• Dan <murpsoft hotmail.com> Nov 23 2007
• "Stewart Gordon" <smjg_1998 yahoo.com> Nov 24 2007
• BCS <ao pathlink.com> Nov 24 2007

Dan <murpsoft hotmail.com> writes:

I thought I recalled D having a "bit[]" type with the restriction that you
could only use bit[n*8]?

I'm asking because I've been developing an interest in DNA.  I had thought it
would be an amazing project to produce an open source DNA format.  Raw DNA uses
two bits, but you could probably need another 2 per base pair for expression. 
From there, you have raw DNA.  You want to build slices, as slices get copied
off to make siRNA and mRNA and such.  So mapping out which parts of the DNA are
used to make copies of what tends to be useful.

D just seems to fit together quite well - but I want an easy 4-bit type to work
with.

Who likes the idea?

"Stewart Gordon" <smjg_1998 yahoo.com> writes:

"Dan" <murpsoft hotmail.com> wrote in message 
news:fi7r8q$l05$1 digitalmars.com...
 I thought I recalled D having a "bit[]" type with the restriction that you 
 could only use bit[n*8]?


There was once a bit type, but it never had such a restriction.  Walter 
removed it, I think because it was found too complicated and bug-prone to 
implement.  Or something like that.

Meanwhile, Phobos has std.bitarray.  I also have a bit array implementation 
capable of arbitrary slicing:
http://pr.stewartsplace.org.uk/d/sutil/

<snip>
 D just seems to fit together quite well - but I want an easy 4-bit type to 
 work with.

 Who likes the idea?


4-bit type?  I guess some of the bit array code could be modified to give 
nibble arrays, if that's what you mean....

Stewart.

-- 
My e-mail address is valid but not my primary mailbox.  Please keep replies 
on the 'group where everybody may benefit. 

BCS <ao pathlink.com> writes:

Reply to Stewart,

 "Dan" <murpsoft hotmail.com> wrote in message
 news:fi7r8q$l05$1 digitalmars.com...
 
 I thought I recalled D having a "bit[]" type with the restriction
 that you could only use bit[n*8]?
 


 Walter removed it, I think because it was found too complicated and
 bug-prone to implement.  Or something like that.
 


I think it was dropped because there was no way to get uniform slice semantics

bit[100] bits;

auto bits[20..25]; 

For that to even be doable it needs a copy that junks reference semantics, 
or needs to  play games with pointers in ways that restrict sliced bit arrays 
to only 1/8th of the address space (low 3 bits of pointer are used for position 
in byte and byte location is got by bit shifting, but now the high order 
bits need to be implicit :P )