is equal to 31525695615402088
An int64 is 8 bytes. I'm sure you can figure it out from there.
This occurs because a 64-bit integer uses 8-bytes of memory, and BitConverter will only convert using the first 8 bytes of the byte array you specified, starting from position 0. Each sample input you provided starts with the same 8 bytes.
For what it's worth, it's not possible to perform loss-less encoding of a string of variable length into an integer data type with a size of 4 - 8 bytes. You may be looking for a hashing algorithm that represents your data in a finite number of bytes.
Well, ToInt64 uses 8 bytes - that's 4 unicode characters.
Because BitConverter.ToInt64 takes only the first 8 bytes of your byte array which are the same for your strings. Try input strings "google.com" and "yahoo.com".
