Compress Lexico

Thread view:

kubek82 - 25 Aug 2006 09:26 GMT

Hi.
I have a list of words stored in a file.Words (Strings) are separated
with new line character.
I need to keep content of the file in memory during run time, but
sometimes, for different languages, size of the file increases to quite
big size.

I would like to compress the content of the file and load it in to
memory.
Is it worth to do that using Binary Tree? Or simply Huffman Coding
would be more optimal - in case of compress factor=2, I save 2 times
more memory with small programming afford.

Is the Binary Tree good solution for Scrabble? Let's say I want to go
through the Lexico and check whether given word can be formed from my
letters.

I don't expect anyone doing my homework for me, I just hope You can
give me some ideas, key words I should look for, maybe links to code
snippets etc...

With regards
Chris

Reply to this Message

Ben Kraufmann - 25 Aug 2006 12:33 GMT

> Hi.
> I have a list of words stored in a file.Words (Strings) are separated
[quoted text clipped - 19 lines]
> With regards
> Chris

Hi!

You can compress and uncompress the file itself using the package
java.util.zip.
Hmmm, scrabble... Have a look at n-ary search trees. The idea is that a
path in the search tree represents a word in your dictionary:

-|-d-o-g.
|-c-a-t.-s.
|-n-d-l-e.

Imagine this was a tree ;)
Each '.' means that you have found a word (end of a path).
I think this is called Directed Acyclic Word Graph.

Ben

Reply to this Message

Compress Lexico

Free Magazines