[m-dev.] Re: diff: higher_order.m termination fix

[Simon TAYLOR]

Fergus wrote:
> >	Don't fill in the clauses_info for specialised versions, since
> >	it just clutters the HLDS dumps.
> 
> I don't think this change is a good idea.
> It breaks an invariant of the data structure (namely the invariant that
> the clauses_info contents is always valid).
> Although there may not be any code that depends on that invariant,
> it is difficult to be sure, and even if this is true, it is
> difficult to be sure that this will always be the case.

If the clauses are expected to be valid after higher_order.m, the 
original code is buggy, since it just copies the clauses for the 
old predicate when creating the pred_info for a specialised version.

We should probably document that the clauses_info is not valid after
mode analysis, since the clauses and the procedures may be different.

Simon.