baseConvert() & 32-bit ops (was Best array population & access optimization)

[Dar Scott]

On Saturday, June 21, 2003, at 01:12 PM, Mark Brownell wrote:
> On Saturday, June 21, 2003, at 09:58  AM, Dar Scott wrote:
>> Wanna see my test script and results?
>
> Yes please. I want to run the different  OS versions of the hexed  
> boxes through it. It will be nice to see what happens with the 
> constants that seem to work if the unsigned numbers are the same.


on mouseUp
   put empty into field "Report"
   --
   put ( (2^32)-1 )&"" into allOnes
   put (allOnes bitAnd allOnes) & LF & LF after field "Report"
   --
   put allOnes                    & LF    after field "Report"
   put baseConvert(allOnes,10, 2) & "   " after field "Report"
   put baseConvert(allOnes,10,10) & "   " after field "Report"
   put baseConvert(allOnes,10,16) & "   " after field "Report"
   put baseConvert(allOnes,10,30) & LF    after field "Report"
   put baseConvert(allOnes,10,10) into ones10
   put (ones10 bitAnd ones10)     & LF    after field "Report"
   put                              LF    after field "Report"
--
   put "FFFFFFFF" into allOnes
   put allOnes                    & LF    after field "Report"
   put baseConvert(allOnes,16, 2) & "   " after field "Report"
   put baseConvert(allOnes,16,10) & "   " after field "Report"
   put baseConvert(allOnes,16,16) & "   " after field "Report"
   put baseConvert(allOnes,16,30) & LF    after field "Report"
   put baseConvert(allOnes,16,10) into ones10
   put baseConvert(ones10 ,10,16) & "   " after field "Report"
   put baseConvert(-ones10,10,16) & "   " after field "Report"
   put (ones10 bitAnd ones10)     & LF    after field "Report"
   put                              LF    after field "Report"

--
   put "80000000" into oneOnes
   put oneOnes                    & LF    after field "Report"
   put baseConvert(oneOnes,16, 2) & "   " after field "Report"
   put baseConvert(oneOnes,16,10) & "   " after field "Report"
   put baseConvert(oneOnes,16,16) & "   " after field "Report"
   put baseConvert(oneOnes,16,30) & LF    after field "Report"
   put baseConvert(oneOnes,16,10) into one10
   put baseConvert(one10  ,10,16) & "   " after field "Report"
   put baseConvert(-one10 ,10,16) & "   " after field "Report"
   put (one10 bitAnd one10)       & LF    after field "Report"
   put                              LF    after field "Report"
end mouseUp

And the not very pretty result is this:


4294967295

4294967295
11111111111111111111111111111111   -4294967295   FFFFFFFF   5QMCPQF
0

FFFFFFFF
11111111111111111111111111111111   -4294967295   FFFFFFFF   5QMCPQF
-FFFFFFFF   FFFFFFFF   0

80000000
10000000000000000000000000000000   -2147483648   80000000   2SB6CS8
-80000000   80000000   0


If I make a test with better output, I'll send you that.

The main things to notice is that all ones comes out as -4294967295 in 
decimal and not 4294967295 or even -1.  Also, note that bitAnd of that 
doesn't get all ones either.

>
> also, not being familiar with binaryEncode how would you write the 
> following using binaryEncode?
>
> put charToNum("F") + (charToNum("o") * 256 ) + ( charToNum("r") * 
> 65536 ) + ( charToNum("m") * 16777216 ) into xL

   local halfBlock  -- required
   put binaryDecode("N","Form",halfBlock) into numConverted

In converting strings to blocks use 'char i to (i+3) of plainText' 
instead of "Form" and pay attention to the end.  The variable 
numConverted with be 0 for short data.

You use binaryEncode for the other direction and, perhaps, in the first 
part of the Blowfish F function.  (The later might take two steps.)

Dar Scott