A few unquoted atoms for your collection

[triska]

One of the issues I sometimes see in new Prolog code is that more quotes than necessary are used.

So, here are a few atoms for your collection:

?- member(Atom, [?,???,++...--,##/>,*./=?]).
   Atom = ? 
;  Atom = ??? 
;  Atom = ++...-- 
;  Atom = ##/> 
;  Atom = *./=? .

None of them require quoting. An example of an atom that does need quoting is 'a b'.

A key attraction of Scryer Prolog is that it can be reliably used to determine what is Prolog syntax and what is not, so you can easily use it to see whether an atom needs quoting.

Enjoy!
Markus

[hurufu]

Is syntax described formally in some document other then ISO Prolog specification? Like how does it know that ++...-- is a single atom and not +(+...--) or +(+, ...---)?

[triska]

Very good question!

The syntax of tokens is specified in subclause 6.4 of the Prolog ISO standard:

6.4 Tokens

Lexically, the syntax of Prolog terms (6.3) shall be a
sequence of tokens. This subclause defines how characters
are combined to form tokens, and the tokens to form terms
and read-terms.

...

A token shall not be followed by characters such that
concatenating the characters of the token with these
characters forms a valid token as specified by the above
syntax.

The text highlighted above is the eager consumer rule. This is important to understand: A token always extends as far as possible. For instance, 123.e defines the tokens 123 . e, but 123.4 defines the single token 123.4. A layout text is sometimes necessary to separate two tokens.

As one of the tokens defined in 6.4, let's consider this one:

name (* 6.4 *)
   = [ layout text sequence (* 6.4.1 *) ] ,
     name token (* 6.4.2 *) ;

combined with:

6.4.2 Names

name token (* 6.4.2 *)
   = letter digit token (* 6.4.2 *)
   | graphic token (* 6.4.2 *)
   | quoted token (* 6.4.2 *)
   | semicolon token (* 6.4.2 *)
   | cut token (* 6.4.2 *) ;

...

graphic token (* 6.4.2 *)
   = graphic token char (* 6.4.2 *),
     { graphic token char (* 6.4.2 *) } ;

graphic token char (* 6.4.2 *)
   = graphic char (* 6.5.1 *)
   | backslash char (* 6.5.5 *) ;

and finally:

6.5.1 Graphic characters

graphic char (* 6.5.1 *)
   = "#" | "$" | "&" | "*" | "+" | "-" | "."
   | "/" | ":" | "<" | "=" | ">" | "?" | "@"
   | "^" | "~" ;

[hurufu]

Thanks for explanation, so that means that I have to fix vim syntax highlighting too.