From: Kim Pedersen <kimersen@...>
Date: 2006-08-02T06:48:01+09:00
Subject: Re: Binary numbers

Dale Martenson wrote:
> On Wednesday, August 02, 2006, at 5:46 AM, Kim Pedersen wrote:
>   
>> Hi,
>>
>> I have some data with binary numbers.
>> e.g. a 3 byte string could be 0x01216f which is 74095 decimal.
>> To do this conversion in ruby I've come up with:
>>
>> def binum(bs)
>>    n = e = 0
>>    bs.reverse.each_byte do |c|
>>        n += c * 256 ** e
>>        e += 1
>>    end
>>    n
>> end
>>
>> bs = ''
>> bs << 0x01 << 0x21 << 0x6f
>>
>> puts binum(bs)  => 74095
>>
>> Now some data are signed binary numbers stored as two's complement.
>> e.g. 0xfede91 which is -74095 decimal.
>>
>> How can I do this conversion in ruby?
>>     
>
> def binum(bs)
>   result = 0
>   bs.unpack("cCC").each {|b| result = (result << 8) | b }
>   result
> end
>
> bs = ''
> bs << 0x01 << 0x21 << 0x6f
>
> puts binum(bs)
>
> bs = ''
> bs << 0xfe << 0xde << 0x91
>
> puts binum(bs)
>
> The above code extracts the first byte as a signed 8-bit integer and the  
> rest of the bytes as unsigned 8-bit integers. Each iteration through the  
> loop shift the result to construct the proper value.
>
> --Dale Martenson
>   
Great!
Thanks
Kim

BTW, the length of the string may vary from 1 to 8.  I'll have to adjust 
for that.