[m-rev.] for review: various mlds_to_il fixes

[Peter Ross]

Fergus wrote:
> On 13-Jul-2001, Peter Ross <peter.ross at miscrit.be> wrote:
> > compiler/mlds_to_il.m:
> >     When generating the initializer for arrays use the type information
> >     to generate arrays of the correct type.  We still treat structs like
> >     they are an array of System.Object.
>
> That is already done by the change that I posted earlier this week
> and just committed:
>
<http://www.cs.mu.oz.au/research/mercury/mailing-lists/mercury-reviews/mercu
ry-reviews.0107/0106.html>.
>
Yes I just saw your recent mail on that, oh well it improved my
understanding of the code.

> >     Set fields with local access to .NET private.
>
> I don't see why that is needed.
> It's an invariant of the MLDS that fields shouldn't have "local" access.
> Only entities declared in "block" statements should have "local" access.
>
The ones that have the local access are the variables next_slots_table and
string_table which are generated in ml_string_switch (or at least they had
local access when I started work on this change).

> For fields of environment structs, which were originally local vars,
> ml_elim_nested changes their access from "local" to "public"
> (see convert_local_to_field).  Perhaps that should be changed to
> "default".
>
> >     Check to see if a field is a field of the current class before
> >     assuming that it is handwritten.
>
> That's OK, but you will get merge conflicts with my changes;
> could you please post a diff after resolving those?
>
I will do.

--------------------------------------------------------------------------
mercury-reviews mailing list
post:  mercury-reviews at cs.mu.oz.au
administrative address: owner-mercury-reviews at cs.mu.oz.au
unsubscribe: Address: mercury-reviews-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-reviews-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------