Re: Higher order arithmetic compressoe

"SiBBi" <iBBiS@xxxxxx> wrote in
news:1151253810.698436.222350@xxxxxxxxxxxxxxxxxxxxxxxxxxxx:

I posted a few new bijective arithmetic compressors at
http://bijective.dogma.net/arbcx.zip

Sorry as someone pointed out in an email I forgot
to put the uncompressors in the zip file.
Now that there is an existing decompressor, I could test your
compressors. I created a directory "arbcx", extracted your archive
there, deleted all *.exe files (contains now sources *.cpp/*.h and
README.TXT) and created a tar archive. Here are the compression results
(file size is given in byte):

112640 arbcx.tar
68273 arbcx.tar.arbc0
39812 arbcx.tar.arbc1
26259 arbcx.tar.arbc2
24624 arbcx.tar.arbc2z
(10019 arbcx.tar.bz2)

Unfortunately, since I assume these to be lossless compressors, they
are all broken [*]:

Not likely since they have been tested. They are better than lossless
they are bijective. I wish you would at least try an more honest test
instead of thinking they are designed for unix. My code is designed on
a PC using DOS.

If you post the exact file you tested say in a zip file I will do
the test myself. Secondly of course its not as small as the bz2 the
code was written mostly so people can see what a zero first and second
order pure arithmetic 256 bijective file compressor can do. Its more for
a bench mark.


112642 arbcx.tar.unarbc0
112644 arbcx.tar.unarbc1
112645 arbcx.tar.unarbc2
112644 arbcx.tar.unarbc2z



I don't know what you did since they work on the data I test for
example each file in the Calgary Corpus. I suspect you think it is
for unix indirection it is not.

Try testing one of your exectuables like my write up says.

arbcx0 file1.xxx file2.xxx

where file to be compressed is file1.xxx
file2.xxx is the compressed output.
then run

unarbcx0 file2.xxx file3.xxx

the output is file3.xxx and is the same as file1.xxx



This is the test file it contain Hello World
note its only 11 bytes long


0000 48 45 4C 4C 4F 20 57 4F 52 4C 44 . . . . . *HELLO WORLD*
number of bytes is 11

arbc0 hw.txt hw1.txt
Next is file hw1.txt the compressed result of hw.txt
0000 A8 6F 53 3A 4B 41 B2 A5 80 . . . . . . . *.oS:KA...*
number of bytes is 9

unarbc0 hw1.txt hw2.txt
Next is file hw2.txt the uncompress of h1.txt same as hw.txt
0000 48 45 4C 4C 4F 20 57 4F 52 4C 44 . . . . . *HELLO WORLD*
number of bytes is 11

unarbc0 hw2.txt hw3.txt
since bijective next is uncompress of "HELLO WORLD"
0000 A8 B1 A6 B1 B2 B1 B2 A2 73 B2 B1 B4 B4 A1 B1 B1 *........s.......*
0010 40 . . . . . . . . . . . . . . . *@*
number of bytes is 17

arbc0 hw3.txt hw4.txt
notice that is compresses back
0000 48 45 4C 4C 4F 20 57 4F 52 4C 44 . . . . . *HELLO WORLD*
number of bytes is 11

If you can't get your executables to do the above your doing
something WRONG. Please since you brought it up at least follow
through and let us know if you figure it out.



Christian

[*] I had to delete the fprintf-line with "HERE AT LAST" in all 4
unarbc*.cpp files.





David A. Scott
--
My Crypto code
http://bijective.dogma.net/crypto/scott19u.zip
http://www.jim.com/jamesd/Kong/scott19u.zip old version
My Compression code http://bijective.dogma.net/
**TO EMAIL ME drop the roman "five" **
Disclaimer:I am in no way responsible for any of the statements
made in the above text. For all I know I might be drugged.
As a famous person once said "any cryptograhic
system is only as strong as its weakest link"

.

Relevant Pages