PostScript == operator

Question

The value of == is not an operator, but rather a built-in procedure.

PLRM page 526 (pdf page 540).

a built-in procedure, does this mean that this is a PostScript procedure:

{ something here }

or an actual function e.g. in Java that you have to define yourself?

Answer 1

You can peek at ghostscript's implementation with /== load ==.

Or get the book Inside Postscript by Frank Meritt Braswell. It has an entire chapter about how the == procedure works in Adobe's printer implementations.

But the simple framework is based around the properties of the type operator which yields a typename. More specifically, it yields an executable name which designates the type of the argument. So you can easily implement a type-switch with a dictionary.

<<
    /integertype { } % handle integer case
    /realtype { }    % handle floating-point case
    /arraytype { }   % handle array case
>> begin
5 type exec
2.0 type exec
{a b c} type exec

The different typed objects cause different procedures to be executed.

Using this, we can handle the different types that might be passed to the procedure.

/my== {
     <<
     /integertype { =string cvs print }
     /realtype { =string cvs print }
     /arraytype { dup xcheck {
                      ({ ) print {my==} forall (} ) print
                  }{
                      ([ ) print {my==} forall (] ) print
                  } }
     >> begin dup type exec
} def

=string (described in the book) is a pre-allocated 128-byte scratch buffer used by the = and == procedures for this exact purpose: supplying the argument to cvs.