[m-rev.] diff: simplify code constructing pred names

[Julien Fischer]

On Mon, 7 Feb 2022, Zoltan Somogyi wrote:

> The diff is not really worth reviewing, but it does raise
> a question. At the moment, there are several modules
> that construct new pred names (this diff affects two of them).
> There is no overall pattern, and what partial pattern does exist
> is, in my opinion, flawed, for reasons documented in the XXXs
> added by this diff.

I suspect the situation may be worse than there just be a lack of an
overall pattern; it's likely some things currently only work by
accident.

> Would anyone object if I moved all the code that constructs
> new pred names to a new module, pred_name.m?

No objections from me, I think that would be a worthwhile (and very
overdue) change.

Julien.