|
|
 | | From: | Alex McDonald | | Subject: | Default & non-default compilation semantics | | Date: | 17 Jan 2005 14:11:14 -0800 |
|
|
 | In
http://www.complang.tuwien.ac.at/forth/dpans-html/comment-semantics.html
in the Implementations strategies section, there's a statement that
intrigues (that is, confuses) me. In a traditional xt is CFA, immediate
flag forth;
"Words without interpretation semantics and with default compilation
semantics (e.g., EXIT) can be implemented as non-immediate; you cannot
implement them as immediate words if you support [COMPILE]."
Does this mean "you cannot [COMPILE] EXIT" if EXIT is immediate?
Clarification of the definition of "non-default compilation semantics"
might help; I can't see the difference between IF and EXIT. Neither
have interpretation semantics; why are the compilation semantics of IF
non-default?
--
Regards
Alex McDonald
|
|
| | From: | Anton Ertl | | Subject: | Re: Default & non-default compilation semantics | | Date: | Mon, 17 Jan 2005 22:29:11 GMT |
|
|
| "Alex McDonald" writes:
>In
>http://www.complang.tuwien.ac.at/forth/dpans-html/comment-semantics.html
>in the Implementations strategies section, there's a statement that
>intrigues (that is, confuses) me. In a traditional xt is CFA, immediate
>flag forth;
>
>"Words without interpretation semantics and with default compilation
>semantics (e.g., EXIT) can be implemented as non-immediate; you cannot
>implement them as immediate words if you support [COMPILE]."
>
>Does this mean "you cannot [COMPILE] EXIT" if EXIT is immediate?
What the system will do in that case is not what the standard requires
(though I don't think that the TC did this intentionally; it's rather
a side effect of the fact that [COMPILE] does not quite fit the
semantics model).
To make this more concrete, consider two simple implementations of
EXIT:
' ;s alias EXIT1
: EXIT2 POSTPONE ;s ; immediate
Now consider what happens when you [COMPILE] these EXITs. There is a
difference between these EXITs. The way EXIT and [COMPILE] are
defined in the standard, the required behaviour is to behave like
[COMPILE] EXIT1.
>Clarification of the definition of "non-default compilation semantics"
>might help; I can't see the difference between IF and EXIT. Neither
>have interpretation semantics; why are the compilation semantics of IF
>non-default?
The definitions of these words in the standard are different:
- For IF, no execution semantics is defined (and therefore there is no
default compilation semantics), and compilation semantics is defined
directly (not through the default mechanism). I consider this to be
non-default compilation semantics (and that's necessary for [COMPILE]
IF to work in the traditional way).
- For EXIT, execution semantics is defined, and the compilation
semantics is the default compilation semantics defined in 3.4.3.3.
- anton
--
M. Anton Ertl http://www.complang.tuwien.ac.at/anton/home.html
comp.lang.forth FAQs: http://www.complang.tuwien.ac.at/forth/faq/toc.html
New standard: http://www.complang.tuwien.ac.at/forth/ansforth/forth200x.html
|
|
|
