Custom ValueType to Accept 128-bits

[Shawn B.]

Greetings,

I know I can create class that "implicit"ly accepts an 8-bit to 64-bit
value. Without using double, or single, I would like to create my own
"unsigned" 128-bit valuetype (to be used with my Binary objects I recently
created http://www.visualassembler.com/binary) but I don't know how I can do
so. How can I make it accept the number 2309823982039203948203948? I don't
want to have to pass the numbers as a string, I just want to assign it a
number and expect that it'll work.

I know I can use 64-bit because I just implicit with a ulong as a parameter
type, but I don't know how to go about creating my own 128-bit type...

Any ideas?

Thanks,
Shawn

[Robert Jeppesen]

Hi Shawn,

I don't think you'll ever get a statement like the following to compile:
Binary128 b = 2309823982039203948203948;
The compiler will have no idea what you're talking about. :)
Perhaps a Binary128 could consist of two Binary64 with some logic to hide that,
but I don't think that you'll be able to initialize it with a value bigger than 64bits, unless it's initialized with another
Binary128.

--
Robert Jeppesen
robert.jeppesen%at%durius-dot-se
"Shawn B." <le****@html.com> wrote in message news:uQ**************@tk2msftngp13.phx.gbl...

Greetings,

I know I can create class that "implicit"ly accepts an 8-bit to 64-bit
value. Without using double, or single, I would like to create my own
"unsigned" 128-bit valuetype (to be used with my Binary objects I recently
created http://www.visualassembler.com/binary) but I don't know how I can do
so. How can I make it accept the number 2309823982039203948203948? I don't
want to have to pass the numbers as a string, I just want to assign it a
number and expect that it'll work.

I know I can use 64-bit because I just implicit with a ulong as a parameter
type, but I don't know how to go about creating my own 128-bit type...

Any ideas?

Thanks,
Shawn

[Jason Smith]

You can do this with two 64 bit values, but you'll have to write a lot of
code. Mathematical operations will be quite difficult, as there is no
overflow bit (and a checked overflow condition throws an exception, which
would kill your performance on simple mathematical operations).

I haven't seen a good arbitrary-precision integer library for C#, but maybe
you should be looking for something along those lines, perhaps in the C/C++
world.

"Shawn B." <le****@html.com> wrote in message
news:uQ**************@tk2msftngp13.phx.gbl...

Greetings,

I know I can create class that "implicit"ly accepts an 8-bit to 64-bit
value. Without using double, or single, I would like to create my own
"unsigned" 128-bit valuetype (to be used with my Binary objects I recently
created http://www.visualassembler.com/binary) but I don't know how I can do so. How can I make it accept the number 2309823982039203948203948? I don't want to have to pass the numbers as a string, I just want to assign it a
number and expect that it'll work.

I know I can use 64-bit because I just implicit with a ulong as a parameter type, but I don't know how to go about creating my own 128-bit type...

Any ideas?

Thanks,
Shawn

[Shawn B.]

I was looking in the Mono source code and they have a
Corlib.Mono.Math.BigInteger object that takes 160 bits of memory. It is a
rather interesting piece of work. There is a possibility my answer lies in
in examining that source code.
Thanks,
Shawn

"Shawn B." <le****@html.com> wrote in message
news:uQ**************@tk2msftngp13.phx.gbl...

Greetings,

I know I can create class that "implicit"ly accepts an 8-bit to 64-bit
value. Without using double, or single, I would like to create my own
"unsigned" 128-bit valuetype (to be used with my Binary objects I recently
created http://www.visualassembler.com/binary) but I don't know how I can do so. How can I make it accept the number 2309823982039203948203948? I don't want to have to pass the numbers as a string, I just want to assign it a
number and expect that it'll work.

I know I can use 64-bit because I just implicit with a ulong as a parameter type, but I don't know how to go about creating my own 128-bit type...

Any ideas?

Thanks,
Shawn

[100]

Hi,
If you don't want to use strings for initializing your "big" value type the
compiler has to support such a big litterals.
I'm not sure MS c# compiler can handle those litterals. Maybe in the MONO
project they have modified their compiler.

B\rgds
100

"Shawn B." <le****@html.com> wrote in message
news:%2****************@TK2MSFTNGP09.phx.gbl...

I was looking in the Mono source code and they have a
Corlib.Mono.Math.BigInteger object that takes 160 bits of memory. It is a
rather interesting piece of work. There is a possibility my answer lies in in examining that source code.
Thanks,
Shawn

"Shawn B." <le****@html.com> wrote in message
news:uQ**************@tk2msftngp13.phx.gbl...

Greetings,

I know I can create class that "implicit"ly accepts an 8-bit to 64-bit
value. Without using double, or single, I would like to create my own
"unsigned" 128-bit valuetype (to be used with my Binary objects I recently created http://www.visualassembler.com/binary) but I don't know how I

can do

so. How can I make it accept the number 2309823982039203948203948? I

don't

want to have to pass the numbers as a string, I just want to assign it a
number and expect that it'll work.

I know I can use 64-bit because I just implicit with a ulong as a

parameter

type, but I don't know how to go about creating my own 128-bit type...

Any ideas?

Thanks,
Shawn