[m-rev.] for post-commit review: give "current dir" extensions their own categories

[Julien Fischer]

On Tue, 6 Jun 2023, Zoltan Somogyi wrote:

>
> On 2023-06-06 03:39 +02:00 CEST, "Julien Fischer" <jfischer at opturion.com> wrote:
>>> +    ;       newext_mh(string)
>>> +            % Machine-dependent header files for generated C code.
>>> +            % The extension string is ".mh".
>>
>> What do .mh files contain that is machine dependent (other than the
>> comment at their head containing FullArchName)?
>
> I don't know. I do remember that .mih stands for "machine independent header",
> and .mh lacks the "i". To the best of my knowledge, Fergus never documented
> his rationale of what goes into each kind of header file.

I always though that .mh == Mercury header
and .mih == Mercury internal header.

>
> I have replaced that sentence with
>
> Information about Mercury code exported to outside C code.
>
> based on what export.m puts into .mh files.

It was looking at what export.m was doing that caused me to ask that
question in the first place.  I cannot see anything that it is
outputting that is machine dependent.

Julien.