Re: [hash2][review] Early review (due to holidays)

From: Claudio DeSouza (cdesouza_at_[hidden])
Date: 2024-12-04 14:41:38

I tend to disagree. In my opinion even this:

   hash.update( as_bytes( span{st} ) );

Is better than this:

   hash.update( v.data(), v.size() );

Something like as_byte_span is just an extra nicety, but I think it is easy
to underestimate how you can avoid mistakes with the first example as
opposed to the second.

On Wed, Dec 4, 2024 at 2:38 PM Peter Dimov via Boost <boost_at_[hidden]>
wrote:

> Claudio DeSouza wrote:
> > > So I switched `update` back to void const*. The "type safety" gains
> > > aren't worth the substantial decrease in usability.
> > >
> >
> > In my experience, the usability/readability gains are on the side of
> using span.
> > This may involve providing extensions, but at the end of the day you get
> more
> > C++-friendly constructs.
>
> Well... no, they aren't.
>
> Suppose you have the data in `std::vector<unsigned char> v`. With the
> current
> `update` that's
>
>     hash.update( v.data(), v.size() );
>
> And if you have the data in `std::string st`, that's
>
>     hash.update( st.data(), st.size() );
>
> Now suppose `update` takes `span<unsigned char const>`. For `v`, this is
>
>     hash.update( v );
>
> which is nice; but for `st`, that would be (assuming C++17)
>
>     hash.update( as_bytes( span{st} ) );
>
> which is not so nice.
>
> In Chromium you have as_byte_span, which does as_bytes(span(x)), so
> it'd be better.
>
>
>
> _______________________________________________
> Unsubscribe & other changes:
> http://lists.boost.org/mailman/listinfo.cgi/boost
>