Hashing (non-technical question)

[computation]

I'm a little unclear as to what hashing is exactly. I'm in the process of writing a word jumbling program and I came up with a scheme to compare a large number of words fairly quickly. I want to know if the following is a true form of hashing.

I got the idea from a letter frequency count algorithm. It seemed like there would be far too many for loops in it for me. So I came up with this. Instead of using arrays, I would use bit fields. I reserve 28 bits. 26 for each letter of the alphabet, and 2 unused bits just to make it a multiple of four (4bits/byte).

The word "cats" would be encoded like this:

**abcdefghijklmnopqrstuvwxyz
0010100000000000000011000000

Now, with a simple bitwise &, I can match all possible permutations AND lengths in one if(). (i.e. cats, cat, cast, a, act, at, sac, etc);

In case you want to tell me this won't work, forget about the actual frequency of each letter. All I'm concerned with here is if contains AT LEAST one occurence of the letter. It will still match no matter how many occurences there are, but will have to be verified after the match is made.

Would this be considered hashing or not?

[Salem]

I'd say it is.

[computation]

sorry for posting in the wrong section....i was originallyl going to ask a c++ question, but changed my mind.

so when i read about hashing and it says "formula", it doesn't necessarily mean a traditional mathematical formula? ( i say "traditional" because i guess technically everything about programming could *possibly* be considered a formula)

[Salem]

Anything which transforms a string (say "cats") into some numeric representation of that string is a hash function.

You're taking a potentially large and sparse space (the sequence of all possible letters which happen to be English words) and reducing it to a much smaller space, with possible collisions (your scheme for example collides on anagrams - maybe this is important, then again, maybe it isn't).