Why "at" is not a reserved word?

[cpicanco]

This example is given here: https://www.freepascal.org/docs-html/ref/refse112.html

raise Exception.Create(msg) at
        get_caller_addr(get_frame),
        get_caller_frame(get_frame);

I was wondering why the "at" is not a reserved word. If not, what it is?

[marcov]

Usually only words that start  a block (like in this case "raise") are reserved.

Words further down the line are not likely to raise conflicts (because you can't call e.g. an identifier there) and thus there is no reason to exclude them from the list of possible identifiers.

[Thaddy]

A.K.A. a compiler intrinsic. These do not always represent a reserved word.
Compiler intrinsics are basically operators. functions and procedures which have a greater range of syntactic freedom from which the complexity is nicely hidden from us.

[marcov]

No, not like intrinsics. Intrinsics are like predefined procedures, but with internal implementation (and usually relaxation of typing0 and thus appear in places where identifiers also would appear.

AT is at a place where a normal identifier should appear. It is part of a block construct, but not a FIRST token.

[Thaddy]

AT is at a place where a normal identifier should appear. It is part of a block construct, but not a FIRST token.

Operators are not a first token... I merely explained how and why in simple terms.

[marcov]

I don't see intrinsics as operators, as said they use pseudo function call syntax, rather than infix expression operators.

Things like the ROR intrinsic doesn't really need special treatment in the parser (except to pass through), all the magic happens in the backend.