Using MARKER within a colon definition?

[SirWumpus]

I'd like to define a word that sets a marker at the start of a group of tests and clean-up those tests at the end of the group. Something like:

: test_group ... MARKER rm_test_group ... ;
: test_group_end ... rm_test_group ;

However MARKER wants to add a word to the word list, which makes it problematic while in the middle of a word definition. Is there some trick or method I'm missing that could achieve this? Or is it just not possible?

[paraplegic]

Perhaps simply using the HERE word to put the current address on the stack?

…

On Thu, 26 Jan 2023, Anthony Howe wrote: Date: Thu, 26 Jan 2023 12:05:08 -0800 From: Anthony Howe ***@***.***> Reply-To: ForthHub/discussion ***@***.***> To: ForthHub/discussion ***@***.***> Cc: Subscribed ***@***.***> Subject: [ForthHub/discussion] Using MARKER within a colon definition? (Discussion #133) I'd like to define a word that sets a marker at the start of a group of tests and clean-up those tests at the end of the group. Something like: : test_group ... MARKER rm_test_group ... ; : test_group_end ... rm_test_group ; However MARKER wants to add a word to the word list, which makes it problematic while in the middle of a word definition. Is there some trick or method I'm missing that could achieve this? Or is it just not possible? ? Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you are subscribed to this thread.[AAS2KEUZMYKW4XC4VEUWCWLWULKHJA5CNFSM6AAAAAAUH5BVMSWGG33NNVSW45C7OR4XAZNKIRUXGY3VONZWS33OVJRW63LNMVXHIX3JMTHAASJN2U.gif] Message ID: ***@***.***>

-- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Robert S. Sciuk ***@***.*** Principal Consultant Phone: 289.312.1278 Cell: 905.706.1354 Control-Q Research 97 Village Rd. Wellesley, ON N0B 2T0

[SirWumpus]

I don't see how that would help / work.

[ruv]

A simple solution is to use evaluate:

: test_group ... "marker rm_test_group" evaluate ... ;
: test_group_end ... "rm_test_group" evaluate  ;

Or execute-parsing in the first case:

: test_group "rm_test_group" ['] marker execute-parsing ... ;

---

The original code:

: test_group ... MARKER rm_test_group ... ;
: test_group_end ... rm_test_group ;

is not correct due to the following problems:

1. The word marker is an ordinary word. Therefore it doesn't parse anything when it's used inside a definition (i.e., when it's compiled), — just like constant and : (colon). So the Forth text interpreter will encounter rm_test_group and raise an error.
2. When test_group_end is being defined, rm_test_group has not been defined yet, so an error will be raised.

[SirWumpus]

Ah. I knew I'd forgotten something, like evaluate. That method works nicely and seems the cleanest expression.

execute-parsing isn't a standard word or implies a particular implementation, which is less portable.

[gerryjackson]

If I understand what you're asking for you could do: defer do-marker : test-group >in @ marker >in ! ' ['] do-marker defer! ; : test-group-end do-marker  ; \ Example wordlist constant mk-wl    \ to see tests defined and removed mk-wl >order definitions test-group foo : test1 ( ... ) ; : test2 ( ... ) ; : test3 ( ... ) ; words test-group-end words \ Works in Gforth Gforth 0.7.9_20180905, Copyright (C) 1995-2017 Free Software Foundation, Inc. Gforth comes with ABSOLUTELY NO WARRANTY; for details type `license' defer do-marker  : test-group >in @ marker >in ! ' ['] do-marker defer! ; : test-group-end do-marker  ;  \ Example Gforth 0.7.9_20180905, Copyright (C) 1995-2017 Free Software Foundation, Inc. Gforth comes with ABSOLUTELY NO WARRANTY; for details type `license' Type `help' for basic help wordlist constant mk-wl    \ not needed - to see tests defined and removed mk-wl >order definitions test-group foo : test1 ( ... ) ; : test2 ( ... ) ; : test3 ( ... ) ; words      \ Display definitions in mk-wl test3 test2 test1 foo test-group-end words       \ Ditto - test group removed \ Alternatively more simply : test-group marker ; : test-group-end  ' execute  ; \ Usage test-group foo \ Test code as above end-test-group foo

…

On 26/01/2023 20:05, Anthony Howe wrote: I'd like to define a word that sets a marker at the start of a group of tests and clean-up those tests at the end of the group. Something like: |: test_group ... MARKER rm_test_group ... ; : test_group_end ... rm_test_group ; | However |MARKER| wants to add a word to the word list, which makes it problematic while in the middle of a word definition. Is there some trick or method I'm missing that could achieve this? Or is it just not possible? — Reply to this email directly, view it on GitHub <#133>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADH775NGSTDNXEKKHKDCC4TWULKHXANCNFSM6AAAAAAUH5BVMQ>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[ruv]

@gerryjackson, nice solution!

Though, as I can see, the idea of Anthony was to have a kind of anonymous lexical blocks of tests:

test_group
  ...
test_group_end

They also work as expected if they are nested.
NB: since test_group_end is a word that should be encountered by the Forth text interpreter, the contained tests shall not shadow this word or make it unfindable (due to changes in the search order).

[gerryjackson]

I realised soon after I'd posted, that in my first solution the word do-marker is unnecesssary. Just have defer test-group-end and replace do-marker in the definition of test-group with test-group-end.

…

On 26/01/2023 23:22, ruv wrote: @gerryjackson <https://github.com/gerryjackson>, nice solution! Though, as I can see, the idea of Anthony was to have a kind of anonymous lexical blocks of tests: |test_group ... test_group_end | They also work as expected if they are nested. NB: since |test_group_end| is a word that should be encountered by the Forth text interpreter, the tests shall not shadow this word or make it unfindable (due to changes in the search order). — Reply to this email directly, view it on GitHub <#133 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADH775JC2HPJ2ZRCYWCLXADWUMBMPANCNFSM6AAAAAAUH5BVMQ>. You are receiving this because you were mentioned.Message ID: ***@***.***>