Forms like (*) or (','), is that intended in Scryer?

[haijinSk]

For example:

?- "*" = [C|_].
   C = (*).

?- "," = [C|_].
   C = (',').

I've encountered it in my personal context of text processing experiments, where, in a case of non-alphabetical characters/atoms, I naively expected forms like: '*' and ','.

Like in this example:

?- "." = [C|_].
   C = '.'.

Thank you.

[UWN]

A term like X = ','. is invalid syntax, because the comma is an operator. Try to enter it at the top level. An atom that is an operator cannot be the argument of another operator. Therefore the round brackets. (Of course, some Prologs are more permissive, but they often interpret invalid syntax differently).

Note, however, that the round brackets are only needed when an atom which is also an operator is an operand of some operator. In other situations, the round brackets are not needed like in the term f(*,[*,*|*],{*},*).

[haijinSk]

Thank you very much!

[infradig]

Interesting. SICStus & Eclipse are fine with it. SWI-Prolog accepts it but prints parens around it.

~/trealla (devel) $ rlwrap /usr/local/sicstus4.8.0/bin/sicstus
SICStus 4.8.0 (x86_64-linux-glibc2.28): Sun Dec 4 13:17:41 UTC 2022
Licensed to Andrew Davison
| ?- X = ','.
X = ',' ? ;
no

EDIT: ok, I see that was covered in a comment. Ignore this.

[UWN]

As to the output of SICStus note that the top level loop is explicitly excluded from the scope of 13211-1. See 1 Scope, note f. Otherwise, you do get the right output:

| ?- write_term(X = (','),[quoted(true),variable_names(['X'=X])]).
X=(',')
yes