In the past I have worked with D, C#, Java, Python and C and some other less popular languages. Most recently working with C# I suddenly realize the convenience and flexibility I had in D. One case in particular is with bit-fields in D they are a pleasure to use and implement, but in C# they are a mystery, sure they have Enums that are flexible for some cases, I suppose I just wanted to say thanks to the D community for their effort.

On Friday, 11 November 2016 at 15:36:18 UTC, Taylor Hillegeist wrote:
> In the past I have worked with D, C#, Java, Python and C and some other less popular languages. Most recently working with C# I suddenly realize the convenience and flexibility I had in D. One case in particular is with bit-fields in D they are a pleasure to use and implement, but in C# they are a mystery, sure they have Enums that are flexible for some cases, I suppose I just wanted to say thanks to the D community for their effort.

D doesn't have bitfields though? Phobos just has a template that emulates them by mixing in functions that do the bit shifting. Because it done through mixins it means you can't attach attributes to the fields. Which I am in need of. Haven't found a good workaround though.

On Friday, November 11, 2016 23:32:11 Jerry via Digitalmars-d wrote:
> D doesn't have bitfields though? Phobos just has a template that emulates them by mixing in functions that do the bit shifting. Because it done through mixins it means you can't attach attributes to the fields. Which I am in need of. Haven't found a good workaround though.

What attributes do you need that aren't there already? UDAs? Looking at the code, the full set of appropriate built-in attributes are on them already.

- Jonathan M Davis

On Saturday, 12 November 2016 at 01:02:05 UTC, Jonathan M Davis wrote:
> On Friday, November 11, 2016 23:32:11 Jerry via Digitalmars-d wrote:
>> D doesn't have bitfields though? Phobos just has a template that emulates them by mixing in functions that do the bit shifting. Because it done through mixins it means you can't attach attributes to the fields. Which I am in need of. Haven't found a good workaround though.
>
> What attributes do you need that aren't there already? UDAs? Looking at the code, the full set of appropriate built-in attributes are on them already.
>
> - Jonathan M Davis

Yah I need to associate some extra information with the bitfields using UDAs.

On Saturday, November 12, 2016 03:30:58 Jerry via Digitalmars-d wrote:
> On Saturday, 12 November 2016 at 01:02:05 UTC, Jonathan M Davis
>
> wrote:
> > On Friday, November 11, 2016 23:32:11 Jerry via Digitalmars-d
> >
> > wrote:
> >> D doesn't have bitfields though? Phobos just has a template that emulates them by mixing in functions that do the bit shifting. Because it done through mixins it means you can't attach attributes to the fields. Which I am in need of. Haven't found a good workaround though.
> >
> > What attributes do you need that aren't there already? UDAs? Looking at the code, the full set of appropriate built-in attributes are on them already.
> >
> > - Jonathan M Davis
>
> Yah I need to associate some extra information with the bitfields using UDAs.

It would definitely be possible to make it so that std.bitmanip.bitfields could handle UDAs, but I don't expect that it would be particularly fun to implement. e.g. the example

struct A
{
    int a;
    mixin(bitfields!(
        uint, "x",    2,
        int,  "y",    3,
        uint, "z",    2,
        bool, "flag", 1));
}
A obj;
obj.x = 2;
obj.z = obj.x;

could become something like

struct A
{
    int a;
    mixin(bitfields!(
        uint, "x",    2, "@Foo @Bar(12)"
        int,  "y",    3,
        uint, "z",    2, "@Foo"
        bool, "flag", 1));
}
A obj;
obj.x = 2;
obj.z = obj.x;

but that would require making it so that bitfields looks at the types of what it's passed to see if it has the extra string (similar to what std.getopt.getopt does), and if it does, then it mixes in that string as attributes on that field. It's a bit a of a pain to do, but it could be done. What would make it worse though is if you want to try and make it so that the attributes can be applied to just the setters or just the getters. Then you'd probably end up with stuff like

struct A
{
    int a;
    mixin(bitfields!(
        uint, "x",    2, "@Foo @Bar(12)", "@Foo"
        int,  "y",    3,
        uint, "z",    2, "@Foo", "",
        bool, "flag", 1));
}
A obj;
obj.x = 2;
obj.z = obj.x;

which is getting a bit ugly IMHO. But it would be possible.

In any case, that's the best approach that I can think of at the moment.

- Jonathan M Davis

On Saturday, 12 November 2016 at 03:59:08 UTC, Jonathan M Davis wrote:
> It would definitely be possible to make it so that std.bitmanip.bitfields could handle UDAs, but I don't expect that it would be particularly fun to implement. e.g. the example
>
> <Snip>
>
> which is getting a bit ugly IMHO. But it would be possible.
>
> In any case, that's the best approach that I can think of at the moment.

 I can't help but remember that in my variant of Bitfields I had it allow default values. I could probably extract that part out and re-inject it into the current bitfields version, although I remember quite the fun time debugging said code, namely extracting the text generated...