Quickly finding duplicates in an ArrayList

[paradox]

Basically I have an ArrayList of strings; I need a fast way of
searching through and comparing the elements of the ArrayList and the
return an ArrayList of items that have 3 Duplicates.

For example, if the ArrayList has the following elements:

elems[0] = "blue"
elems[1] = "red"
elems[2] = "blue"
elems[3] = "green"
elems[4] = "red"
elems[5] = "red"

I want to have a function that goes through the list and then returns a
new arraylist like:

newElems[0] = "red"
newElems[1] = "red"
newElems[2] = "red"

In actuality my arraylist is an array of objects, which have a string
property and I'm referencing like: elems[0].filename = "red"; but the
above example is easier to follow.

I would like to be able to perform this search as fast as possible. I
personally don't have any real experience or familiarity with topics
such as binary trees, etc; would something like that be the route to
take? Thanks.

[paradox]

Yeah, I did google it immediately after posting the question and was
able to find sufficient information. If I could of deleted the post, I
would have. ;)

I do have a small question though. Does the .NET Hashtable work in
exactly the same manner as a general Hashtable? Also, I tried using
reflector to track down the GetHashCode function to see how it worked;
I ended up at the String class (taking into consideration the type of
key I was using was a string), and the code I found for the GetHashCode
function was:

[MethodImpl(Meth odImplOptions.I nternalCall)]
public override extern int GetHashCode();

I couldn't figure out where to find anymore info...

[Brian Gideon]

The GetHashCode method is a virtual member of an Object so everything
has a default implementation for it. The string's version of the
method is internal to the .NET framework. That's why you can't see it.
One of the most frequently used ways of getting the hash code of an
object is to XOR all of the object's fields. In the case of a string
you might XOR all characters together. I seriously doubt that's what
the framework is doing though. XOR works well because it is fast and
produces a very random distribution of values. That's important when
it comes time to map the hash code to a slot in the memory array. A
good distribution of hash codes lowers the probability that two keys
will get mapped to the same slot in memory. This is called a
collision. There are many strategies for handling collisions. I
believe the Hashtable in dotnet uses quadratic probing. This strategy
involves number theory and the use of prime numbers.

paradox wrote:

Yeah, I did google it immediately after posting the question and was
able to find sufficient information. If I could of deleted the post,
I would have. ;)

I do have a small question though. Does the .NET Hashtable work in
exactly the same manner as a general Hashtable? Also, I tried using
reflector to track down the GetHashCode function to see how it
worked; I ended up at the String class (taking into consideration
the type of key I was using was a string), and the code I found for
the GetHashCode function was:

[MethodImpl(Meth odImplOptions.I nternalCall)]
public override extern int GetHashCode();

I couldn't figure out where to find anymore info...