| biject 2007-12-19, 9:57 pm |
| On Dec 19, 11:46 am, Klaus Stengel <Klaus.Sten...@asamnet.de> wrote:
> Klaus Stengel wrote:
>
>
>
>
>
>
> There is a mistake in that last sentence. If we always reset the marker
> to the start of the new cycle, we'll miss some subdivisions. For
> example, "FOOFOOFO1" would be partitioned into "FOO|FOOFO|1" instead of
> the correct "FOO|FOO|FO|1".
>
> If the marker is already in the new cycle, it has to stay in that
> position. Only if it stayed in the old cycle we need to advance it
> to the start of the new one.
>
> This has no impact on the fact that this algorithm is O(n), as the
> maximum number of partitioning steps is still limited to the number of
> input characters.
>
> Bye,
> Klaus.- Hide quoted text -
>
> - Show quoted text -
Let me gave an other example by what I meant the breaking into cycles
then merging
|b,,iththj...v|a....eththd...x|
above is eample of the two splits where |a...ethtjd...y| is
lessthen |b...iththj...z|
when sorting the partitions you get two lists
a....eththd,,x label for other 1 the partiotion its self is first
for the cycle
d...xa...etheh 2
eththd...xa... 3
hd,,,xa,,,.etht 4
hthd...xa....et 5
thd....xa...eth 6
ththd...xa...e.. 7
xa....eththd... 8
of course many other rows missing just showing relative order of these
next partition
b.. iththj,,,,v A again many rows missing just to show
trend
iththj.....zb... B
j......vb...ithth C
hj.....vb.....itht D
hthj.....vb.....it E
thj......vb....ith F
ththj.....vb itht G
vb...,,.iththj.... H
to do BWTS you know megre these lists
1 of course the master cycle is top dog
A though some other missed in master cycle could be above
this
2
3
B
C
4 goes to t the mergincan increase runs
D goes to t
5 goes to t
E goes to t
6
F
7
G
H
8
Note I mention to compare funcntions the one for serpate cycles
( ) ( )
and the one for embedded cycles ( ( ))
well many people like having a $ super heavy symbol ( $)
versus ( $)
I think that compare could would be the same as ( ))
( $)) for compare when
making partitions. There are many ways to slice and dice it I lke
the one tha looks
more like the original BWT method but thats me tomorrow I may change
my mind in fact
once you see one more jump out at you
Take Care
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"
|