Some top-level goodies maybe?

[Jean-Luc-Picard-2021]

In another ticket it was stated:

Note that atoms are still not quoted properly in the output. We'll fix that soon too. https://github.com/ciao-lang/ciao/issues/53

I got more ideas, like suppressing variables that did not get instantiated. Currently I get in the Ciao WASM Playground:

?- forall(member(X,[1,4,3]), X > 0).
X=_8862 ? 
yes
?- forall(member(X,[1,4,3]), X > 1).
no

Many Prolog systems only show, like SWI-Prolog, which also uses true/false instead yes/no, but the later doesn't matter:

?- forall(member(X,[1,4,3]), X > 0).
true.

?- forall(member(X,[1,4,3]), X > 1).
false.

[mherme]

Thanks! Regarding the comments on the top level, the one in the standard distribution of Ciao Prolog also has most of those classic features of course. The issue here is that, for technical reasons, the playground currently only has a simplified top level. What we meant when we said it would be fixed soon is that we are planning on replacing that specialized top level with the full, standard one shortly, and hopefully all the classic features will come about automatically.

[jfmc]

We've just updated the playground with the changes mentioned by Manuel. The toplevel reuses more code from the standard toplevel and most of the features are magically in:

• [X] suppressing variables that did not get instantiated

Regarding these two:

• [ ] more compact layout, true/false vs yes/no
• [ ] avoid '?' when there are no more choice points (automatic determinism detection)

we'll consider them. We played with more compact layouts in the past (and had some flags to enable them) and it looks nice but we must make sure that nothing breaks.

[mherme]

Thanks to you for the very useful feedback, please keep it up!

[jfmc]

Summary of pending tasks in issue:

• implement forall/2
• detect determinism (w.r.t. choicepoint) in toplevel queries