1. Huffman endoder / decoder
- Posted by DB James <larch at adelphia.net>
Oct 22, 2005
-
Last edited Oct 23, 2005
Hi all,
I have submitted a program called Huffit.ex that is an elaboration on the two
programs by Junko Miura called huff.ex and dehuff.ex. It should be available to
download soon.
My purpose in writing it was to put her two programs into one, and with with a
user interface which allowed file navigation and selection. The effect is not
only to have compressed files, but to make files unreadable until they are
decoded for whatever privacy concerns the user may have.
Whether life seems without savor if you don't do Huffman compression on a daily
basis or not, it is interesting to see how fast Euphoria is in doing the encoding
and decoding: I tested it on 104 files with a total of 10,678,803 bytes and the
interpreted version did the encoding / decoding in 15 / 11 seconds. The C
compiled (free) version (which I included in the zip file) did the job in 12 / 8
seconds. And, though all data is processed in memory, it handled a 38 MB file on
my machine which has 512 MB of memory.
--Quark
2. Re: Huffman endoder / decoder
- Posted by Greg Haberek <ghaberek at gmail.com>
Oct 22, 2005
-
Last edited Oct 23, 2005
> I have submitted a program called Huffit.ex that is an elaboration on the=
two programs by Junko Miura called huff.ex and dehuff.ex. It should be av=
ailable to download soon.
Ever looked into Arithmetic Coding? I've seen it turn out some
impressive compression ratios, better than Huffman. Just a thought.
~Greg
3. Re: Huffman endoder / decoder
Greg Haberek wrote:
>
> > I have submitted a program called Huffit.ex that is an elaboration on the=
> two programs by Junko Miura called huff.ex and dehuff.ex. It should be av=
> ailable to download soon.
>
> Ever looked into Arithmetic Coding? I've seen it turn out some
> impressive compression ratios, better than Huffman. Just a thought.
>
> ~Greg
Hi Greg,
No, I haven't checked out the Arithmetic Coding, but I downloaded some info on
it and will look into it. However, my effort with Huffman was to make use of
Junko's good work and possibly make it more accessible, as well as for practice
in doing that sort of thing: modifying a program for more general use as,
understandably, many coders are focused on getting something to work, but haven't
the time or maybe patience to finish the job as far as the user is concerned.
The result is that we have good demonstrations of something useful or
interesting, but not quite practical for casual use...
--Quark
