le Redoutable
Ich bin El Glouglou :)
this cruncher is based on offsets from bytes;
also important is the homogeneity of the file to get crunchet ===> first pack it with Winzip, then apply the I.P. ;
the first thing to do is establish what bytes values are the most common ( like a value of, say 232 or 120, 0, 255 etc )
then you put data concerning the offset from location of first occurence from the given byte value, then from offset in between first occ to the second, then to the third etc up to the last value from the input file.
while you put offset values, you place flags from where the offsets appear, like in first occ = 20 ===> flag put at 20; then next occ = 62 , so an offset of 62-20=42, but you place a flag at 62;
now, because not every offset will stand to a value of 1 ---> 255 we must use the value of 255 where you will note a value beyond 254 ( say you have an offset of 888 )
ok now you determine offsets from the second highest byte value;
say the first offset from first byte value = 20,
and first offset from second byte value = 33, you will decrement the value 33 by one point ( and will print 32 to output file )
now the first value from third byte value = 12,
you will gently print a 12 to output
second value from third byte value = 35,
, so offset should be 35-12=23 , minus one point for the flag at 20 and minus the flag at 33 to a total of 21
you should see that, near the end of the input file you will get bargains on the offsets, and there should be more of small values than of high ( i.e. more 1,5,25 than 200, 128 etc )
so, we happen to a heterogeneity file, which Winzip is fond of
but all this is pure theory
( I must be overconfident )
what do you think of that ??
also important is the homogeneity of the file to get crunchet ===> first pack it with Winzip, then apply the I.P. ;
the first thing to do is establish what bytes values are the most common ( like a value of, say 232 or 120, 0, 255 etc )
then you put data concerning the offset from location of first occurence from the given byte value, then from offset in between first occ to the second, then to the third etc up to the last value from the input file.
while you put offset values, you place flags from where the offsets appear, like in first occ = 20 ===> flag put at 20; then next occ = 62 , so an offset of 62-20=42, but you place a flag at 62;
now, because not every offset will stand to a value of 1 ---> 255 we must use the value of 255 where you will note a value beyond 254 ( say you have an offset of 888 )
ok now you determine offsets from the second highest byte value;
say the first offset from first byte value = 20,
and first offset from second byte value = 33, you will decrement the value 33 by one point ( and will print 32 to output file )
now the first value from third byte value = 12,
you will gently print a 12 to output
second value from third byte value = 35,
, so offset should be 35-12=23 , minus one point for the flag at 20 and minus the flag at 33 to a total of 21
you should see that, near the end of the input file you will get bargains on the offsets, and there should be more of small values than of high ( i.e. more 1,5,25 than 200, 128 etc )
so, we happen to a heterogeneity file, which Winzip is fond of
but all this is pure theory
( I must be overconfident )
what do you think of that ??