llvm dev - Nov 2015 - Why StringRef is not a ArrayRef<char>

Hi,

Why is StringRef not an ArrayRef<char> extension? That might simplify some
interfaces.

Cheers,
Paweł
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20151105/ef56dcdc/attachment.html>

I think it's the same difference we have between int * and
ArrayRef<int>, for instance. If you have a raw pointer, you are
expected to keep track of the size out-of-band. In aggregate types,
size is part of the type, making it very difficult to confuse a [1 x
i8] with a [3 x i8]. Many people probably use the raw pointer as a way
to hold a reference to dynamically allocated memory, and as a
type-escape hatch.

On Thu, Nov 5, 2015 at 11:58 AM, Paweł Bylica <llvm-dev at lists.llvm.org>
wrote:
> Hi,
>
> Why is StringRef not an ArrayRef<char> extension? That might simplify
some
> interfaces.
>
> Cheers,
> Paweł
>
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>

On Fri, Nov 06, 2015 at 09:07:43AM -0500, Ramkumar Ramachandra via llvm-dev
wrote:
> I think it's the same difference we have between int * and
> ArrayRef<int>, for instance. If you have a raw pointer, you are
> expected to keep track of the size out-of-band. In aggregate types,
> size is part of the type, making it very difficult to confuse a [1 x
> i8] with a [3 x i8]. Many people probably use the raw pointer as a way
> to hold a reference to dynamically allocated memory, and as a
> type-escape hatch.
StringRef is explicitly sized. I'd be more interested in which
interfaces are actually simplified by implicit or explicit conversions
between StringRef and ArrayRef<char>.

Joerg