Hashing

Overview

https://www.youtube.com/watch?v=_IyMIQvAiwc&index=10&list=PLCiOXwirraUD0G290WrVKpVYd3leGRRMW

https://www.youtube.com/watch?v=Kl70jizXl0s&list=PLCiOXwirraUD0G290WrVKpVYd3leGRRMW&index=11

TRC Video

https://www.youtube.com/watch?v=BJL-JxZDAQc

What is hashing

Hashing is the process of calculating a fixed length value from another value, such as a string. A value is passed into a Hash Function and the function returns a Hash Value. This process is used in cryptography. Rather than storing a 'secret' (such as a password) as plain text, it's hash value is calculated and stored in its place. Therefore, if the storage database was compromised and an attacker could see the stored passwords, they wouldn't know their actual value - only the hash. It is intentionally difficult to reverse the hashing process without the original parameters and function used to create it.

Hashing is also used for storing large quantities of data in a hashing table. The data is converted to an integer hash, which is then used to 'search' the table to fetch the relevant data. This has the advantage that finding an integer from a list of integers is much faster than searching for a string within a list of strings.

Key Terms

Hash Value

The Hash value is the value generated after passing the plain text through a hashing function.

Hash Key

The hash key is the specific modifier passed into the hash function in order to generate the returned value.

Hash Function

The hash function is the method in which plain text is converted to a hash value. It contains all the logic to alter the data into something that is more challenging to use.

Collisions

Collisions occur when two values, when entered into the same hashing function, return the same hash value. These are inevitable when dealing with huge quantities of values no matter how complex the function. The creator can take steps to reduce the number of collisions that will occur but they are impossible to prevent.

Handling Collisions

Rehashing / Closed

Linear rehashing would increment the hash value to the next available location. This will obviously require additional logic in your look up routine, because the value returned should be the location the data is stored. You would also need to check the nearest locations to see if the data collided with the current location.

Rehashing could also involve building a new hash table with a new algorithm and rehash all of the values using the new algorithm.

Linked List of Collided Values / Open

When you have a collision the value at that location could be linked to any values that have collided. This would essentially make the location a linked list of collided values.

You will have to be careful when removing a value from your hash table because removing could remove the whole list. Your logic will need to check each item in the list if collisions have occurred. When you delete an item you should remove it from the list if a list exists.

Example

PHP's crypt() function is a hashing function that supports multiple methods & standards, designed for password cryptography. They use a salt, which is additional data that is added to the value to be hashed in order to make it longer and more complex.

From PHP.net Manual on crypt:

1 echo 'SHA-512: ' . crypt('rasmuslerdorf', '$6$rounds=5000$usesomesillystringforsalt$') . "<br>";
2 Output: SHA-512: $6$rounds=5000$usesomesillystri$D4IrlXatmP7rx3P3InaxBeoomnAihCKRVQP22JZ6EY47Wc6BkroIuUUBOov1i.S5KPgErtP/EN5mcO.ChWQW21

The function hashes 'rasmuslerdorf' (creator of PHP) and outputs it, using the salt '$6$rounds=5000$usesomesillystringforsalt$'.