[m-rev.] for review: read Mercury files using read_file_as_string

[Zoltan Somogyi]

On Sat, 27 Jul 2019 17:47:23 +1000 (AEST), Julien Fischer <jfischer at opturion.com> wrote:

> I don't think the less accurate error message is significant; the
> compiler doesn't pass that test in non-C grades anyway.  (Because the
> target languages are fine with null chars in strings.)

Ok, thanks.

> > The second is that the speedup could be slightly higher if read_file_as_string
> > returned the max offset in the string it returns, which it has to know in order to
> > allocate the right amount of memory. While I could write a new variant of
> > the C version of read_file_as_string that does that, maybe called read_file_as_
> > string_and_length,
> 
> Don't use length, since that's ambiguous.

What would you suggest? I presume "size" would also be ambiguous.

> The diff looks fine otherwise.

Thanks; I will follow your suggestions.

Zoltan.