Re: Implementation of LEAVE?

From	"HAA" <someone@microsoft.com>
Newsgroups	comp.lang.forth
Subject	Re: Implementation of LEAVE?
Date	2016-06-26 15:31 +1000
Organization	Aioe.org NNTP Server
Message-ID	<nknp7r$12ei$1@gioia.aioe.org> (permalink)
References	<cf94fe95-4400-43db-8981-0c24d63e0749@googlegroups.com>

Show all headers | View raw

alexinbeijing@gmail.com wrote:
> Dear comp.lang.forth,
>
> Greetings! I'm hoping one of you Forth wizards can enlighten me on the most practical
> way to implement (ANS Forth) LEAVE. (I'm implementing Forth as a way of learning about
> it.)
>
> IF..ELSE..THEN, BEGIN..AGAIN, BEGIN..WHILE..REPEAT, etc. all use the data stack to
> store unresolved jump origins and destinations. But because LEAVE follows a different
> nesting discipline from all of these other control flow structures, it can't use the
> data stack to store its jump origin, or things get messed up when (for example) you put
> a LEAVE inside an IF..THEN block.
>
> I have read that some Forths use the retstack for storing LEAVE jump origins, then
> resolve all of them when the corresponding LOOP or +LOOP is reached. This seems to make
> sense.
>
> But then I have also read things like this:
>
>     (http://www.figuk.plus.com/build/heart.htm)
>     The behaviour of ?DO and LEAVE has led to some strange
>     implementations. For example, F83 DO pushes the exit
>     address of the loop on the return stack so that it can
>     be picked up and used by LEAVE, and it does so whether
>     that particular loop has a LEAVE or not. This version
>     is more rational. LEAVE compiles its own exit address
>     and doesn't mess with the return stack.

It doesn't need to be complicated.  This one is for ITC/DTC.  Compiler security
checking not shown.

0 value leav
: DO  postpone (do)  leav  >mark dup to leav  postpone begin ; immediate
: ?DO  postpone (?do)  leav  >mark  dup to leav  postpone begin ; immediate
: LEAVE  postpone (leave)  leav , ; immediate
: LOOP  postpone (loop)  <resolve  postpone then  to leav ; immediate
: +LOOP  postpone (+loop)  <resolve  postpone then  to leav ; immediate

Thread

Implementation of LEAVE? alexinbeijing@gmail.com - 2016-06-23 02:01 -0700
  Re: Implementation of LEAVE? Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-06-23 05:45 -0500
    Re: Implementation of LEAVE? anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-06-23 11:53 +0000
      Re: Implementation of LEAVE? Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-06-23 12:06 -0500
        Re: Implementation of LEAVE? Paul Rubin <no.email@nospam.invalid> - 2016-06-23 12:19 -0700
          Re: Implementation of LEAVE? Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-06-24 03:05 -0500
            Re: Implementation of LEAVE? JennyB <jennybrien@googlemail.com> - 2016-06-24 02:14 -0700
              Re: Implementation of LEAVE? Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-06-25 04:19 -0500
                Re: Implementation of LEAVE? JennyB <jennybrien@googlemail.com> - 2016-06-25 03:38 -0700
                Re: Implementation of LEAVE? Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-06-25 11:00 -0500
                Re: Implementation of LEAVE? anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-06-25 17:34 +0000
                Re: Implementation of LEAVE? Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-06-26 12:14 -0500
                Re: Implementation of LEAVE? hughaguilar96@gmail.com - 2016-06-26 16:36 -0700
                Re: Implementation of LEAVE? hughaguilar96@gmail.com - 2016-06-26 17:57 -0700
                Re: Implementation of LEAVE? Alex McDonald <alex@rivadpm.com> - 2016-06-27 14:16 +0100
                Re: Implementation of LEAVE? anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-06-27 16:12 +0000
                Re: Implementation of LEAVE? Alex McDonald <alex@rivadpm.com> - 2016-06-27 22:31 +0100
                Re: Implementation of LEAVE? anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-06-27 14:21 +0000
                Re: Implementation of LEAVE? Andrew Haley <andrew29@littlepinkcloud.invalid> - 2016-06-27 11:05 -0500
              Re: Implementation of LEAVE? JennyB <jennybrien@googlemail.com> - 2016-06-25 03:49 -0700
            Re: Implementation of LEAVE? JennyB <jennybrien@googlemail.com> - 2016-06-24 08:45 -0700
            Re: Implementation of LEAVE? JennyB <jennybrien@googlemail.com> - 2016-06-24 02:33 -0700
        Re: Implementation of LEAVE? anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-06-24 16:23 +0000
  Re: Implementation of LEAVE? anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-06-23 11:43 +0000
  Re: Implementation of LEAVE? alexinbeijing@gmail.com - 2016-06-24 11:41 -0700
    Re: Implementation of LEAVE? anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-06-25 05:18 +0000
      Re: Implementation of LEAVE? Brad Eckert <hwfwguy@gmail.com> - 2016-06-27 17:45 -0700
        Re: Implementation of LEAVE? anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-06-28 06:14 +0000
  Re: Implementation of LEAVE? Brad Eckert <hwfwguy@gmail.com> - 2016-06-24 15:30 -0700
  Re: Implementation of LEAVE? minforth@arcor.de - 2016-06-24 04:36 -0700
  Re: Implementation of LEAVE? Mark Wills <markwills1970@gmail.com> - 2016-06-25 07:51 -0700
    Re: Implementation of LEAVE? Coos Haak <chforth@hccnet.nl> - 2016-06-26 00:37 +0200
  Re: Implementation of LEAVE? Mark Wills <markwills1970@gmail.com> - 2016-06-25 08:29 -0700
  Re: Implementation of LEAVE? "HAA" <someone@microsoft.com> - 2016-06-26 15:31 +1000
    Re: Implementation of LEAVE? anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2016-06-26 10:16 +0000
      Re: Implementation of LEAVE? "HAA" <someone@microsoft.com> - 2016-06-27 22:38 +1000
  Re: Implementation of LEAVE? Alex Dowad <alexinbeijing@gmail.com> - 2016-06-26 13:53 -0700

csiph-web