This code works for *ALL* kinds of data 
types, it supports all the ASCII 
characters. But the best ratio is given 
on textfiles, where some characters 
appear more often than others and some 
characters does not appear at all. 
Binary files often contains all ASCII 
values and therefore the Huffman 
Algorithm can not compress the data 
much (or at all).

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Hi Frederik,
I have compressed a file 
that was
>6 MB.
The Result :
Source 
size = 6117120 bytes
Dest   size = 
1534173 bytes
Ratio       = 25%
Time 
spend  = 1,5632 sec.
>
I mean this is 
a exelent work from you
but I have 
rated so.
I´m looking your next work 
impatient :-)
>
by michael

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Fredrik: You said you wanted to post a 
Lempel-Ziv Compression Program.
Why 
not use both Huffman and Lempel-Ziv? 
This is actually called "Deflation" 
(What used in Winzip).
Here are some 
sites for this compression 
method:
http://www.cdrom.com/pub/info
zip/zlib/
http://www.cdrom.com/pub/info
zip/zlib/zlib_docs.html
http://www.faqs
.org/faqs/compression-faq/
http://www.c
drom.com/pub/infozip/zlib/rfc-deflate.ht
ml
http://www.internz.com/compression-p
ointers.html
Hope that 
helps,
Yaron 
B.
Budowski@bezeqint.net

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

I might look into that Deflate routine 
(but i am using the ZLib DLL already 
and this would be probably 7-8x slower 
in VB so I don't know..)
For you who 
requested a self-extracting EXE you can 
use the code 11335 and replace the .bas 
file with my .cls file to improve 
performance.
"You can put more 
afford (effort?) to enhance it, 
 it 
may bring a great challenging for 
you..."
Is it *possible* to improve 
upon this code? I think not (at least 
not in VB). I believe this is actually 
the fastest possible solution. Then you 
can design 1000 different 
implementations but that's not what a 
.cls file is designed for. ;)

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

I uploaded a new version, where the 
only bug I could find is fixed. I also 
added some error trapping so now you 
can not try to compress a file that 
does not exist (if you tried it before 
you got a Subscript-out-of-bounds 
error). New is also support for a 
Progress Event! ;)

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Will this zip be compatible for 
unzipping with PKZIP, WINZIP, etc?

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

FYI - Fredrik
I noticed that your 
latest screen shot shows a longer 
compression duration of the same file 
over the older screen shot.
The 
previous screen shot showed a 
compressed time of 0,988125 s and of 
course the current screen shot shows a 
time of 1,318125 s.
This seems to be 
a roughly 25% decrease in performance.

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Some of the decrease in performance can 
probably be found in the new Event that 
is raised, if you comment everything 
that has with the new event to do you 
might get some better performance. I 
might have run that old run before I 
added the simple crc too. I also had my 
CPU at a higher Mhz but recently I had 
to downclock it 'cause I forgot to 
plugin the fan a couple of weeks ago. 
Doh. ;)
Unfortunately this 
compression is not compatible with 
Winzip (as that uses a LZ77 routine 
just like ZLib on top of the Huffman? 
Not sure).

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Really a great code. Im waiting for 
your other works.

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

This is nice code, but can you make it 
search to more than two leafs?  Maybe 
you could add a parameter 

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Could you give me your email address, i 
have something want to discuss with you 
and want to ask for help. it was really 
urgent. 

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Bravo! this is truely ingenius work 
you've posted here, I have a question, 
how d you think this compression does 
on BMP files compared to a RLE 
compcression (keep in mind that most 
BMPs have at least a fair bit of 
repition, depending on the graphic.)???

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Is there any way of password protecting 
these files as in winzip?

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Tim_Miron:
I believe you will get 
better performance from the RLE 
compression when dealing with large 
bitmaps, and specially those with high 
color-depth (as you said lot of 
repeting data). I saw a RLE post on PSC 
a week or so ago you could compare 
with, it's 
at..
http://www.planet-source-code.co
m/xq/ASP/txtCodeId.12790/lngWId.1/qx/vb/
scripts/ShowCode.htm

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Could you kindly explain what are the 
progress values,and how come the values 
are 7,5 and 88 for encoded function;and 
89 and 11 for decoded function.Could 
you kindly explain how is your 
algorithm optimised OR different from 
the conventional algorithm.
   I am a final year engineering 
student and your program has come in 
real handy.If it is possible for you to 
find time please mail me at 
nimitsarup@Hotmail.com.
Thanks a million.
             

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Hmm. found a bug in the source code...

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Weird... posting screwed up... Let me 
try again.
In the sub 
EncodeByte(),
If (ByteLen = 0) 
Then
ReDim Preserve ByteArray(0 To 
ByteLen + 3)
If (ByteLen > 0) Then 
...
If ByteLen = 0, why is there a 
need for
the 2nd "If"? Since it won't 
be used at
all.
After which, there 
is 
"If ((lLength = 0) Or (lLength > 
ByteLen)) Then"
And that is before 
adding the CRC and other data,
why not 
put that near the end? And why
just 
"lLength > ByteLen" since even 
if
lLength = ByteLen, the "compressed" 
length
will be BIGGER than the 
original file.
Regards

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

First off, I'd like to praise you for 
the wonderful piece of code!
I did 
some comparisons of the compression 
ratios between RLE and Huffman. (Using 
my RLE example: 
http://www.planet-source-code.com/xq/ASP
/txtCodeId.8418/lngWId.1/qx/vb/scripts/S
howCode.htm )
On the images I've 
used, Huffman produced better 
compression most of the time. However, 
I also found that compressing a file 
with RLE and then with Huffman yielded 
even better results! Results comparable 
with JPEG for large enough files! (Well 
this is quite reasonable since the JPEG 
algorithm uses RLE and/or Huffman after 
doing the quantizations)
Anyways 
keep up the great work.

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

What can i say other than this >> This 
piece of code is great!

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

how do i make the program recognize the 
original extention of the file after 
decompression process(.jpg,.exe 
....etc) ???? after i decompress the 
file i can't know what it was and i 
dont wana write saol.txt.new and all 
these things...p.s. its a gr8 code

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

if anyone knows how plzzzzzzz contact 
me ..i need to know and by the way..do 
u have any documentaions explaining 
that algorithm and y u choosed these 
valuse i'll apreciate if u send it to 
me sir..thanks alot

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Let's see if we can get PSC to keep my 
post as one piece this time... when 
compressing an 8,192 byte of all char 
code 255, an Illegal Page Fault is 
brought up.  I checked the code, and a 
CopyMemory call is causing this.

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Uploading a modified version of this 
code that will function with Microsoft 
Access as a Module.  - I'm working on 
the code to further enhance it allowing 
multiple files to be compressed into a 
single file which is where compression 
routines really shine I think.  Search 
my submissions to view the MS Access 
code.

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

This code is so sweet it'll make you 
cry. 5 Globes

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

I have only one question:
Is it 
possible to change this class file to 
unpack a file to memory?
Imagine 
this:
I have about 1000 texts packed 
with your program (1 text per file) and 
they are recorded in a cd-rom, well i 
NEED to unpack them directly to a 
textbox/listbox/richtextbox without 
creating any temporary file first, 
which means that the unpack routine, 
should NOT write any information unless 
to the text/list/rich box.
this may 
seem a little to much to ask, but since 
your code is so GREAT i was expecting 
that maybe you could be so kind to 
arrange something like this...
The 
only this i could get was copying the 
ByteArray() with Clippboard.Copy to 
memory :( and not the actual unpack 
data in ascii format but binary 
:(
is there any function/class to 
convert binary data to string?
hope 
that you ALL can help,
Best 
Regards
RootShell
(RootShell@netcabo
.pt)

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

This code has gone a long way. I was 
toying with it and yes, you can still 
improve the speed of 
compression/decompression by replacing 
he CharCount array with a Dictionary, 
in this case you don't have to parse 
the entire ascii character set for 
non-zero counts. There are several 
FOR-NEXT loops that can still be 
optimize, including the progress bar. 
For both text and binary file test, the 
time it takes to compress and 
decompress has been reduced an average 
of 46.7%. Another issue with this code 
is, it reads all of the file data in 
memory, what if we are dealing with a 
10MB file? An alternative approach is 
to process the file in a 2Kb file read 
increment. The code runs faster too and 
there's no degradation of performance. 
At some point, speed is measured in 
ticks and not in seconds. 

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Very useful. Thanks for sharing.

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason:
 

Just tried your compression code on 
some access databases. Which are being 
compressed from 70k to about 27k, which 
is great. But after uncompressing, 
Access and my VB application throws an 
error, "Unrecognised database format". 
Have you had any previous reports of 
this problem. Cheers Nick

Keep the Planet clean! If this comment was disrespectful, please report it:
Reason: